text
stringlengths 649
4.42k
| synonym_substitution
stringlengths 759
4.5k
| butter_fingers
stringlengths 649
4.42k
| random_deletion
stringlengths 453
2.31k
| change_char_case
stringlengths 649
4.42k
| whitespace_perturbation
stringlengths 764
5.02k
| underscore_trick
stringlengths 649
4.42k
|
|---|---|---|---|---|---|---|
call any one of these rule-sets the *sRules*, indicating that they are auxiliary rules that accompany standard quantum mechanics.
The Born rule is not necessary. It is possible to achieve very desirable results by dropping the Born rule from the governing auxiliary rules and introducing probability into quantum mechanics through *probability current* alone. This has been done with two very different auxiliary rule-sets called the *nRules* and the *oRules*. In this paper we are only concerned with the nRules.
The nRules {#the-nrules.unnumbered}
==========
For the most part, solutions to Schrödinger’s equation represent a continuous and classical-like evolution of a quantum mechanical wave function. Occasionally this evolution engages a state that is discontinuous with its immediate predecessor in some variable. This may be called a quantum *jump* or quantum *gap*. The nRules are concerned with what happens at one of these gaps.
We choose the nRules with two objectives in mind. First, the rules should locate any measurement inside of the system being measured, and should apply equally to all systems independent of size – microscopic or macroscopic. A measurement is understood to mean a stochastically initiated collapse of the wave that occurs independent of an ‘external’ measuring device or observer. Second, the rules should apply to individual trials, not just to ensembles of trials. When these objectives are met, we find that the system is able to include conscious observers without the ambiguity associated with the Schrödinger cat experiment, or with the many worlds of Everett.
The first of these rules recognizes two different kinds of states: *ready states* and *realized states*. A ready state $\underline{S}$ is underlined, and a realized state $S$ is not. The first rule tells us that ready states are introduced into solutions of Schrödinger’s equation immediately after a quantum jump. A state is otherwise realized. We define a *ready component* as one that includes a ready state. It too exists immediately after a quantum jump.
**nRule (1)**: *If an irreversible interaction produces a complete component that is discontinuous with its predecessor in some variable, then it is a ready component.*
Complete components contain all of the symmetrized objects in the universe. Each included object is itself complete in that it is not just a partial expansion in some representation. The ready states in a discontinuous component are the discontinuous states appearing there.
The second rule establishes the existence of a stochastic | call any one of these rule - sets the * sRules *, indicating that they are accessory rule that accompany standard quantum mechanics.
The digest principle is not necessary. It is possible to achieve very desirable results by drop the Born rule from the regulate auxiliary rules and insert probability into quantum mechanics through * probability current * alone. This has been done with two very different auxiliary rule - hardening called the * nRules * and the * oRules *. In this paper we are only implicated with the nRules.
The nRules { # the-nrules.unnumbered }
= = = = = = = = = =
For the most part, solutions to Schrödinger ’s equality represent a continuous and authoritative - like evolution of a quantum mechanical wave function. Occasionally this evolution engages a state that is discontinuous with its contiguous predecessor in some variable. This may be called a quantum * jump * or quantum * gap *. The nRules are concerned with what happen at one of these gap.
We choose the nRules with two objectives in mind. First, the rules should locate any measurement inside of the system being measure, and should use equally to all systems autonomous of size – microscopic or macroscopic. A measurement is understand to think of a stochastically initiated collapse of the wave that occurs independent of an ‘ external ’ measuring device or observer. Second, the rules should enforce to individual trials, not just to ensembles of trials. When these objectives are converge, we find that the system is able to include conscious observers without the ambiguity associated with the Schrödinger cat experiment, or with the many populace of Everett.
The first of these rules recognizes two different kinds of country: * ready states * and * realized state *. A ready state $ \underline{S}$ is underlined, and a realized country $ S$ is not. The first rule state us that ready state are introduced into solution of Schrödinger ’s equation immediately after a quantum startle. A state is otherwise realized. We define a * ready part * as one that includes a ready state. It besides exists immediately after a quantum jump.
* * nRule (1) * *: * If an irreversible interaction produces a complete component that is discontinuous with its harbinger in some variable, then it is a ready part. *
Complete components contain all of the symmetrize objects in the universe. Each included object is itself complete in that it is not just a partial expansion in some representation. The ready states in a discontinuous component are the discontinuous state appear there.
The second convention establishes the existence of a stochastic | capl any one of these rule-rets the *sRules*, indicaving thzt they xre auxiliary rules that acclmpany wtandard quantum mechavics.
The Blrn rule is iot necessary. It is posslyle tk achnete very desirabke results by dropping tve Blrn rule from the governing auxiliawy rulex wnd introducind prpfabimptn into quantum mechanics through *probabplity current* alome. This has been done with two very different audiliary rulg-aetf called the *vRules* and the *oRules*. Jn this paper we are only concefned cith the nRolss.
Hve nRules {#tie-nrulvs.unnumbered}
==========
For the mmst pary, solutions to Scirödibger’s equation represxnt a continuous and classican-lnke evolution of a quqnrum mgchanhcal qavd fhnrtikn. Occwsiknally thia evolution engages a state thst px discontinukus wieh its immediate predecessor in some varidbls. This may be called a wuantum *jump* or quantom *gap*. The nRules are concerned with what happens at one of thesx eapw.
Wc chuisf the nRules with two objectives in mind. Firse, tne rules should jocate any kewsitement inside uf the syatem being measuref, and srould apply eqtallu to all systems independenr of size – mpcrowcopic or macrosco'ic. A measurzment os uncerstood to mean a stockasticzlly initiahed collalre of the wave tfat owcurs independent of an ‘exeernal’ meesuriug devicd or obserder. Second, the vgles should apply ho inbividgal trials, not just to ensembles of trials. When these onjactpves are iet, wc find that the system is ablg to inclbde covscious obaervers without thq ambiguity avdociated witi the Schwödinter xat expdfiment, or with the many worlds of Everett.
The first pf jhsse rules recogunzts two different kivds ov xtwdes: *ready stdtes* ana *realkzed states*. X resdy state $\underline{S}$ is hnderlined, and a rrajized stqte $S$ is not. The firsy rule tells us thwt reedy stetes ate yntroduced into solutions of Sdhrödinger’d ezuation immeqiatcly wfter a qucntum jump. A state is otherwise realized. We define a *readi component* as one thct includes a readr state. Id too exists immediarely after a quanuum jump.
**nRule (1)**: *If an irdevershble lnteraction produces a complete component that is discontinuous with its predecxsfor in some varoable, tken it is a rxavy component.*
Complvte components contain all of thx symmetriseb objects in the universe. Eacn kncluded objezt is itself complete in that it is not just a partial expansiom in some representation. The reaey stztes in a viscontinuous componrnt arz the eiscontunuoms states appexrimg there.
The secpnb rule establishes the existence if a srofhastic | call any one of these rule-sets the that are auxiliary that accompany standard is necessary. It is to achieve very results by dropping the Born rule the governing auxiliary rules and introducing probability into quantum mechanics through *probability current* This has been done with two very different auxiliary rule-sets called the *nRules* the In paper are only concerned with the nRules. The nRules {#the-nrules.unnumbered} ========== For the most part, solutions to equation represent a continuous and classical-like evolution of quantum mechanical wave function. this evolution engages a state is with its predecessor some This may be a quantum *jump* or quantum *gap*. The nRules are concerned with what happens at one of these We choose with two in First, rules should locate inside of the system being measured, equally to all systems independent of size – or macroscopic. measurement is understood to mean a initiated collapse of the wave that occurs independent an ‘external’ measuring device or observer. Second, the rules should apply to individual trials, not ensembles of trials. When objectives are met, find the is to include observers without the ambiguity associated with the Schrödinger cat experiment, or the many worlds of Everett. The first of these rules different of states: *ready and *realized states*. A state is underlined, and a $S$ not. tells that states are introduced into of Schrödinger’s equation immediately after quantum jump. A state a *ready component* as one that includes a state. It too exists immediately after a jump. **nRule (1)**: *If an irreversible interaction produces a complete component that discontinuous with in some variable, then it is a ready Complete components contain all the symmetrized objects in the universe. Each included object itself in that is not just partial expansion in representation. The ready a discontinuous are discontinuous The second rule establishes the existence a | call any one of these rule-sets The *sRules*, iNdicaTinG thAt They Are aUxiliary rules tHAt acCompany standard quantum MechaNiCS.
The bOrN rule Is not neCEsSARy. IT iS pOssIbLE tO achiEve Very desIrable resuLts By Dropping the BORn Rule from thE goVerning auxilIarY rules AnD inTRoducIng ProbaBility INto quaNtum mechaNiCS throuGH *probabILItY curRent* alone. This has bEEn DOne with two very DifferEnT AuXILiaRy rUle-sets calLeD the *nrUles* and THe *OrULes*. iN this paper we aRe only conceRNed With thE nrulES.
The nRUles {#tHe-NRulEs.unnumbereD}
==========
For The most paRt, soluTIons to SCHrödingEr’s equAtiOn rEpreSEnT a ConTiNUouS AnD clASsiCal-like eVoLuTion oF a quANTUM mecHanIcal Wave fUnction. OccasiOnaLly tHIs eVolutIon enGageS a State That is DiscoNtInuous with its imMediAte predecEssOr In sOmE variABle. ThiS maY be Called a Quantum *JUmp* Or QUANtUm *gap*. The nRules are cOnCERnEd with whAt happENs At ONe of thesE gAps.
we chOOSe the NRulES wIth two obJectivES iN mInd. FirsT, tHe ruleS sHouLd lOcate ANy meAsuremEnt insidE of thE System being meaSUred, and should APpLY EqUAlly To aLl systems inDepeNDent Of siZE – mIcrOScopiC or maCrOScOPic. A measurement is unDeRstood To meaN a stochasticaLly initiatED COllapse oF the WAvE That occurs indePendeNt of an ‘exteRNal’ measuRing dEvice or oBserver. SeCONd, the rulEs sHouLd aPplY TO iNdividual triaLS, Not jUsT to enseMblEs of triAls. wheN thEse ObJectives aRe met, we fInD tHaT tHe sYstem IS able to iNcLudE cOnsCious OBserveRs witHout ThE aMBigUity assOCiATEd wiTh ThE SchRödInGer caT expERimEnt, or wiTh the many WorLDs of evErEtt.
The fIrst of these ruLeS recognizeS tWo dIffereNT Kinds of sTates: *ready states* and *realIZed statEs*. A Ready StatE $\underlinE{S}$ iS underLinED, and a rEalizeD statE $S$ Is nOT. the fiRST rUle TeLls us that rEADy sTates ArE intRoduced Into solutions of SchRÖdiNger’s equation ImmEdiaTELy AftER a QUanTuM JumP. a State is otherwisE realized. WE dEFiNe a *ready coMPonEnT* as one tHat inclUdes a REady staTe. It too exIsts immedIaTely AFTer A quantum juMp.
**nRule (1)**: *IF an irreveRSible INtEractIon ProducEs A coMpletE compoNEnt That iS discoNtInuous With iTs PredecesSor in some variable, then it Is a reaDy comPonEnt.*
CompleTe cOMpoNents contAin aLl of the symMetRizEd objEctS In the UnivERsE. EaCH inclUded OBject is itSElF coMPLeTe in that it iS NOT juSt a paRtiAL expanSion In some representatIOn. The ready statEs in A DIscOntINuouS cOmponent are the DisCoNTInuous stAtEs appearing There.
The SeCOnd ruLe estaBlisheS the exiSTEnCE of a stOchaStiC | call any one of these rul e-sets the *sRu les *,in dica ting that they are auxi liary rules that accom panyst a ndar d q uantu m mecha n ic s .
T he B orn r u le is n otnecessa ry. It ispos si ble to achie v every desir abl e results by dr opping t heB orn r ule from the g o vernin g auxilia ry rulesa nd intr o d uc ingprobability intoq ua n tum mechanicsthroug h* pr o b abi lit y current* a lone. This ha s b e e n do n e with two ve ry differen t au xiliar yrul e -setscalle dt he*nRules* an d th e *oRules *. Int his pap e r we ar e only co nce rned wi th th en Rul e s.
T h e n Rules {# th e- nrule s.un n u m b ered }
= ==== =====
For the mos t p art, sol ution s toSchr öd inger ’s equ ation r epresent a cont inuo us and cl ass ic al- li ke ev o lution of aquantum mechan i cal w a v e f unction. Occasiona ll y th is evolu tion e n ga ge s a state t hat isd i scont inuo u swith its immed i at epredece ss or inso mevar iable . Thi s maybe calle d a q u antum *jump* o r quantum *gap * .T h en Rule s a re concerne d wi t h wh at h a pp ens at on e ofth e se gaps.
We choose th enRules with two objectiv es in mind . F irst, th e ru l es should locateany m easurement inside o f the systembeing mea s u red, and sh oul d a ppl y eq ually to alls y stem sindepen den t of si ze– m icr osc op ic or mac roscopic .Ame as ure menti s unders to odto me an as tochas tical ly i ni ti a ted collap s eo f the w av e th atoc cursinde p end ent ofan ‘exter nal ’ mea su ri ng devi ce or observe r. Second, t he ru les sh o u ld apply to individual trials,n ot just to ense mble s of tria ls. Whenthe s e obje ctives areme t,w e find t ha t t he system is a ble to i nc lude consci ous observers with o utthe ambiguity as soci a t ed wi t ht heSc h röd i n ger cat experim ent, or wi th th e many wor l dsof Everet t.
The firs t of the se rulesrecognize stwod i ffe rent kinds of stat es: *read y stat e s* and*re alized s tat es*.A read y st ate $ \under li ne{S}$ is u nd erlined, and a realized state $ S$ isnot.The first ru let ell s us that rea dy statesare in trodu ced intosolu t io nso f Sch rödi n ger’s equ a ti oni m me diately aft e r a q uantu m j u mp. Astat e is otherwise re a lized. We defi ne a * rea dyc ompo ne nt* as one tha t i nc l u des a re ad y state. It too exi st s imme diatel y afte r a qua n t um jump.
**n Rul e (1)**:*If a n irreve rs ib l e inte ract io n prod uces a comp l e te component tha t isd i scont i nuo us wi th its pr e dece ssor in so me variable , then itis aready c om ponent .*
C omplete co m ponents c ontai n all o fthesym metriz ed o b j ectsin t he un iverse. E a c hi nc lu d edobje ct is i tsel f complet e in that it is notju sta partia l e x p ansion insom e rep r e sentation. Ther ea d y sta tes in a dis continu o usco mponent ar e the disco ntinuouss tat es app earing t he re.
T hese cond rule estab lishes t he exist e n ce ofa s t ocha s tic | call_any one_of these rule-sets the_*sRules*, indicating_that_they are_auxiliary_rules that accompany_standard quantum mechanics.
The_Born rule is not_necessary. It is_possible_to achieve very desirable results by dropping the Born rule from the governing auxiliary_rules_and introducing_probability_into_quantum mechanics through *probability current*_alone. This has been done_with two_very different auxiliary rule-sets called the *nRules* and_the_*oRules*. In this_paper we are only concerned with the nRules.
The nRules_{#the-nrules.unnumbered}
==========
For the most part, solutions to_Schrödinger’s equation represent_a_continuous_and classical-like evolution of_a quantum mechanical wave function. Occasionally_this evolution engages a state that_is discontinuous with its immediate predecessor in_some variable. This may be called_a quantum *jump* or quantum_*gap*. The_nRules are concerned with what_happens at one_of these_gaps.
We choose the_nRules with two objectives in mind._First, the rules_should locate any measurement inside of_the_system being measured,_and_should_apply equally_to all systems_independent_of size_–_microscopic or macroscopic. A measurement is_understood_to mean a stochastically initiated collapse of_the wave that occurs_independent_of an ‘external’ measuring_device or observer. Second, the_rules should apply to individual trials,_not just_to ensembles_of trials. When these objectives are met, we find that the_system is able to include conscious_observers without the ambiguity_associated with_the_Schrödinger cat experiment,_or_with the_many worlds of Everett.
The first of these_rules recognizes_two different kinds of states: *ready_states* and *realized states*._A_ready state $\underline{S}$ is underlined, and_a realized state $S$ is not._The first rule tells us_that_ready_states are introduced into solutions_of Schrödinger’s equation immediately after a_quantum jump. A_state is otherwise realized. We define a_*ready_component* as one that includes a_ready_state. It too exists immediately after_a_quantum_jump.
**nRule (1)**: *If an irreversible_interaction produces a complete component that_is discontinuous with its predecessor in some variable, then_it is a_ready component.*
Complete components contain all_of_the_symmetrized objects in the universe. Each included object is itself_complete in_that it is_not just a partial expansion in some representation. The ready_states in a discontinuous component are the_discontinuous states appearing there.
The second rule establishes the existence of a_stochastic |
the wrong predictions for the $\sigma$-meson because the corresponding mass obtained is of an order $1/\bar\rho\sim 1~GeV$ (moreover, the adiabaticity approximation is certainly broken then). The result Eq. (\[simas\]) aggravates, in a sense, the situation since the hard component leaves for the region of harder masses whereas the light component persists to become lighter. On the other hand, the present observations signal rather the existence of the scalar meson of mass about $0.5~GeV$ what apparently does not coincide with the light component $M_2$. Nevertheless, our consideration having no claims of the quantitative agreement shows the scalar meson could be the mixed particle of pretty large width (similar to the superposition of oscillators with the various fundamental frequencies) describing the excitations of chiral and gluon condensates.
VI. Conclusion {#vi.-conclusion.unnumbered}
==============
In this paper we have developed the consistent approach to describe the interaction of quarks with IL. Theoretically, it is based (and justified) on the particular choice of the configurations saturating the functional integral what occurs to be not merely a technical exercise. They are the deformable (crampled) (anti-)instantons with the variable parameters $\gamma(x)$ and in the concrete treatment of this paper we play with the variation of the PP size $\rho(x,z)$. In a sense, such an ansatz is strongly motivated by the form of quark determinant which is solely dependent on the average instanton size in the SCSB theory. We have demonstrated that in the long-length wave approximation the variational problem of the deformation field optimization turns into the construction of effective Lagrangian for the scalar phononlike $\varphi$ and quark fields with the Yukawa interaction. Physically, it allows us to analyze the inverse influence of quarks on the instanton vacuum. We have pointed out this influence on the IL parameters as negligible. The modification of the SCSB parameters occurs pretty poor as well. In particular, the scale of quark condensate change amounts to a few $MeV$ only. Nevertheless, switching on the phononlike excitations of IL leads to several qualitatively new and interesting effects. The propagation of the quark condensate disturbances over IL happens in this approach to be in close analogy with well-known polaron problem. We imply a necessity to take into account the medium feedback while elementary excitations propagating. The intriguing conclusion comes with realizing it leads to generation of current quark masses $m_f$ within QCD itself with their values | the wrong predictions for the $ \sigma$-meson because the corresponding mass prevail is of an decree $ 1/\bar\rho\sim 1 ~ GeV$ (moreover, the adiabaticity approximation is certainly break then). The result Eq. (\[simas\ ]) aggravates, in a common sense, the situation since the hard component entrust for the region of harder batch whereas the light component persists to become lighter. On the other hand, the present observation signal rather the existence of the scalar meson of mass about $ 0.5 ~ GeV$ what obviously does not coincide with the light part $ M_2$. Nevertheless, our retainer having no claim of the quantitative agreement shows the scalar meson could be the mixed particle of reasonably large width (similar to the superposition of oscillators with the various fundamental frequency) describing the excitations of chiral and gluon condensates.
VI. Conclusion { # vi.-conclusion.unnumbered }
= = = = = = = = = = = = = =
In this paper we have developed the consistent approach to describe the interaction of quarks with IL. Theoretically, it is establish (and justified) on the particular choice of the shape saturate the functional integral what occurs to be not merely a technical exercise. They are the deformable (crampled) (anti-)instantons with the varying parameters $ \gamma(x)$ and in the concrete discussion of this paper we play with the variation of the PP size $ \rho(x, z)$. In a common sense, such an ansatz is strongly motivated by the form of quark determinant which is solely dependent on the median instanton size in the SCSB theory. We have demonstrated that in the long - length wave estimate the variational problem of the contortion battlefield optimization turns into the construction of effective Lagrangian for the scalar phononlike $ \varphi$ and quark fields with the Yukawa interaction. Physically, it allows us to analyze the inverse influence of quarks on the instanton vacuum. We have pointed out this influence on the IL parameters as negligible. The alteration of the SCSB parameters occurs pretty poor equally well. In particular, the scale of quark condensate change amounts to a few $ MeV$ only. however, switching on the phononlike excitations of IL leads to several qualitatively fresh and interesting effects. The generation of the quark condensate disturbances over IL happens in this approach to be in cheeseparing doctrine of analogy with well - known polaron problem. We imply a necessity to claim into account the medium feedback while elementary excitations propagate. The challenging conclusion comes with realize it leads to generation of current quark masses $ m_f$ within QCD itself with their values | thf wrong predictions for uhe $\sigma$-meson beequse tie corrsspondine mass obtained is of an ordxr $1/\bqr\rho\wim 1~GeV$ (moreover, the aaiabaticiny approxumatmon is certainly broken bken). Tgc resblv Eq. (\[simas\]) aggrsvates, in d sense, the siduxtnon since the hard component leaves sor the rfgion of hardet masxqs wgvrtas the light component persists fo becoke lighter. On the other hand, the presenh obdervations signal gather the gsisewnce of the rcalar meson of mass agout $0.5~GeV$ what apparently does nut conncide with tye phght componxnt $M_2$. Gevertheless, our convideratoon having no glaimv od the quantitative agceement shows the scwlar mesot eould be the mixed paetucle mf psetth lafge wmdtg (simipar to the sulerposition of oscillators witn eye various fuhdameneaj frequencies) describing the excitationv or chiral and gluon condwnsates.
VI. Conclusion {#gi.-conclusyon.unnumbered}
==============
In this paper we have developed the wonsiagenu qpproxxh to describe the interaction of quarks with IJ. Tnepretically, it ls based (and justigifd) jn the particolar choics of the configurahions swturaring the suncyional integral what occurs to be not mvrelt a technical exereise. They arz the ceforkable (crampled) (anti-)instcntons with the vwriable pzfameters $\gamma(x)$ xnd it the concrete treatment os this pa'er wz play wkth jhe varyation of hhe PI size $\rho(x,z)$. In a sfnse, duwh an ansahz is strongly motivated by the form of quark dategminant wkich ix solely depegdent on the arerage iustantun size in the SCVB theory. Wq have demonsdtated that in the lond-lentth qave apofoximation the variatioucl problen of the deformatipn wjeld optimizatiin uuens into the cpnsgrustpon os effective Ldgravgixm for the scalar ohonpnlike $\varphi$ and qudrk rields with the Yulara interqction. Prysically, it sllows us to analyee the interse onfjuence of quarks on the instanfon vacuul. Wc have pointeq oub thys influenee on the IL parameters as negligible. Thx modification of the SXSB parameters occuts kretty poor ax welj. In parthcular, the scale of wuark condensate ghange amounts to a fes $MeV$ mnly. Jevertheless, switching on the phononlike excitations of IL leads to severql qualitatively nsw amd innerzstnng effqcts. Vhe propagation og the quark condensate disturbaices over HL happens in this approach to bd in close anxlogy with well-known polzron pronlem. We imply a necessity to tale into account the medimm feedbaxk whjle elvjeitary excitations prppagatnng. Thw intrituinn conclusion cumex aith reallzinj ic leads to generation of current wuark nadses $m_f$ within QRD itself wibh theit values | the wrong predictions for the $\sigma$-meson because mass is of order $1/\bar\rho\sim 1~GeV$ certainly then). The result (\[simas\]) aggravates, in sense, the situation since the hard leaves for the region of harder masses whereas the light component persists to lighter. On the other hand, the present observations signal rather the existence of scalar of about what apparently does not coincide with the light component $M_2$. Nevertheless, our consideration having no claims the quantitative agreement shows the scalar meson could the mixed particle of large width (similar to the of with the fundamental describing excitations of chiral gluon condensates. VI. Conclusion {#vi.-conclusion.unnumbered} ============== In this paper we have developed the consistent approach to describe interaction of IL. Theoretically, is (and on the particular the configurations saturating the functional integral be not merely a technical exercise. They are deformable (crampled) with the variable parameters $\gamma(x)$ and the concrete treatment of this paper we play the variation of the PP size $\rho(x,z)$. In a sense, such an ansatz is strongly the form of quark which is solely on average size the SCSB We have demonstrated that in the long-length wave approximation the variational of the deformation field optimization turns into the construction of for scalar phononlike $\varphi$ quark fields with the interaction. it allows us to inverse of instanton We pointed out this influence the IL parameters as negligible. modification of the SCSB well. In particular, the scale of quark condensate amounts to a few $MeV$ only. Nevertheless, on the phononlike excitations of IL leads to several qualitatively new and effects. The the quark condensate disturbances over IL happens in approach to be in analogy with well-known polaron problem. We imply a necessity take account the feedback while elementary propagating. The intriguing comes with realizing to generation current masses itself with their values | the wrong predictions for the $\Sigma$-meson BecauSe tHe cOrRespOndiNg mass obtained IS of aN order $1/\bar\rho\sim 1~GeV$ (moreOver, tHe ADiabATiCity aPproximATiON Is cErTaInlY bROkEn theN). ThE result eq. (\[simas\]) aggRavAtEs, in a sense, thE SiTuation sinCe tHe hard componEnt Leaves FoR thE RegioN of HardeR masseS WhereaS the light CoMPonent PErsists TO BeCome Lighter. On the other HAnD, The present obseRvatioNs SIgNAL raTheR the existeNcE of thE Scalar mESoN OF MasS About $0.5~GeV$ what aPparently doES noT coincIdE wiTH the liGht coMpONenT $M_2$. NeverthelEss, oUr consideRation HAving no CLaims of The quaNtiTatIve aGReEmEnt ShOWs tHE sCalAR meSon could Be ThE mixeD parTICLE of pRetTy laRge wiDth (similar to tHe sUperPOsiTion oF osciLlatOrS with The varIous fUnDamental frequenCies) DescribinG thE eXciTaTions OF chiraL anD glUon condEnsates.
vi. CoNcLUSIoN {#vi.-conclusion.unnumBeRED}
==============
IN this papEr we haVE dEvELoped the CoNsiStenT APproaCh to DEsCribe the InteraCTiOn Of quarkS wIth IL. THeOreTicAlly, iT Is baSed (and JustifieD) on thE Particular choiCE of the configuRAtIONs SAturAtiNg the functiOnal INtegRal wHAt OccURs to bE not mErELy A Technical exercise. ThEy Are the DeforMable (crampled) (Anti-)instanTONS with the VariABlE Parameters $\gammA(x)$ and In the concrETe treatmEnt of This papeR we play wiTH The variaTioN of The pP sIZE $\rHo(x,z)$. In a sense, sUCH an aNsAtz is stRonGly motiVatEd bY thE foRm Of quark deTerminanT wHiCh Is SolEly dePEndent on ThE avErAge InstaNTon sizE in thE SCSb tHeORy. WE have deMOnSTRateD tHaT in tHe lOnG-lengTh waVE apProximaTion the vaRiaTIonaL pRoBlem of tHe deformation FiEld optimizAtIon Turns iNTO the consTruction of effective LagrANgian foR thE scalAr phOnonlike $\vArpHi$ and qUarK Fields With thE YukaWa IntERActioN. pHySicAlLy, it allows US To aNalyzE tHe inVerse inFluence of quarks on tHE inStanton vacuum. we hAve pOINtEd oUT tHIs iNfLUenCE On the IL parameteRs as negligIbLE. THe modificaTIon Of The SCSB ParametErs ocCUrs pretTy poor as wEll. In partIcUlar, THE scAle of quark CondensaTe change aMOunts TO a Few $Mev$ onLy. NeveRtHelEss, swItchinG On tHe phoNonlikE eXcitatIons oF Il leads to Several qualitatively new And intErestIng Effects. ThE prOPagAtion of thE quaRk condensaTe dIstUrbanCes OVer IL HappENs In tHIs appRoacH To be in cloSE aNalOGY wIth well-knowN POLarOn proBleM. we implY a neCessity to take into ACcount the mediuM feeDBAck WhiLE eleMeNtary excitatioNs pRoPAGating. ThE iNtriguing coNclusion CoMEs witH realiZing it Leads to GENeRAtion oF curRenT quark masSes $M_f$ WIthin QCd iTsELf with TheiR vAlues | the wrong predictions for the $\sig ma$-m eso n b ec ause the corresponding mass obtained is of an ord er $1 /\ b ar\r h o\ sim 1 ~GeV$ ( m or e o ver ,th e a di a ba ticit y a pproxim ation is c ert ai nly broken t h en ). The res ult Eq. (\[sima s\] ) aggr av ate s , ina s ense, the s i tuatio n since t he hard c o mponent l ea vesfor the region of ha r der masses whe reas t he li g h t c omp onent pers is ts to becomel ig h t e r.O n the other h and, the pr e sen t obse rv ati o ns sig nal r at h erthe existen ce o f the sca lar me s on of m a ss abou t $0.5 ~Ge V$what ap pa ren tl y do e snot coi ncide wi th t he li ghtc o m p onen t $ M_2$ . Nev ertheless, ou r c onsi d era tionhavin g no c laims of th e qua nt itative agreeme nt s hows thesca la r m es on co u ld bethe mi xed par ticle o f pr et t y la rge width (similar t o th e superp ositio n o fo scillato rs wi th t h e vari ousf un damental frequ e nc ie s) desc ri bing t he ex cit ation s ofchiral and glu on co n densates.
VI. Conclusion {# v i. - c on c lusi on. unnumbered}
=== = ==== ==== = =
In thispaper w e h a ve developed the co ns istent appr oach to descr ibe the in t e r action o f qu a rk s with IL. Theo retic ally, it i s based ( and j ustified ) on thep a rticular ch oic e o f t h e c onfigurations s atur at ing the fu nctiona l i nte gra l w ha t occursto be no tme re ly atechn i cal exer ci se. T hey aret he def ormab le ( cr am p led ) (anti - )i n s tant on swith th evaria blep ara meters$\gamma(x )$a nd i nth e concr ete treatment o f this pap er we playw i th the v ariation of the PP size $\rho(x ,z) $. In a s ense, suc h a n ansa tzi s stro ngly m otiva te d b y the f o r mofqu ark determ i n ant whic his s olely d ependent on the av e rag e instanton s ize int h eSCS B t h eor y. Weh a ve demonstrated that in t he lo ng-lengthw ave a pproxim ation t he va r iationa l problem of the d ef orma t i onfield opti mization turns in t o the co nstru cti on ofef fec tiveLagran g ian forthe sc al ar pho nonli ke $\varph i$ and quark fields wit h theYukaw a i nteractio n.P hys ically, i t al lows us to an aly ze th e i n verse inf l ue nce of qu arks on the in s ta nto n va cuum. We ha v e poi ntedout this i nflu ence on the IL pa r ameters as neg ligi b l e.The modi fi cation of theSCS Bp a rameters o ccurs prett y poor a sw ell.In par ticula r, thes c al e of qu arkcon densate c han ge amounts t oa few $ MeV$ o nly. N everth e less , switching on the phon o n likee xci tatio ns of ILl eads to severa l qualitati vely n ew a nd in teresti ng effec ts. T he propaga t ion of th e qua rk cond en sate di sturba nces o ver I L ha pp ens in thisa p pr o ac ht o b e in clos eanal ogy withw ell-know n p o laron p ro ble m . We im p ly a necessity to take i nto accoun t the me d ium f eedbac k whil e eleme n tar yexcitat ion s propagati ng. The i n tri gu ingconclusi on com es wi th realizing it l eads togenerati o n of cu rre n t qu a rk mas se s $m_f$ with in QC Di tself w i t h t heir va l ues | the_wrong predictions_for the $\sigma$-meson because_the corresponding_mass_obtained is_of_an order $1/\bar\rho\sim_1~GeV$ (moreover, the_adiabaticity approximation is certainly_broken then). The_result_Eq. (\[simas\]) aggravates, in a sense, the situation since the hard component leaves for_the_region of_harder_masses_whereas the light component persists_to become lighter. On the_other hand,_the present observations signal rather the existence of_the_scalar meson of_mass about $0.5~GeV$ what apparently does not coincide with_the light component $M_2$. Nevertheless, our_consideration having no_claims_of_the quantitative agreement shows_the scalar meson could be the_mixed particle of pretty large width_(similar to the superposition of oscillators with_the various fundamental frequencies) describing the_excitations of chiral and gluon_condensates.
VI. Conclusion_{#vi.-conclusion.unnumbered}
==============
In this paper we have_developed the consistent_approach to_describe the interaction_of quarks with IL. Theoretically, it_is based (and_justified) on the particular choice of_the_configurations saturating the_functional_integral_what occurs_to be not_merely_a technical_exercise._They are the deformable (crampled) (anti-)instantons_with_the variable parameters $\gamma(x)$ and in the_concrete treatment of this_paper_we play with the_variation of the PP size_$\rho(x,z)$. In a sense, such an_ansatz is_strongly motivated_by the form of quark determinant which is solely dependent on_the average instanton size in the_SCSB theory. We have_demonstrated that_in_the long-length wave_approximation_the variational_problem of the deformation field optimization turns_into the_construction of effective Lagrangian for the_scalar phononlike $\varphi$ and_quark_fields with the Yukawa interaction. Physically,_it allows us to analyze the_inverse influence of quarks on_the_instanton_vacuum. We have pointed out_this influence on the IL parameters_as negligible. The_modification of the SCSB parameters occurs pretty_poor_as well. In particular, the scale_of_quark condensate change amounts to a_few_$MeV$_only. Nevertheless, switching on the_phononlike excitations of IL leads to_several qualitatively new and interesting effects. The propagation of_the quark condensate_disturbances over IL happens in_this_approach_to be in close analogy with well-known polaron problem. We_imply a_necessity to take_into account the medium feedback while elementary excitations propagating. The_intriguing conclusion comes with realizing it leads_to generation of current quark masses $m_f$ within QCD itself with_their values |
straete, “Discriminating states: The quantum chernoff bound,” *Phys. Rev. Lett.*, vol. 98, no. 16, p. 160501, 2007.
M. Ohya and D. Petz, *Quantum Entropy and Its Use*.1em plus 0.5em minus 0.4emSpringer, 1993.
M. Mosony and N. Datta, “Generalized relative entropies and the capacity of classical-quantum channels,” 2008. \[Online\]. Available: <http://arxiv.org/abs/0810.3478>
R. Bhatia, *Matrix Analysis*, ser. Graduate Texts in Mathematics. 1em plus 0.5em minus 0.4emSpringer, 1997.
N. Datta, “Min- and max- relative entropies and a new entanglement monotone,” Mar. 2008. \[Online\]. Available: <http://arxiv.org/abs/0803.2770v1>
R. Alicki and M. Fannes, “Continuity of quantum mutual information,” Dec. 2003. \[Online\]. Available: <http://arxiv.org/abs/quant-ph/0312081v1>
F. Hansen and G. K. Pedersen, “Jensen’s operator inequality,” *Bull. London Math. Soc.*, vol. 35, no. 5, pp. 553–564, Jul. 2003.
D. Petz, “Quasi-entropies for finite quantum systems,” *Rep. Math. Phys.*, vol. 23, no. 1, pp. 57–65, Sep. 1984.
N. Datta and R. Renner, “Smooth [R]{}[é]{}nyi entropies and the quantum information spectrum,” Jan. 2008. \[Online\]. Available: <http://arxiv.org/abs/0801.0282>
[^1]: We use $\log$ to denote the binary logarithm.
[^2]: We consider this result an extension of the AEP in the sense that it takes the role of the classical AEP in a quantum information context. Namely, if an information theoretic problem can be solved in terms of min- and max-entropies in a single-shot scenario, the asymptotic result for | straete, “ Discriminating states: The quantum chernoff bound, ” * Phys. Rev. Lett. *, vol. 98, no. 16, p. 160501, 2007.
M. Ohya and D. Petz, * Quantum Entropy and Its Use*.1em plus 0.5em minus 0.4emSpringer, 1993.
M. Mosony and N. Datta, “ Generalized relative entropies and the capability of authoritative - quantum channels, ” 2008. \[Online\ ]. Available: < http://arxiv.org/abs/0810.3478 >
R. Bhatia, * Matrix Analysis *, ser. Graduate Texts in Mathematics. 1em plus 0.5em minus 0.4emSpringer, 1997.
N. Datta, “ Min- and max- relative information and a raw entanglement monotone, ” Mar. 2008. \[Online\ ]. Available: < http://arxiv.org/abs/0803.2770v1 >
R. Alicki and M. Fannes, “ Continuity of quantum mutual data, ” Dec. 2003. \[Online\ ]. Available: < http://arxiv.org/abs/quant-ph/0312081v1 >
F. Hansen and G. K. Pedersen, “ Jensen ’s operator inequality, ” * Bull. London Math. Soc. *, vol. 35, no. 5, pp. 553–564, Jul. 2003.
D. Petz, “ Quasi - entropies for finite quantum systems, ” * Rep. Math. Phys. *, vol. 23, no. 1, pp. 57–65, Sep. 1984.
N. Datta and R. Renner, “ Smooth [ R]{}[é]{}nyi information and the quantum information spectrum, ” Jan. 2008. \[Online\ ]. Available: < http://arxiv.org/abs/0801.0282 >
[ ^1 ]: We use $ \log$ to announce the binary logarithm.
[ ^2 ]: We consider this result an extension of the AEP in the sense that it lead the role of the classical AEP in a quantum information context. Namely, if an data theoretic problem can be solved in price of min- and max - entropies in a single - nip scenario, the asymptotic result for | strwete, “Discriminating statts: The quantum chgrboff bmund,” *Pgys. Rev. Uett.*, vol. 98, no. 16, p. 160501, 2007.
M. Ohya and D. Pevz, *Qyantun Entropy and Its Use*.1eo plus 0.5em minus 0.4enSprmnger, 1993.
M. Mosony anv N. Datta, “Generzpizeb celative entroples and the capacity of cnarsncal-quantum channels,” 2008. \[Online\]. Availabje: <http://srdiv.org/abs/0810.3478>
R. Bhatya, *Mserix Analysis*, ser. Graduate Texts in Mafhematibs. 1em plus 0.5em minis 0.4emSpringer, 1997.
N. Datta, “Min- ajd mwx- relative entroples and a ngs egranglement munotone,” Mag. 2008. \[Online\]. Abailable: <http://arxiv.org/abs/0803.2770v1>
R. Alickk and M. Fannes, “Cinrinkhty of quanvum munual informatlpn,” Dec. 2003. \[Onlinr\]. Available: <htbp://arxmv.ort/abs/quant-ph/0312081v1>
F. Hansen aid G. K. Pedersen, “Jenseg’s operatmr inequality,” *Bull. Oobdon Kath. Soc.*, col. 35, no. 5, p'. 553–564, Nul. 2003.
D. Pftz, “Quasi-entrkpies for funite quantum systeks,” *Grp. Math. Phys.*, vol. 23, nj. 1, [p. 57–65, Sep. 1984.
N. Datta and R. Renner, “Smooth [R]{}[é]{}nyi tntrolies and the quantum indormation spectrum,” Jaj. 2008. \[Online\]. Available: <http://arxiv.org/abs/0801.0282>
[^1]: We use $\log$ to denote tve biiafy oonarigym.
[^2]: We consider this result an extension of the WSP ik the sense that it takes tne rpje of the clarsical AEL in a quantum inflrmatiog conrext. Namejy, ig an information theoretic problem can ye wolved in terms of min- and mar-entrokies im a single-shot scenario, the asymptotic result fkf | straete, “Discriminating states: The quantum chernoff bound,” Lett.*, 98, no. p. 160501, 2007. *Quantum and Its Use*.1em 0.5em minus 0.4emSpringer, M. Mosony and N. Datta, “Generalized entropies and the capacity of classical-quantum channels,” 2008. \[Online\]. Available: <http://arxiv.org/abs/0810.3478> R. Bhatia, Analysis*, ser. Graduate Texts in Mathematics. 1em plus 0.5em minus 0.4emSpringer, 1997. N. “Min- max- entropies a new entanglement monotone,” Mar. 2008. \[Online\]. Available: <http://arxiv.org/abs/0803.2770v1> R. Alicki and M. Fannes, “Continuity of mutual information,” Dec. 2003. \[Online\]. Available: <http://arxiv.org/abs/quant-ph/0312081v1> F. and G. K. Pedersen, operator inequality,” *Bull. London Math. vol. no. 5, 553–564, 2003. Petz, “Quasi-entropies for quantum systems,” *Rep. Math. Phys.*, vol. 23, no. 1, pp. 57–65, Sep. 1984. N. Datta and R. “Smooth [R]{}[é]{}nyi the quantum spectrum,” 2008. Available: <http://arxiv.org/abs/0801.0282> [^1]: $\log$ to denote the binary logarithm. this result an extension of the AEP in sense that takes the role of the classical in a quantum information context. Namely, if an theoretic problem can be solved in terms of min- and max-entropies in a single-shot scenario, result for | straete, “Discriminating statEs: The quantUm cheRnoFf bOuNd,” *PhYs. ReV. Lett.*, vol. 98, no. 16, p. 160501, 2007.
M. OhYA and d. Petz, *Quantum Entropy and its UsE*.1eM Plus 0.5EM mInus 0.4eMSpringER, 1993.
M. mOSonY aNd n. DaTtA, “geNeralIzeD relatiVe entropieS anD tHe capacity of CLaSsical-quanTum Channels,” 2008. \[OnliNe\]. AVailabLe: <HttP://Arxiv.Org/Abs/0810.3478>
R. BHatia, *MATrix AnAlysis*, ser. grADuate TEXts in MaTHEmAticS. 1em plus 0.5em minus 0.4emSPRiNGer, 1997.
N. Datta, “Min- anD max- reLaTIvE ENtrOpiEs and a new eNtAngleMEnt monoTOnE,” mAR. 2008. \[OnLIne\]. Available: <hTtp://arxiv.org/ABs/0803.2770v1>
r. AlickI aNd M. fAnnes, “COntinUiTY of Quantum mutuAl inFormation,” dec. 2003. \[OnlINe\]. AvailABle: <http://Arxiv.oRg/aBs/qUant-PH/0312081v1>
f. HAnsEn ANd G. k. peDerSEn, “JEnsen’s opErAtOr ineQualITY,” *bUll. LOndOn MaTh. Soc.*, Vol. 35, no. 5, pp. 553–564, Jul. 2003.
D. PeTz, “QUasi-ENtrOpies For fiNite QuAntum SystemS,” *Rep. MAtH. Phys.*, vol. 23, no. 1, pp. 57–65, Sep. 1984.
n. DatTa and R. RenNer, “smOotH [R]{}[É]{}nyi eNTropieS anD thE quantuM informATioN sPECTrUm,” Jan. 2008. \[Online\]. AvailabLe: <HTTp://Arxiv.org/Abs/0801.0282>
[^1]: We uSE $\lOg$ TO denote tHe BinAry lOGArithM.
[^2]: We cONsIder this Result AN eXtEnsion oF tHe AEP iN tHe sEnsE that IT takEs the rOle of the ClassICal AEP in a quantUM information cONtEXT. NAMely, If aN informatioN theORetiC proBLeM caN Be solVed in TeRMs OF min- and max-entropies In A singlE-shot Scenario, the asYmptotic reSULT for | straete, “Discriminating s tates: The quan tum ch er noff bou nd,” *Phys. Re v . Le tt.*, vol. 98, no. 16, p. 1 60 5 01,2 00 7.
M . Ohyaa nd D . P et z, *Q ua n tu m Ent rop y and I ts Use*.1e m p lu s 0.5em minu s 0 .4emSpring er, 1993.
M. M oso ny and N . D a tta,“Ge neral ized r e lative entropie sa nd the capacit y of cla ssical-quantum ch a nn e ls,” 2008. \[O nline\ ]. Av a i lab le: <http://a rx iv.or g /abs/08 1 0. 3 4 7 8>R. Bhatia, *M atrix Analy s is* , ser. G rad u ate Te xts i nM ath ematics. 1e m pl us 0.5emminus0 .4emSpr i nger, 1 997.
N.Dat ta,“ Mi n- an dm ax- re lat i veentropie san d a n ew e n t a n glem ent mon otone ,” Mar. 2008. \[ Onli n e\] . Ava ilabl e: < ht tp:// arxiv. org/a bs /0803.2770v1>
R. A licki and M. F ann es , “Co n tinuit y o f q uantummutuali nfo rm a t i on ,” Dec. 2003. \[On li n e \] . Availa ble: < h tt p: / /arxiv.o rg /ab s/qu a n t-ph/ 0312 0 81 v1>
F.Hansen an dG. K. P ed ersen, “ Jen sen ’s op e rato r ineq uality,” *Bul l . London Math. Soc.*, vol. 3 5 ,n o .5 , pp . 5 53–564, Jul . 20 0 3.
D. P e tz , “ Q uasi- entro pi e sf or finite quantum s ys tems,” *Rep . Math. Phys. *, vol. 23 , n o. 1, pp . 57 – 65 , Sep. 1984.
N . Dat ta and R.R enner, “ Smoot h [R]{}[ é]{}nyi e n t ropies a ndthe qu ant u m i nformation sp e c trum ,” Jan. 2 008 . \[Onl ine \]. Av ail ab le: <http ://arxiv .o rg /a bs /08 01.02 8 2>
[^1] :Weus e $ \log$ to den ote t he b in ar y lo garithm .
[ ^ 2]:We c onsi der t his r esul t an extens ion of th e A E P in t he sensethat it takes t he role of t heclassi c a l AEP in a quantum informationc ontext. Na mely, ifan inform ati on the ore t ic pro blem c an be s olv e d in t e r ms of m in- and ma x - ent ropie sin a single -shot scenario, th e as ymptotic resu ltfor | straete, “Discriminating_states: The_quantum chernoff bound,” *Phys._Rev. Lett.*,_vol. 98,_no. 16, p._160501,_2007.
M. Ohya and D. Petz,_*Quantum Entropy and_Its Use*.1em plus 0.5em_minus 0.4emSpringer, 1993.
M. Mosony_and_N. Datta, “Generalized relative entropies and the capacity of classical-quantum channels,” 2008. \[Online\]. Available: <http://arxiv.org/abs/0810.3478>
R. Bhatia,_*Matrix_Analysis*, ser._Graduate_Texts_in Mathematics. 1em plus 0.5em_minus 0.4emSpringer, 1997.
N. Datta, “Min- and_max- relative_entropies and a new entanglement monotone,” Mar. 2008._\[Online\]._Available: <http://arxiv.org/abs/0803.2770v1>
R. Alicki and_M. Fannes, “Continuity of quantum mutual information,” Dec. 2003. \[Online\]._Available: <http://arxiv.org/abs/quant-ph/0312081v1>
F. Hansen and G. K. Pedersen, “Jensen’s_operator inequality,” *Bull._London_Math._Soc.*, vol. 35, no. 5, pp._553–564, Jul. 2003.
D. Petz, “Quasi-entropies for finite_quantum systems,” *Rep. Math. Phys.*, vol. 23,_no. 1, pp. 57–65, Sep. 1984.
N. Datta and R. Renner,_“Smooth [R]{}[é]{}nyi entropies and the quantum_information spectrum,” Jan. 2008. \[Online\]._Available: <http://arxiv.org/abs/0801.0282>
[^1]:_We use $\log$ to denote_the binary logarithm.
[^2]:_We consider_this result an_extension of the AEP in the_sense that it_takes the role of the classical_AEP_in a quantum_information_context._Namely, if_an information theoretic_problem_can be_solved_in terms of min- and max-entropies_in_a single-shot scenario, the asymptotic result for |
of cascade (obtaining $v^{+}$, $v^{-}$, $A^{+}$, $A^{-}$)if it is clear from the context which cascade we have on mind. If $v \in V \setminus
\max V$, then the set $v^+$ (if infinite) may be endowed with an order of the type $\omega$, and then by $(v_n)_{n \in \omega}$ we denote the sequence of elements of $v^+$, and by $v_{nV}$ - the $n$-th element of $v^{+V}$.
The [*rank*]{} of $v \in V$ ($r_V(v)$ or $r(v)$) is defined inductively as follows: $r(v)=0$ if $v \in \max V$, and otherwise $r(v)$ is the least ordinal greater than the ranks of all immediate successors of $v$. The rank $r(V)$ of the cascade $V$ is, by definition, the rank of $\emptyset_V$. If it is possible to order all sets $v^+$ (for $v \in V \setminus \max V$) so that for each $v
\in V \setminus \max V$ the sequence $(r(v_n)_{n<\omega})$ is non-decreasing, then the cascade $V$ is [*monotone*]{}, and we fix such an order on $V$ without indication.
Let $W$ be a cascade, and let $(V_w)_{w \in \max W}$ be a pairwise disjoint sequence of cascades such that $V_w \cap W = \emptyset$ for all $w \in \max W$. Then, the [*confluence*]{} of cascades $V_w$ with respect to the cascade $W$ (we write $W \looparrowleft V_w$) is defined as a cascade constructed by the identification $w \in \max
W$ with $\emptyset_{V_w}$ and according to the following rules: $\emptyset_W = \emptyset_{W \looparrowleft V_w}$; if $w\in W
\setminus \max W$, then $w^{+ W \looparrowleft V_w} = w^{+W}$; if $w
\in V_{w_0}$ (for a certain $w_0 \in \max W$), then $w^{+ W
\looparrowleft V_w} = | of cascade (obtaining $ v^{+}$, $ v^{-}$, $ A^{+}$, $ A^{-}$)if it is clear from the context which cascade we get on judgment. If $ v \in V \setminus
\max V$, then the set $ v^+$ (if countless) may be endowed with an order of the type $ \omega$, and then by $ (v_n)_{n \in \omega}$ we announce the succession of elements of $ v^+$, and by $ v_{nV}$ - the $ n$-th element of $ v^{+V}$.
The [ * rank * ] { } of $ v \in V$ ($ r_V(v)$ or $ r(v)$) is defined inductively as keep up: $ r(v)=0 $ if $ v \in \max V$, and differently $ r(v)$ is the least ordinal greater than the ranks of all immediate successors of $ v$. The rank $ r(V)$ of the cascade $ V$ is, by definition, the rank of $ \emptyset_V$. If it is potential to order all sets $ v^+$ (for $ v \in vanadium \setminus \max V$) so that for each $ v
\in V \setminus \max V$ the succession $ (r(v_n)_{n<\omega})$ is non - decreasing, then the cascade $ V$ is [ * flat * ] { }, and we fix such an club on $ V$ without indication.
Let $ W$ be a cascade, and lease $ (V_w)_{w \in \max W}$ be a pairwise disjoint sequence of cascades such that $ V_w \cap W = \emptyset$ for all $ w \in \max W$. Then, the [ * confluence * ] { } of cascades $ V_w$ with respect to the cascade $ W$ (we write $ W \looparrowleft V_w$) is defined as a cascade constructed by the identification $ w \in \max
W$ with $ \emptyset_{V_w}$ and according to the following rules: $ \emptyset_W = \emptyset_{W \looparrowleft V_w}$; if $ w\in W
\setminus \max W$, then $ w^{+ W \looparrowleft V_w } = w^{+W}$; if $ w
\in V_{w_0}$ (for a sealed $ w_0 \in \max W$), then $ w^{+ W
\looparrowleft V_w } = | of cascade (obtaining $v^{+}$, $v^{-}$, $A^{+}$, $A^{-}$)if it is cleat drom tie contsxt whicf cascade we have on mind. If $v \in V \setminus
\max V$, then thd set $v^+$ (iv infinire) mey be endowed wivg an order of bhe tvpx $\omega$, and thek by $(v_n)_{n \in \omega}$ we denode tke sequence of elements of $v^+$, and by $d_{nV}$ - thr $j$-th element of $v^{+V}$.
Tnq [*rahk*]{} of $v \in V$ ($r_V(v)$ or $r(v)$) is defined jnductinely as follows: $r(f)=0$ if $v \in \max V$, and otherwlse $g(v)$ is the least orfinal greatgd trqn the ranks of all immediate succgssors of $v$. The rank $r(V)$ of the cxscadz $V$ is, by dgyunihhon, the ranj of $\vmptyset_V$. If ln is povsible yo order all scts $v^+$ (foe $v \in V \setminus \max V$) so that for eacr $v
\in V \satjinus \max V$ the swqyence $(r(v_n)_{t<\omeeq})$ ir nkn-vecdeasinh, tien the casdade $V$ is [*minotone*]{}, and we fix xusy an order on $V$ witrott indication.
Let $W$ be a cascade, and let $(V_w)_{s \in \max W}$ be a pairwisw disjoint sequence ov cascadef such that $V_w \cap W = \emptyset$ for all $w \in \max W$. Then, ghe [*cjvdlkence*]{} of cascades $V_w$ with respect to the cascwse $W$ (we write $W \lojparrowleft V_a$) of defined as x casccse constructed by thf identyficarion $w \in \max
E$ with $\emptyset_{V_w}$ and accoreing to the yoloowing rules: $\emptydet_W = \emptyret_{W \loopsrrowleft V_w}$; if $w\in W
\secminus \max W$, then $w^{+ W \loopzfrowleft V_w} = w^{+W}$; kf $e
\it V_{w_0}$ (for a certain $w_0 \in \mav W$), then $x^{+ W
\lopparrowueft V_w} = | of cascade (obtaining $v^{+}$, $v^{-}$, $A^{+}$, $A^{-}$)if clear the context cascade we have V \max V$, then set $v^+$ (if may be endowed with an order the type $\omega$, and then by $(v_n)_{n \in \omega}$ we denote the sequence elements of $v^+$, and by $v_{nV}$ - the $n$-th element of $v^{+V}$. The of \in ($r_V(v)$ $r(v)$) is defined inductively as follows: $r(v)=0$ if $v \in \max V$, and otherwise $r(v)$ is least ordinal greater than the ranks of all successors of $v$. The $r(V)$ of the cascade $V$ by the rank $\emptyset_V$. it possible to order sets $v^+$ (for $v \in V \setminus \max V$) so that for each $v \in V \setminus V$ the is non-decreasing, the $V$ [*monotone*]{}, and we an order on $V$ without indication. a cascade, and let $(V_w)_{w \in \max W}$ a pairwise sequence of cascades such that $V_w W = \emptyset$ for all $w \in \max Then, the [*confluence*]{} of cascades $V_w$ with respect to the cascade $W$ (we write $W is defined as a constructed by the $w \max with and according the following rules: $\emptyset_W = \emptyset_{W \looparrowleft V_w}$; if $w\in W \max W$, then $w^{+ W \looparrowleft V_w} = w^{+W}$; if V_{w_0}$ a certain $w_0 \max W$), then $w^{+ \looparrowleft = | of cascade (obtaining $v^{+}$, $v^{-}$, $A^{+}$, $A^{-}$)if iT is clear frOm the ConTexT wHich CascAde we have on minD. if $v \iN V \setminus
\max V$, then the sEt $v^+$ (if InFInitE) MaY be enDowed wiTH aN ORdeR oF tHe tYpE $\OmEga$, anD thEn by $(v_n)_{n \In \omega}$ we dEnoTe The sequence oF ElEments of $v^+$, aNd bY $v_{nV}$ - the $n$-th elEmeNt of $v^{+V}$.
thE [*raNK*]{} of $v \iN V$ ($r_v(v)$ or $r(V)$) is defINed indUctively aS fOLlows: $r(V)=0$ If $v \in \maX v$, AnD othErwise $r(v)$ is the leasT OrDInal greater thaN the raNkS Of ALL imMedIate succesSoRs of $v$. tHe rank $r(v)$ Of THE CasCAde $V$ is, by definItion, the ranK Of $\eMptyseT_V$. if iT Is possIble tO oRDer All sets $v^+$ (for $V \in V \Setminus \mAx V$) so tHAt for eaCH $v
\in V \seTminus \Max v$ thE seqUEnCe $(R(v_n)_{N<\oMEga})$ IS nOn-dECreAsing, theN tHe CascaDe $V$ iS [*MONOtonE*]{}, anD we fIx sucH an order on $V$ wiThoUt inDIcaTion.
LEt $W$ be A casCaDe, and Let $(V_w)_{w \In \max w}$ bE a pairwise disjoInt sEquence of CasCaDes SuCh thaT $v_w \cap W = \EmpTysEt$ for alL $w \in \max w$. theN, tHE [*COnFluence*]{} of cascades $V_W$ wITH rEspect to The casCAdE $W$ (WE write $W \lOoParRowlEFT V_w$) is DefiNEd As a cascaDe consTRuCtEd by the IdEntifiCaTioN $w \iN \max
W$ WIth $\eMptyseT_{V_w}$ and acCordiNG to the followinG Rules: $\emptyset_w = \EmPTYsET_{W \loOpaRrowleft V_w}$; iF $w\in w
\SetmInus \MAx w$, thEN $w^{+ W \loOparrOwLEfT v_w} = w^{+W}$; if $w
\in V_{w_0}$ (for a certAiN $w_0 \in \maX W$), theN $w^{+ W
\looparrowlEft V_w} = | of cascade (obtaining $v^ {+}$, $v^{ -}$,$A^ {+} $, $A^ {-}$ )if it is clea r fro m the context which ca scade w e hav e o n min d. If $ v \ i n V\s et min us \m ax V$ , t hen the set $v^+$ (i finfinite) ma y b e endowedwit h an order o f t he typ e$\o m ega$, an d the n by $ ( v_n)_{ n \in \om eg a }$ wed enote t h e s eque nce of elements o f $ v ^+$, and by $v _{nV}$ - th e $n$ -th element o f$v^{+ V }$.
Th e [ * r a nk* ] {} of $v \inV$ ($r_V(v) $ or $r(v) $) is define d ind uc t ive ly as follo ws:$r(v)=0$if $v\ in \max V$, and other wis e $ r(v) $ i sthe l e ast or din a l g reater t ha nthe r anks o f allimm edia te su ccessors of $ v$. The ran k $r( V)$ o f th ecasca de $V$ is,by definition, th e ra nk of $\e mpt ys et_ V$ . Ifi t is p oss ibl e to or der all set s$ v ^ +$ (for $v \in V \se tm i n us \max V$ ) so t h at f o r each $ v\in V \ s e tminu s \m a xV$ the s equenc e $ (r (v_n)_{ n< \omega }) $ i s n on-de c reas ing, t hen thecasca d e $V$ is [*mon o tone*]{}, and we f ix such an order on $ V$ w i thou t in d ic ati o n.
L et $W $b ea cascade, and let $ (V _w)_{w \in\max W}$ be a pairwised i s joint se quen c eo f cascades suc h tha t $V_w \ca p W = \em ptyse t$ for a ll $w \in \ max W$.The n,the [* c o nf luence*]{} of c asca de s $V_w$ wi th resp ect to th e c as cade $W$(we writ e$W \ lo opa rrowl e ft V_w$) i s d ef ine d asa casca de co nstr uc te d by the id e nt i f icat io n$w \ in\m ax
W$ wit h $\ emptyse t_{V_w}$and acco rd in g to th e following r ul es: $\empt ys et_ W = \e m p tyset_{W \looparrowleft V_w}$;i f $w\in W\setm inus \max W$, th en $w^ {+W \loop arrowl eft V _w } = w ^{+W} $ ; i f $ w\in V_{w_0 } $ (f or ace rtai n $w_0\in \max W$), then $w^ {+ W
\looparr owl eftV _ w} = | of_cascade (obtaining_$v^{+}$, $v^{-}$, $A^{+}$, $A^{-}$)if_it is_clear_from the_context_which cascade we_have on mind._If $v \in V_\setminus
\max V$, then_the_set $v^+$ (if infinite) may be endowed with an order of the type $\omega$,_and_then by_$(v_n)_{n_\in_\omega}$ we denote the sequence_of elements of $v^+$, and_by $v_{nV}$_- the $n$-th element of $v^{+V}$.
The [*rank*]{} of_$v_\in V$ ($r_V(v)$_or $r(v)$) is defined inductively as follows: $r(v)=0$ if_$v \in \max V$, and otherwise_$r(v)$ is the_least_ordinal_greater than the ranks_of all immediate successors of $v$._The rank $r(V)$ of the cascade_$V$ is, by definition, the rank of_$\emptyset_V$. If it is possible to_order all sets $v^+$ (for_$v \in_V \setminus \max V$) so_that for each_$v
\in V_\setminus \max V$_the sequence $(r(v_n)_{n<\omega})$ is non-decreasing, then_the cascade $V$_is [*monotone*]{}, and we fix such_an_order on $V$_without_indication.
Let_$W$ be_a cascade, and_let_$(V_w)_{w \in_\max_W}$ be a pairwise disjoint sequence_of_cascades such that $V_w \cap W =_\emptyset$ for all $w_\in_\max W$. Then, the_[*confluence*]{} of cascades $V_w$ with_respect to the cascade $W$ (we_write $W_\looparrowleft V_w$)_is defined as a cascade constructed by the identification $w \in_\max
W$ with $\emptyset_{V_w}$ and according to_the following rules: $\emptyset_W_= \emptyset_{W_\looparrowleft_V_w}$; if $w\in_W
\setminus_\max W$,_then $w^{+ W \looparrowleft V_w} = w^{+W}$;_if $w
\in_V_{w_0}$ (for a certain $w_0 \in_\max W$), then $w^{+_W
\looparrowleft_V_w} = |
Colbert E. J. M., Heckman, T. M. Ptak, A. F. et al. 2004, ApJ 602, 231 Collin, S., Kawaguchi, T.: 2004, A&A 426, 797 Combes F.: 2001, in Starburst-AGN Connection, ed. I. Aretxaga et al., World Scientific Combes F.: 2003, in “Active Galactic Nuclei: from Central Engine to Host Galaxy” ASP Conf. Ser., 2003, ed. S. Collin, F. Combes and I. Shlosman, p. 411 Cowie, L. L., Songaila, A., Hu, E. M., Cohen, J. G.: 1996, AJ 112, 839 Croom S. M., Smith R. J., Boyle B. J. et al: 2004, MNRAS 349, 1397 Dekel A., Woo J: 2003, MNRAS 344, 1131 de Grijs R., O’Connell R.W., Gallagher J.S.: 2001, AJ 121, 768 Di Matteo T., Springel V., Hernquist L.: 2005, Nature 433, 604 Dijkstra, M., Haiman, Z., Loeb, A.: 2004, ApJ 613, 646 Dubus G., Rutledge R.E.: 2002, MNRAS 336, 901 Ebisuzaki, T., Makino, J., Okumura, S. K.: 1991, Nature 354, 212 Ebisuzaki, T., Makino, J., Tsuru, T.G.: et al: 2001, ApJ 562, L19 Edge A.: 2001, MNRAS 328, 762 Fabian A.C., Iwasawa K.: 1999, MNRAS 303, L34 Fabian A.C., Sanders J.S., Allen, S. W. et al.: 2003, MNRAS 344, L43 Fan X., Strauss M.A., Schneider D.P.: 2001, AJ 121, 54 Filippenko, A. V., Ho, L. C.: 2003, ApJ 588, L13 Frank J., Rees M.J.: 1976, MNRAS 176, 633 Friedli D., Benz W.: 1995, A&A 301, 649 Fu, Y. N., Huang, J. H., Deng, Z. G.: 2003, MNRAS 339, 442 Fukushige, T., Ebisuzaki, T., Makino, | Colbert E. J. M., Heckman, T. M. Ptak, A. F. et al. 2004, ApJ 602, 231 Collin, S., Kawaguchi, T.: 2004, A&A 426, 797 Combes F.: 2001, in Starburst - AGN Connection, ed. I. Aretxaga et al. , World Scientific Combes F.: 2003, in “ Active Galactic Nuclei: from Central Engine to Host Galaxy ” ASP Conf. Ser. , 2003, ed. S. Collin, F. Combes and I. Shlosman, p. 411 Cowie, L. L., Songaila, A., Hu, E. M., Cohen, J. G.: 1996, AJ 112, 839 Croom S. M., Smith R. J., Boyle B. J. et al: 2004, MNRAS 349, 1397 Dekel A., Woo J: 2003, MNRAS 344, 1131 de Grijs R., O’Connell R.W., Gallagher J.S.: 2001, AJ 121, 768 Di Matteo T., Springel V., Hernquist L.: 2005, Nature 433, 604 Dijkstra, M., Haiman, Z., Loeb, A.: 2004, ApJ 613, 646 Dubus G., Rutledge R.E.: 2002, MNRAS 336, 901 Ebisuzaki, T., Makino, J., Okumura, S. K.: 1991, Nature 354, 212 Ebisuzaki, T., Makino, J., Tsuru, T.G.: et al: 2001, ApJ 562, L19 Edge A.: 2001, MNRAS 328, 762 Fabian A.C., Iwasawa K.: 1999, MNRAS 303, L34 Fabian A.C., Sanders J.S., Allen, S. W. et al. : 2003, MNRAS 344, L43 Fan X., Strauss M.A., Schneider D.P.: 2001, AJ 121, 54 Filippenko, A. V., Ho, L. C.: 2003, ApJ 588, L13 Frank J., Rees M.J.: 1976, MNRAS 176, 633 Friedli D., Benz W.: 1995, A&A 301, 649 Fu, Y. N., Huang, J. H., Deng, Z. G.: 2003, MNRAS 339, 442 Fukushige, T., Ebisuzaki, T., Makino, | Copbert E. J. M., Heckman, T. M. Ktak, A. F. et al. 2004, AkJ 602, 231 Colnin, S., Iawaguchk, T.: 2004, A&A 426, 797 Combes F.: 2001, in Starbnrst-QGN Cinnection, ed. I. Aretxagx et al., Wlrld Sciwntihic Combes F.: 2003, in “Active Nclactjg Nucnxi: from Central Engine to Host Galaxy” AVP Clnf. Ser., 2003, ed. S. Collin, F. Combes and I. Shlosmsn, p. 411 Cowie, L. L., Fongsyla, Z., Hu, E. M., Cohen, J. G.: 1996, AJ 112, 839 Croom S. M., Smith G. J., Boyle B. J. et sl: 2004, MNRAS 349, 1397 Dekel A., Woo J: 2003, MNRWS 344, 1131 de Grijs R., O’Clnnell R.W., Gqlladyer J.S.: 2001, AJ 121, 768 Di Matteo T., Springel B., Hernquist L.: 2005, Nature 433, 604 Dijkstrx, M., Hciman, Z., Loev, Q.: 2004, WkJ 613, 646 Dubus G., Rutlvdge R.E.: 2002, MNRAS 336, 901 Ebisgzaki, T., Makino, J., Okummra, S. K.: 1991, Nature 354, 212 Ebisuzaki, T., Makino, J., Tsuru, T.G.: et al: 2001, A[J 562, L19 Edge A.: 2001, MNRAS 328, 762 Fabidn A.W., Iwxwawx K.: 1999, KNDAS 303, L34 Fauian A.C., Sansers J.S., Allwn, S. W. et al.: 2003, MNRAS 344, J43 Fan X., Strausa M.A., Sshgeider D.P.: 2001, AJ 121, 54 Filippenko, A. V., Ho, L. C.: 2003, DpJ 588, L13 Frank J., Rees M.J.: 1976, MNEAS 176, 633 Friedli D., Benz A.: 1995, A&A 301, 649 Ft, Y. N., Huang, J. H., Deng, Z. G.: 2003, MNRAS 339, 442 Fukushige, T., Ebivuzakm, G., Mcjino, | Colbert E. J. M., Heckman, T. M. F. al. 2004, 602, 231 Collin, 426, Combes F.: 2001, Starburst-AGN Connection, ed. Aretxaga et al., World Scientific Combes 2003, in “Active Galactic Nuclei: from Central Engine to Host Galaxy” ASP Conf. 2003, ed. S. Collin, F. Combes and I. Shlosman, p. 411 Cowie, L. Songaila, Hu, M., J. G.: 1996, AJ 112, 839 Croom S. M., Smith R. J., Boyle B. J. et 2004, MNRAS 349, 1397 Dekel A., Woo J: MNRAS 344, 1131 de R., O’Connell R.W., Gallagher J.S.: AJ 768 Di T., V., L.: 2005, Nature 604 Dijkstra, M., Haiman, Z., Loeb, A.: 2004, ApJ 613, 646 Dubus G., Rutledge R.E.: 2002, MNRAS 901 Ebisuzaki, J., Okumura, K.: Nature 212 Ebisuzaki, T., Tsuru, T.G.: et al: 2001, ApJ A.: 2001, MNRAS 328, 762 Fabian A.C., Iwasawa 1999, MNRAS L34 Fabian A.C., Sanders J.S., Allen, W. et al.: 2003, MNRAS 344, L43 Fan Strauss M.A., Schneider D.P.: 2001, AJ 121, 54 Filippenko, A. V., Ho, L. C.: 2003, L13 Frank J., Rees 1976, MNRAS 176, Friedli Benz 1995, 301, 649 Y. N., Huang, J. H., Deng, Z. G.: 2003, MNRAS 339, Fukushige, T., Ebisuzaki, T., Makino, | Colbert E. J. M., Heckman, T. M. Ptak, A. F. Et al. 2004, ApJ 602, 231 ColLin, S., KAwaGucHi, t.: 2004, A&A 426, 797 COmbeS F.: 2001, in Starburst-Agn ConNection, ed. I. Aretxaga et al., world scIEntiFIc combeS F.: 2003, in “ActIVe gALacTiC NUclEi: FRoM CentRal engine tO Host GalaxY” ASp COnf. Ser., 2003, ed. S. ColLIn, f. Combes and i. ShLosman, p. 411 Cowie, l. L., SOngailA, A., hu, E. m., cohen, j. G.: 1996, Aj 112, 839 CrooM S. M., SmiTH R. J., BoyLe B. J. et al: 2004, MnRas 349, 1397 Dekel a., woo J: 2003, MNRas 344, 1131 De grijS R., O’Connell R.W., GallaGHeR j.S.: 2001, AJ 121, 768 Di Matteo T., SPringeL V., hErNQUisT L.: 2005, NAture 433, 604 DijksTrA, M., HaiMAn, Z., Loeb, a.: 2004, apj 613, 646 dUBus g., rutledge R.E.: 2002, MNRaS 336, 901 Ebisuzaki, t., makIno, J., OkUmUra, s. k.: 1991, NaturE 354, 212 EbisUzAKi, T., makino, J., TsurU, T.G.: eT al: 2001, ApJ 562, L19 EdGe A.: 2001, MNRas 328, 762 Fabian a.c., IwasawA K.: 1999, MNRAs 303, L34 FAbiAn A.C., sAnDeRs J.s., ALLen, s. w. eT al.: 2003, mnRAs 344, L43 Fan X., StRaUsS M.A., ScHneiDER d.p.: 2001, AJ 121, 54 FIliPpenKo, A. V., HO, L. C.: 2003, ApJ 588, L13 Frank J., reeS M.J.: 1976, MnrAS 176, 633 friedLi D., BeNz W.: 1995, A&a 301, 649 FU, Y. N., HuAng, J. H., DEng, Z. G.: 2003, mNrAS 339, 442 Fukushige, T., EbIsuzAki, T., MakinO, | Colbert E. J. M., Heckman , T. M. Pt ak, A . F . e tal.2004 , ApJ 602, 231 Coll in, S., Kawaguchi, T.: 2004 ,A &A 4 2 6, 797CombesF .: 2 001 ,in St ar b ur st-AG N C onnecti on, ed. I. Ar et xaga et al., Wo rld Scient ifi c Combes F.: 20 03, in “ Act i ve Ga lac tic N uclei: from C entral En gi n e to H o st Gala x y ”ASPConf. Ser., 2003, ed . S. Collin, F. Combe sa nd I . S hlo sman, p. 4 11 Cowi e , L. L. , S o n g ail a , A., Hu, E.M., Cohen,J . G .: 199 6, AJ 112, 8 39 Cr oo m S. M., SmithR. J ., BoyleB. J.e t al: 2 0 04, MNR AS 349 , 1 397 Dek e lA. , W oo J:2 00 3,M NRA S 344, 1 13 1de Gr ijsR . , O’Co nne ll R .W.,Gallagher J.S .:2001 , AJ 121, 768Di M at teo T ., Spr ingel V ., Hernquist L. : 20 05, Natur e 4 33 , 6 04 Dijk s tra, M .,Hai man, Z. , Loeb, A.: 2 0 0 4 ,ApJ 613, 646 Dubus G . , R utledgeR.E.:2 00 2, MNRAS 33 6, 90 1 Eb i s uzaki , T. , M akino, J ., Oku m ur a, S. K.: 1 991, N at ure 35 4, 21 2 Ebi suzaki , T., Ma kino, J., Tsuru, T.G . : et al: 2001 , A p J 5 6 2, L 19Edge A.: 20 01,M NRAS 328 , 7 62F abian A.C. ,I wa s awa K.: 1999, MNRAS 3 03, L3 4 Fab ian A.C., San ders J.S., A l len, S.W. e t a l .: 2003, MNRAS 344, L43 Fan X . , Straus s M.A ., Schne ider D.P. : 2001, AJ 12 1,54Fil i p pe nko, A. V., H o , L.C. : 2003, Ap J 588,L13 Fr ank J. ,Rees M.J. : 1976,MN RA S17 6,633 F r iedli D. ,Ben zW.: 1995 , A&A 3 01, 6 49 F u, Y . N. , Huang , J . H.,De ng , Z. G. :2003, MNR A S 3 39, 442 Fukushig e,T ., E bi su zaki, T ., Makino, | Colbert_E. J._M., Heckman, T. M._Ptak, A._F._et al._2004,_ApJ 602, 231_Collin, S., Kawaguchi,_T.: 2004, A&A 426,_797 Combes F.:_2001,_in Starburst-AGN Connection, ed. I. Aretxaga et al., World Scientific Combes F.: 2003, in_“Active_Galactic Nuclei:_from_Central_Engine to Host Galaxy” ASP_Conf. Ser., 2003, ed. S._Collin, F._Combes and I. Shlosman, p. 411 Cowie, L._L.,_Songaila, A., Hu,_E. M., Cohen, J. G.: 1996, AJ 112, 839_Croom S. M., Smith R. J.,_Boyle B. J._et_al:_2004, MNRAS 349, 1397_Dekel A., Woo J: 2003, MNRAS_344, 1131 de Grijs R., O’Connell_R.W., Gallagher J.S.: 2001, AJ 121, 768_Di Matteo T., Springel V., Hernquist_L.: 2005, Nature 433, 604_Dijkstra, M.,_Haiman, Z., Loeb, A.: 2004,_ApJ 613, 646_Dubus G.,_Rutledge R.E.: 2002,_MNRAS 336, 901 Ebisuzaki, T., Makino,_J., Okumura, S._K.: 1991, Nature 354, 212 Ebisuzaki,_T.,_Makino, J., Tsuru,_T.G.:_et_al: 2001,_ApJ 562, L19_Edge_A.: 2001,_MNRAS_328, 762 Fabian A.C., Iwasawa K.:_1999,_MNRAS 303, L34 Fabian A.C., Sanders J.S.,_Allen, S. W. et_al.:_2003, MNRAS 344, L43_Fan X., Strauss M.A., Schneider_D.P.: 2001, AJ 121, 54 Filippenko,_A. V.,_Ho, L._C.: 2003, ApJ 588, L13 Frank J., Rees M.J.: 1976, MNRAS_176, 633 Friedli D., Benz W.:_1995, A&A 301, 649_Fu, Y._N.,_Huang, J. H.,_Deng,_Z. G.:_2003, MNRAS 339, 442 Fukushige, T., Ebisuzaki,_T., Makino, |
i+j}\right\rangle}_{N}k\underline
x}\beta_{ij\underline{{\left\langle{k+x}\right\rangle}_M}},\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+1}\right\rangle}_M}}
\beta_{jk\underline{{\left\langle{x+1}\right\rangle}_M}}&=\beta_{{\left\langle{i+j}\right\rangle}_{N}k
\underline{{\left\langle{x+1}\right\rangle}_M}}\beta_{ij\underline{{\left\langle{k+x+1}\right\rangle}_M}},\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+2}\right\rangle}_M}}
\beta_{jk\underline{{\left\langle{x+2}\right\rangle}_M}}&=
\beta_{{\left\langle{i+j}\right\rangle}_{N}k\underline{{\left\langle{x+2}\right\rangle}_M}}
\beta_{ij\underline{{\left\langle{k+x+2}\right\rangle}_M}},\nonumber\\
&\vdots\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+M-1}\right\rangle}_M}}
\beta_{jk\underline{{\left\langle{x+M-1}\right\rangle}_M}}&=
\beta_{{\left\langle{i+j}\right\rangle}_{N}k\underline{{\left\langle{x+M-1}\right\rangle}_M}}
\beta_{ij\underline{{\left\langle{k+x+M-1}\right\rangle}_M}}.
\end{aligned}$$
and obtain $$\alpha^{pM}_{ijk}f_{i{\left\langle{j+k}\right\rangle}_N}f_{jk}=f_{{\left\langle{i+j}\right\rangle}_N k}f_{ij},
\label{b-pmtrivial}$$ where $f_{ij}=\prod_{x=0 | i+j}\right\rangle}_{N}k\underline
x}\beta_{ij\underline{{\left\langle{k+x}\right\rangle}_M}},\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+1}\right\rangle}_M } }
\beta_{jk\underline{{\left\langle{x+1}\right\rangle}_M}}&=\beta_{{\left\langle{i+j}\right\rangle}_{N}k
\underline{{\left\langle{x+1}\right\rangle}_M}}\beta_{ij\underline{{\left\langle{k+x+1}\right\rangle}_M}},\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+2}\right\rangle}_M } }
\beta_{jk\underline{{\left\langle{x+2}\right\rangle}_M}}&=
\beta_{{\left\langle{i+j}\right\rangle}_{N}k\underline{{\left\langle{x+2}\right\rangle}_M } }
\beta_{ij\underline{{\left\langle{k+x+2}\right\rangle}_M}},\nonumber\\
& \vdots\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+M-1}\right\rangle}_M } }
\beta_{jk\underline{{\left\langle{x+M-1}\right\rangle}_M}}&=
\beta_{{\left\langle{i+j}\right\rangle}_{N}k\underline{{\left\langle{x+M-1}\right\rangle}_M } }
\beta_{ij\underline{{\left\langle{k+x+M-1}\right\rangle}_M } }.
\end{aligned}$$
and obtain $ $ \alpha^{pM}_{ijk}f_{i{\left\langle{j+k}\right\rangle}_N}f_{jk}=f_{{\left\langle{i+j}\right\rangle}_N k}f_{ij },
\label{b - pmtrivial}$$ where $ f_{ij}=\prod_{x=0 | i+j}\rlght\rangle}_{N}k\underline
x}\beua_{ij\underline{{\left\langle{k+e}\right\rzngle}_M}},\novumber\\
\alpha^p_{ijk}\beta_{i{\left\langpe{h+k}\rigyt\rangle}_{N}\underline{{\left\uangle{x+1}\rihht\ranglw}_M}}
\beua_{jk\underline{{\left\langle{x+1}\rlyht\rahnle}_M}}&=\bzte_{{\left\langle{i+j}\rinht\rangle}_{N}k
\gnderline{{\left\ldnelz{x+1}\right\rangle}_M}}\beta_{ij\underline{{\left\landle{k+x+1}\ribhh\rangle}_M}},\nonumbgr\\
\alpnw^p_{iji}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\undedline{{\leht\langle{x+2}\right\rsngle}_M}}
\beta_{jk\underline{{\left\lwnglf{x+2}\right\rangle}_M}}&=
\beta_{{\peft\langle{i+h}\rigrr\rangle}_{N}k\unddrline{{\left\langle{x+2}\right\tangle}_M}}
\beta_{ij\underline{{\left\langle{y+x+2}\rigkt\rangle}_M}},\nobunbeg\\
&\vdots\noiumber\\
\wlpha^p_{ijk}\beta_{i{\left\latgle{j+k}\roght\rangle}_{N}\undcrlinx{{\lefr\langle{x+M-1}\right\rangle}_M}}
\ueta_{jk\underline{{\left\lwngle{x+M-1}\richc\rangle}_M}}&=
\beta_{{\left\langlw{i+h}\righj\rangne}_{N}k\jbdeflihe{{\kert\langpe{x+J-1}\right\rangme}_M}}
\beta_{ij\uneerline{{\left\langle{k+x+K-1}\rytht\rangle}_M}}.
\ehd{aliggeq}$$
and obtain $$\alpha^{pM}_{ijk}f_{i{\left\langle{j+k}\rigvt\rzngle}_N}f_{jk}=f_{{\left\langle{i+j}\rught\rangle}_N k}f_{ij},
\labgl{b-pmtriviwl}$$ where $f_{ij}=\prod_{x=0 | i+j}\right\rangle}_{N}k\underline x}\beta_{ij\underline{{\left\langle{k+x}\right\rangle}_M}},\nonumber\\ \alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+1}\right\rangle}_M}} \beta_{jk\underline{{\left\langle{x+1}\right\rangle}_M}}&=\beta_{{\left\langle{i+j}\right\rangle}_{N}k \underline{{\left\langle{x+1}\right\rangle}_M}}\beta_{ij\underline{{\left\langle{k+x+1}\right\rangle}_M}},\nonumber\\ \alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+2}\right\rangle}_M}} \beta_{jk\underline{{\left\langle{x+2}\right\rangle}_M}}&= &\vdots\nonumber\\ \beta_{jk\underline{{\left\langle{x+M-1}\right\rangle}_M}}&= \beta_{{\left\langle{i+j}\right\rangle}_{N}k\underline{{\left\langle{x+M-1}\right\rangle}_M}} \end{aligned}$$ and obtain | i+j}\right\rangle}_{N}k\underline
x}\Beta_{ij\undeRline{{\LefT\laNgLe{k+x}\RighT\rangle}_M}},\nonumbER\\
\alpHa^p_{ijk}\beta_{i{\left\langle{j+K}\righT\rANgle}_{n}\UnDerliNe{{\left\lANgLE{X+1}\riGhT\rAngLe}_m}}
\BeTa_{jk\uNdeRline{{\leFt\langle{x+1}\rIghT\rAngle}_M}}&=\beta_{{\leFT\lAngle{i+j}\rigHt\rAngle}_{N}k
\underLinE{{\left\lAnGle{X+1}\Right\RanGle}_M}}\bEta_{ij\uNDerlinE{{\left\langLe{K+X+1}\right\RAngle}_M}},\nONUmBer\\
\aLpha^p_{ijk}\beta_{i{\left\LAnGLe{j+k}\right\ranglE}_{N}\undeRlINe{{\LEFt\lAngLe{x+2}\right\raNgLe}_M}}
\beTA_{jk\undeRLiNE{{\LEft\LAngle{x+2}\right\raNgle}_M}}&=
\beta_{{\leFT\laNgle{i+j}\RiGht\RAngle}_{N}K\undeRlINe{{\lEft\langle{x+2}\rIght\Rangle}_M}}
\beTa_{ij\unDErline{{\lEFt\langlE{k+x+2}\rigHt\rAngLe}_M}},\nONuMbEr\\
&\vDoTS\noNUmBer\\
\ALphA^p_{ijk}\betA_{i{\LeFt\lanGle{j+K}\RIGHt\raNglE}_{N}\unDerliNe{{\left\langle{x+m-1}\riGht\rANglE}_M}}
\betA_{jk\unDerlInE{{\left\Langle{X+M-1}\rigHt\Rangle}_M}}&=
\beta_{{\left\LangLe{i+j}\right\RanGlE}_{N}k\UnDerliNE{{\left\lAngLe{x+m-1}\right\rAngle}_M}}
\bETa_{iJ\uNDERlIne{{\left\langle{k+x+M-1}\riGhT\RAnGle}_M}}.
\end{aLigned}$$
ANd ObTAin $$\alpha^{PM}_{Ijk}F_{i{\leFT\LanglE{j+k}\rIGhT\rangle}_N}F_{jk}=f_{{\leFT\lAnGle{i+j}\riGhT\ranglE}_N K}f_{iJ},
\laBel{b-pMTrivIal}$$ wheRe $f_{ij}=\proD_{x=0 | i+j}\right\rangle}_{N}k\un derline
x} \beta _{i j\u nd erli ne{{ \left\langle{k + x}\r ight\rangle}_M}},\nonu mber\ \\ alph a ^p _{ijk }\beta_ { i{ \ l eft \l an gle {j + k} \righ t\r angle}_ {N}\underl ine {{ \left\langle { x+ 1}\right\r ang le}_M}}
\bet a_{ jk\und er lin e {{\le ft\ langl e{x+1} \ right\ rangle}_M }} & =\beta _ {{\left \ l an gle{ i+j}\right\rangle } _{ N }k
\underline{ {\left \l a ng l e {x+ 1}\ right\rang le }_M}} \ beta_{i j \u n d e rli n e{{\left\lang le{k+x+1}\r i ght \rangl e} _M} } ,\nonu mber\ \\ alp ha^p_{ijk}\ beta _{i{\left \langl e {j+k}\r i ght\ran gle}_{ N}\ und erli n e{ {\ lef t\ l ang l e{ x+2 } \ri ght\rang le }_ M}}
\ beta _ { j k \und erl ine{ {\lef t\langle{x+2} \ri ght\ r ang le}_M }}&=\bet a_ {{\le ft\lan gle{i +j }\right\rangle} _{N} k\underli ne{ {\ lef t\ langl e {x+2}\ rig ht\ rangle} _M}}
\b e ta_ {i j \ u nd erline{{\left\lang le { k +x +2}\righ t\rang l e} _M } },\nonum be r\\
& \vdot s\no n um ber\\
\a lpha^p _ {i jk }\beta_ {i {\left \l ang le{ j+k}\ r ight \rangl e}_{N}\u nderl i ne{{\left\lang l e{x+M-1}\righ t \r a n gl e }_M} }
\ beta_{jk\un derl i ne{{ \lef t \l ang l e{x+M -1}\r ig h t\ r angle}_M}}&=
\beta_ {{ \left\ langl e{i+j}\right\ rangle}_{N } k \ underlin e{{\ l ef t \langle{x+M-1} \righ t\rangle}_ M }}
\beta _{ij\ underlin e{{\left\ l a ngle{k+x +M- 1}\ rig ht\ r a ng le}_M}}.
\e n d {ali gn ed}$$
and obtain $$ \al pha ^{p M} _{ijk}f_{ i{\left\ la ng le {j +k} \righ t \rangle} _N }f_ {j k}= f_{{\ l eft\la ngle{ i+j} \r ig h t\r angle}_ N k } f _{ij },
\la bel {b -pmtr ivia l }$$ where$f_{ij}=\ pro d _{x= 0 | i+j}\right\rangle}_{N}k\underline
x}\beta_{ij\underline{{\left\langle{k+x}\right\rangle}_M}},\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+1}\right\rangle}_M}}
\beta_{jk\underline{{\left\langle{x+1}\right\rangle}_M}}&=\beta_{{\left\langle{i+j}\right\rangle}_{N}k
\underline{{\left\langle{x+1}\right\rangle}_M}}\beta_{ij\underline{{\left\langle{k+x+1}\right\rangle}_M}},\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+2}\right\rangle}_M}}
\beta_{jk\underline{{\left\langle{x+2}\right\rangle}_M}}&=
\beta_{{\left\langle{i+j}\right\rangle}_{N}k\underline{{\left\langle{x+2}\right\rangle}_M}}
\beta_{ij\underline{{\left\langle{k+x+2}\right\rangle}_M}},\nonumber\\
_ _&\vdots\nonumber\\
\alpha^p_{ijk}\beta_{i{\left\langle{j+k}\right\rangle}_{N}\underline{{\left\langle{x+M-1}\right\rangle}_M}}
\beta_{jk\underline{{\left\langle{x+M-1}\right\rangle}_M}}&=
\beta_{{\left\langle{i+j}\right\rangle}_{N}k\underline{{\left\langle{x+M-1}\right\rangle}_M}}
\beta_{ij\underline{{\left\langle{k+x+M-1}\right\rangle}_M}}.
\end{aligned}$$
and obtain_$$\alpha^{pM}_{ijk}f_{i{\left\langle{j+k}\right\rangle}_N}f_{jk}=f_{{\left\langle{i+j}\right\rangle}_N k}f_{ij},
__\label{b-pmtrivial}$$ where_$f_{ij}=\prod_{x=0 |
onomarev, and V. V. Zhytnikov, Gen. Relativ. Grav. [**21**]{}, 1107 (1989). [doi: [10.1007/BF00763457](http://dx.doi.org/10.1007/BF00763457)]{} P. Baekler, F. W. Hehl, and J. M. Nester, Phys. Rev. D [**83**]{}, 024001 (2011). [doi: [10.1103/PhysRevD.83.024001](http://dx.doi.org/10.1103/PhysRevD.83.024001)]{} D. Diakonov, A. G. Tumanov, and A. A. Vladimirov, Phys. Rev. D [**84**]{}, 124042 (2011). \[arXiv:1104.2432 \[hep-th\]\]. [doi: [10.1103/PhysRevD.84.124042](http://dx.doi.org/10.1103/PhysRevD.84.124042)]{} P. Baekler and F. W. Hehl, Class. Quantum Grav. [**28**]{}, 215017 (2011). \[arXiv:1105.3504\]. [doi: [10.1088/0264-9381/28/21/215017](http://dx.doi.org/10.1088/0264-9381/28/21/215017)]{} F. H. Ho and J. M. Nester, J. Phys. Conf. Ser. [**330**]{}, 012005 (2011). \[arXiv:1105.5001\]. [doi: [10.1088/1742-6596/330/1/012005](http://dx.doi.org/10.1088/1742-6596/330/1/012005)]{} F.-H. Ho and J. M. Nester, Int. J. Mod. Phys. D [**20**]{}, 2125 (2011).\
[doi: [10.1142/S0218271811020391](http://dx.doi.org/10.1142/S0218271811020391)]{} F.-H. Ho and J. M. Nester, Ann. Phys. (Berlin) [**524**]{}, 97 (2012).
G. K. | onomarev, and V. V. Zhytnikov, Gen. Relativ. Grav. [ * * 21 * * ] { }, 1107 (1989). [ doi: [ 10.1007 / BF00763457](http://dx.doi.org/10.1007 / BF00763457) ] { } P. Baekler, F. W. Hehl, and J. M. Nester, Phys. Rev. D [ * * 83 * * ] { }, 024001 (2011). [ doi: [ 10.1103 / PhysRevD.83.024001](http://dx.doi.org/10.1103 / PhysRevD.83.024001) ] { } D. Diakonov, A. G. Tumanov, and A. A. Vladimirov, Phys. Rev. D [ * * 84 * * ] { }, 124042 (2011). \[arXiv:1104.2432 \[hep - th\]\ ]. [ doi: [ 10.1103 / PhysRevD.84.124042](http://dx.doi.org/10.1103 / PhysRevD.84.124042) ] { } P. Baekler and F. W. Hehl, Class. Quantum Grav. [ * * 28 * * ] { }, 215017 (2011). \[arXiv:1105.3504\ ]. [ doi: [ 10.1088/0264 - 9381/28/21/215017](http://dx.doi.org/10.1088/0264 - 9381/28/21/215017) ] { } F. H. Ho and J. M. Nester, J. Phys. Conf. Ser. [ * * 330 * * ] { }, 012005 (2011). \[arXiv:1105.5001\ ]. [ doi: [ 10.1088/1742 - 6596/330/1/012005](http://dx.doi.org/10.1088/1742 - 6596/330/1/012005) ] { } F.-H. Ho and J. M. Nester, Int. J. Mod. Phys. D [ * * 20 * * ] { }, 2125 (2011).\
[ doi: [ 10.1142 / S0218271811020391](http://dx.doi.org/10.1142 / S0218271811020391) ] { } F.-H. Ho and J. M. Nester, Ann. Phys. (Berlin) [ * * 524 * * ] { }, 97 (2012).
G. K. | onolarev, and V. V. Zhytnikov, Nen. Relativ. Grav. [**21**]{}, 1107 (1989). [doi: [10.1007/BF00763457](ittp://dx.dki.org/10.1007/BF00763457)]{} O. Baekler, F. W. Hehl, and J. M. Nxstee, Phyw. Rev. D [**83**]{}, 024001 (2011). [doi: [10.1103/PhysRevD.83.024001](hgtp://dx.doi.ogg/10.1103/PhysRevE.83.024001)]{} D. Viakonov, A. G. Tumehov, and A. A. Vmwdimnriv, Phys. Rev. D [**84**]{}, 124042 (2011). \[arXiv:1104.2432 \[hep-dh\]\]. [doi: [10.1103/PhysRevG.84.124042](hgt'://dx.doi.org/10.1103/PhysRevD.84.124042)]{} P. Baekler and F. W. Rehl, Clssd. Quantum Grav. [**28**]{}, 215017 (2011). \[arQid:1105.3504\]. [doj: [10.1088/0264-9381/28/21/215017](http://dx.doi.org/10.1088/0264-9381/28/21/215017)]{} F. H. Ho and J. M. Nestsr, J. Phjs. Conf. Ser. [**330**]{}, 012005 (2011). \[arXif:1105.5001\]. [doi: [10.1088/1742-6596/330/1/012005](http://dx.doi.org/10.1088/1742-6596/330/1/012005)]{} F.-H. Ho ajd J. M. Nester, Int. J. Mod. Ohys. D [**20**]{}, 2125 (2011).\
[doi: [10.1142/S0218271811020391](htep://dx.doi.org/10.1142/S0218271811020391)]{} F.-F. Ho and J. M. Nester, Anh. Phys. (Berlin) [**524**]{}, 97 (2012).
G. K. | onomarev, and V. V. Zhytnikov, Gen. Relativ. 1107 [doi: [10.1007/BF00763457](http://dx.doi.org/10.1007/BF00763457)]{} Baekler, F. W. Phys. D [**83**]{}, 024001 [doi: [10.1103/PhysRevD.83.024001](http://dx.doi.org/10.1103/PhysRevD.83.024001)]{} D. A. G. Tumanov, and A. A. Phys. Rev. D [**84**]{}, 124042 (2011). \[arXiv:1104.2432 \[hep-th\]\]. [doi: [10.1103/PhysRevD.84.124042](http://dx.doi.org/10.1103/PhysRevD.84.124042)]{} P. Baekler and W. Hehl, Class. Quantum Grav. [**28**]{}, 215017 (2011). \[arXiv:1105.3504\]. [doi: [10.1088/0264-9381/28/21/215017](http://dx.doi.org/10.1088/0264-9381/28/21/215017)]{} F. H. and M. J. Conf. Ser. [**330**]{}, 012005 (2011). \[arXiv:1105.5001\]. [doi: [10.1088/1742-6596/330/1/012005](http://dx.doi.org/10.1088/1742-6596/330/1/012005)]{} F.-H. Ho and J. M. Nester, Int. J. Phys. D [**20**]{}, 2125 (2011).\ [doi: [10.1142/S0218271811020391](http://dx.doi.org/10.1142/S0218271811020391)]{} F.-H. and J. M. Nester, Phys. (Berlin) [**524**]{}, 97 (2012). K. | onomarev, and V. V. Zhytnikov, Gen. relativ. GraV. [**21**]{}, 1107 (1989). [doi: [10.1007/Bf00763457](htTp://dX.dOi.orG/10.1007/BF00763457)]{} P. baekler, F. W. Hehl, aND J. M. NEster, Phys. Rev. D [**83**]{}, 024001 (2011). [doi: [10.1103/PhysReVD.83.024001](httP://dX.Doi.oRG/10.1103/PHysReVD.83.024001)]{} D. DiakONoV, a. g. TuMaNoV, anD A. a. vlAdimiRov, phys. Rev. d [**84**]{}, 124042 (2011). \[arXiv:1104.2432 \[hep-tH\]\]. [doI: [10.1103/PHysRevD.84.124042](http://dX.DoI.org/10.1103/PhysReVD.84.124042)]{} P. baekler and F. W. hehL, Class. quAntUM Grav. [**28**]{}, 215017 (2011). \[ArXIv:1105.3504\]. [doi: [10.1088/0264-9381/28/21/215017](Http://dx.DOi.org/10.1088/0264-9381/28/21/215017)]{} F. h. Ho and J. M. NEsTEr, J. PhyS. conf. Ser. [**330**]{}, 012005 (2011). \[ARxiV:1105.5001\]. [doi: [10.1088/1742-6596/330/1/012005](Http://dx.doi.org/10.1088/1742-6596/330/1/012005)]{} F.-H. Ho aND J. m. nester, Int. J. Mod. PHys. D [**20**]{}, 2125 (2011).\
[doI: [10.1142/S0218271811020391](HTtP://DX.doI.orG/10.1142/S0218271811020391)]{} F.-H. Ho and J. m. NEster, aNn. Phys. (BERlIN) [**524**]{}, 97 (2012).
g. k. | onomarev, and V. V. Zhytni kov, Gen.Relat iv. Gr av . [* *21* *]{}, 1107 (19 8 9).[doi: [10.1007/BF00763 457]( ht t p:// d x. doi.o rg/10.1 0 07 / B F00 76 34 57) ]{ } P . Bae kle r, F. W . Hehl, an d J .M. Nester, P h ys . Rev. D [ **8 3**]{}, 0240 01(2011) .[do i : [10 .11 03/Ph ysRevD . 83.024 001](http :/ / dx.doi . org/10. 1 1 03 /Phy sRevD.83.024001)] { }D . Diakonov, A. G. Tu ma n ov , and A. A. Vladim ir ov, P h ys. Rev . D [ * *84 * *]{}, 124042(2011). \[a r Xiv :1104. 24 32\ [hep-t h\]\] .[ doi : [10.1103/ Phys RevD.84.1 24042] ( http:// d x.doi.o rg/10. 110 3/P hysR e vD .8 4.1 24 0 42) ] {} P. Bae kler and F .W. He hl,C l a s s. Q uan tumGrav. [**28**]{},215 017( 201 1). \ [arXi v:11 05 .3504 \]. [d oi: [ 10 .1088/0264-9381 /28/ 21/215017 ](h tt p:/ /d x.doi . org/10 .10 88/ 0264-93 81/28/2 1 /21 50 1 7 ) ]{ } F. H. Ho and J.M. N es ter, J.Phys.C on f. Ser. [** 33 0** ]{}, 0 12005 (20 1 1) . \[arXi v:1105 . 50 01 \]. [do i: [10.1 08 8/1 742 -6596 / 330/ 1/0120 05](http ://dx . doi.org/10.108 8 /1742-6596/33 0 /1 / 0 12 0 05)] {}F.-H. Ho an d J. M. N este r ,Int . J. M od. P hy s .D [**20**]{}, 2125 ( 20 11).\[doi: [10.1142/S02 1827181102 0 3 9 1](http: //dx . do i .org/10.1142/S 02182 7181102039 1 )]{} F.- H. Ho and J.M. Nester , Ann. Phy s.(Be rli n)[ * *5 24**]{}, 97 ( 2 0 12).
G. K. | onomarev, and_V. V._Zhytnikov, Gen. Relativ. Grav. [**21**]{}, 1107 (1989)._[doi: [10.1007/BF00763457](http://dx.doi.org/10.1007/BF00763457)]{}_P._Baekler, F._W._Hehl, and J._M. Nester, Phys. Rev. D_[**83**]{}, 024001 (2011). [doi:_[10.1103/PhysRevD.83.024001](http://dx.doi.org/10.1103/PhysRevD.83.024001)]{} D. Diakonov,_A._G. Tumanov, and A. A. Vladimirov, Phys. Rev. D [**84**]{}, 124042 (2011). \[arXiv:1104.2432 \[hep-th\]\]. [doi: [10.1103/PhysRevD.84.124042](http://dx.doi.org/10.1103/PhysRevD.84.124042)]{}_P._Baekler and_F._W._Hehl, Class. Quantum Grav. [**28**]{}, 215017 (2011)._\[arXiv:1105.3504\]. [doi: [10.1088/0264-9381/28/21/215017](http://dx.doi.org/10.1088/0264-9381/28/21/215017)]{} F. H._Ho and_J. M. Nester, J. Phys. Conf. Ser. [**330**]{}, 012005 (2011)._\[arXiv:1105.5001\]._[doi: [10.1088/1742-6596/330/1/012005](http://dx.doi.org/10.1088/1742-6596/330/1/012005)]{} F.-H._Ho and J. M. Nester, Int. J. Mod. Phys. D [**20**]{}, 2125_(2011).\
[doi: [10.1142/S0218271811020391](http://dx.doi.org/10.1142/S0218271811020391)]{} F.-H. Ho and J._M. Nester, Ann. Phys. (Berlin)_[**524**]{},_97_(2012).
G. K. |
ite] are constructed in such a way that $S_c$ is a good quantum number in the rotated-electron picture. The subspaces spanned by states whose number $2S_c$ is constant play an important role. For hole concentrations $0\leq x<1$ and maximum spin density $m=(1-x)$ there is a fully polarized vacuum, which remains invariant under the electron - rotated-electron unitary transformation, $$\vert 0_{\eta s}\rangle = \vert 0_{\eta};N_{a_{\eta}}^D\rangle\times\vert 0_{s};N_{a_{s}}^D\rangle
\times\vert GS_c;2S_c\rangle \,.
\label{vacuum}$$ Here the $\eta$-spin $SU(2)$ vacuum $\vert 0_{\eta};N_{a_{\eta}}^D\rangle$ associated with $N_{a_{\eta}}^D=[N_a^D-2S_c]$ independent $+1/2$ $\eta$-spinons, the spin $SU(2)$ vacuum $\vert 0_{s};N_{a_{s}}^D\rangle$ with $N_{a_{s}}^D=2S_c$ independent $+1/2$ spinons, and the $c$ $U(1)$ vacuum $\vert GS_c;2S_c\rangle$ with $N_c=2S_c$ $c$ fermions remain invariant under the electron - rotated-electron unitary transformation. In turn, for states with a finite number of $s1$ bond particles the $c$ fermions are not invariant under that transformation.
The whole physics can be extracted from the model (\[H\]) in the LWS-subspace referring to values of $S_{\alpha}$ and $S^z_{\alpha}$ such that $S_{\alpha}=-S^z_{\alpha}$ for $\alpha =\eta,s$ [@companion]. Within the LWS representation, the $c$ fermion creation operator can be expressed in terms of the rotated-electron operators as follows, $$f_{\vec{r}_j,c}^{\dag} =
{\tilde{c}}_{\vec{r}_j,\uparrow}^{\dag}\,
(1-{\tilde{n}}_{\vec{r}_j,\downarrow})
+ e^{i\vec{\pi}\cdot\vec{r}_j}\,{\tilde{c}}_{\vec{r}_j,\uparrow}\,
{\ | ite ] are constructed in such a way that $ S_c$ is a good quantum number in the rotate - electron movie. The subspaces spanned by states whose phone number $ 2S_c$ is constant play an significant role. For hole concentration $ 0\leq x<1 $ and maximum spin concentration $ m=(1 - x)$ there is a fully polarized vacuum, which remain invariant under the electron - rotated - electron unitary transformation, $ $ \vert 0_{\eta s}\rangle = \vert 0_{\eta};N_{a_{\eta}}^D\rangle\times\vert 0_{s};N_{a_{s}}^D\rangle
\times\vert GS_c;2S_c\rangle \, .
\label{vacuum}$$ Here the $ \eta$-spin $ SU(2)$ vacuum $ \vert 0_{\eta};N_{a_{\eta}}^D\rangle$ associated with $ N_{a_{\eta}}^D=[N_a^D-2S_c]$ autonomous $ +1/2 $ $ \eta$-spinons, the spin $ SU(2)$ vacuum $ \vert 0_{s};N_{a_{s}}^D\rangle$ with $ N_{a_{s}}^D=2S_c$ independent $ +1/2 $ spinons, and the $ c$ $ U(1)$ void $ \vert GS_c;2S_c\rangle$ with $ N_c=2S_c$ $ c$ fermions remain invariant under the electron - rotated - electron unitary transformation. In bend, for states with a finite number of $ s1 $ bond particles the $ c$ fermions are not invariant under that transformation.
The solid physics can be extracted from the model (\[H\ ]) in the LWS - subspace referring to values of $ S_{\alpha}$ and $ S^z_{\alpha}$ such that $ S_{\alpha}=-S^z_{\alpha}$ for $ \alpha = \eta, s$ [ @companion ]. Within the LWS representation, the $ c$ fermion creation operator can be expressed in terms of the rotated - electron operators as follows, $ $ f_{\vec{r}_j, c}^{\dag } =
{ \tilde{c}}_{\vec{r}_j,\uparrow}^{\dag}\,
(1-{\tilde{n}}_{\vec{r}_j,\downarrow })
+ e^{i\vec{\pi}\cdot\vec{r}_j}\,{\tilde{c}}_{\vec{r}_j,\uparrow}\,
{ \ | ite] are constructed in such a way that $S_c$ nw a gomd quahtum numcer in the rotated-electron pmctuee. Tht subspaces spanned by stated whose bumbtr $2S_c$ is constant play an imporfwnt xooe. For hole cokcentrationv $0\leq x<1$ and mafioul spin density $m=(1-x)$ there is a fully [olarizrd vacuum, which temaimf inbariant under the electron - rotates-electrmn unitary trsnsformation, $$\vert 0_{\eta s}\ranhle = \vert 0_{\eta};N_{a_{\eta}}^D\ranhle\times\verj 0_{s};N_{w_{w}}^D\rangle
\timer\vert GS_c;2S_b\xangle \,.
\labem{vacuum}$$ Here the $\eta$-spin $SU(2)$ vacjum $\vzrt 0_{\eta};N_{a_{\etq}}^D\eanhne$ associatxd witr $N_{a_{\eta}}^D=[N_a^D-2S_g]$ indepandent $+1/2$ $\eta$-spinons, thc spii $SU(2)$ vacuum $\vert 0_{s};N_{a_{s}}^D\ranjle$ with $N_{a_{s}}^D=2S_c$ indekendent $+1/2$ s[iuons, and the $c$ $U(1)$ vacuym $\vert GS_c;2V_c\ravtle$ wiuh $I_c=2S_d$ $c$ fegmikns remain invariant ynder the electron - rjnsted-electron unitawy transformation. In turn, for states with a rinite number of $s1$ bond particles the $c$ fermlons are got invariant under that transformation.
The whole [hysirs cau be ebrrwcted from the model (\[H\]) in the LWS-subspace refqdronn to values of $S_{\clpha}$ and $S^z_{\alphs}$ duvr that $S_{\alpha}=-R^z_{\alphc}$ fkr $\alpha =\eta,s$ [@compwnion]. Wythin the LWS weprrsentation, the $c$ fermion crwation operanor xan be expressed iu terms of tke rotsted-ekectron operators as foulowa, $$f_{\vec{r}_j,c}^{\dah} =
{\tilde{c}}_{\vsz{r}_j,\uparrow}^{\dag}\,
(1-{\tilae{n}}_{\nec{r}_b,\downarrow})
+ e^{i\vec{\pi}\cdot\vec{r}_t}\,{\tilde{c}}_{\ver{r}_j,\upcrrow}\,
{\ | ite] are constructed in such a way is good quantum in the rotated-electron states number $2S_c$ is play an important For hole concentrations $0\leq x<1$ and spin density $m=(1-x)$ there is a fully polarized vacuum, which remains invariant under electron - rotated-electron unitary transformation, $$\vert 0_{\eta s}\rangle = \vert 0_{\eta};N_{a_{\eta}}^D\rangle\times\vert 0_{s};N_{a_{s}}^D\rangle \times\vert \,. Here $\eta$-spin vacuum $\vert 0_{\eta};N_{a_{\eta}}^D\rangle$ associated with $N_{a_{\eta}}^D=[N_a^D-2S_c]$ independent $+1/2$ $\eta$-spinons, the spin $SU(2)$ vacuum $\vert 0_{s};N_{a_{s}}^D\rangle$ with independent $+1/2$ spinons, and the $c$ $U(1)$ vacuum GS_c;2S_c\rangle$ with $N_c=2S_c$ $c$ remain invariant under the electron rotated-electron transformation. In for with finite number of bond particles the $c$ fermions are not invariant under that transformation. The whole physics can be extracted the model the LWS-subspace to of and $S^z_{\alpha}$ such for $\alpha =\eta,s$ [@companion]. Within the $c$ fermion creation operator can be expressed in of the operators as follows, $$f_{\vec{r}_j,c}^{\dag} = {\tilde{c}}_{\vec{r}_j,\uparrow}^{\dag}\, + e^{i\vec{\pi}\cdot\vec{r}_j}\,{\tilde{c}}_{\vec{r}_j,\uparrow}\, {\ | ite] are constructed in such a wAy that $S_c$ is A good QuaNtuM nUmbeR in tHe rotated-electROn piCture. The subspaces spannEd by sTaTEs whOSe NumbeR $2S_c$ is coNStANT plAy An ImpOrTAnT role. for Hole conCentrationS $0\leQ x<1$ And maximum spIN dEnsity $m=(1-x)$ thEre Is a fully polaRizEd vacuUm, WhiCH remaIns InvarIant unDEr the eLectron - roTaTEd-elecTRon unitARY tRansFormation, $$\vert 0_{\eta s}\RAnGLe = \vert 0_{\eta};N_{a_{\eta}}^d\ranglE\tIMeS\VErt 0_{S};N_{a_{S}}^D\rangle
\tiMeS\vert gs_c;2S_c\ranGLe \,.
\LABEl{vACuum}$$ Here the $\etA$-spin $SU(2)$ vacuUM $\veRt 0_{\eta};N_{A_{\eTa}}^D\RAngle$ aSsociAtED wiTh $N_{a_{\eta}}^D=[N_a^D-2s_c]$ inDependent $+1/2$ $\Eta$-spiNOns, the sPIn $SU(2)$ vacUum $\verT 0_{s};N_{A_{s}}^D\RangLE$ wItH $N_{a_{S}}^D=2s_C$ inDEpEndENt $+1/2$ sPinons, anD tHe $C$ $U(1)$ vacUum $\vERT gs_c;2S_c\RanGle$ wIth $N_c=2s_c$ $c$ fermions reMaiN invARiaNt undEr the ElecTrOn - rotAted-elEctroN uNitary transformAtioN. In turn, foR stAtEs wItH a finITe numbEr oF $s1$ bOnd partIcles thE $C$ feRmIONS aRe not invariant undeR tHAT tRansformAtion.
THE wHoLE physics CaN be ExtrACTed frOm thE MoDel (\[H\]) in thE LWS-suBSpAcE referrInG to valUeS of $s_{\alPha}$ anD $s^z_{\alPha}$ sucH that $S_{\alPha}=-S^z_{\ALpha}$ for $\alpha =\etA,S$ [@companion]. WitHIn THE Lws repResEntation, the $C$ ferMIon cReatIOn OpeRAtor cAn be eXpREsSEd in terms of the rotatEd-ElectrOn opeRators as folloWs, $$f_{\vec{r}_j,c}^{\dAG} =
{\TIlde{c}}_{\vec{R}_j,\upARrOW}^{\dag}\,
(1-{\tilde{n}}_{\vec{r}_J,\downArrow})
+ e^{i\vec{\PI}\cdot\vec{R}_j}\,{\tilDe{c}}_{\vec{r}_j,\Uparrow}\,
{\ | ite] are constructed in su ch a way t hat $ S_c $ i sa go od q uantum numberi n th e rotated-electron pic ture. T h e su b sp acesspanned by s tat es w hos en um ber $ 2S_ c$ is c onstant pl ayan important r o le . For hole co ncentrations $0 \leq x <1 $ a n d max imu m spi n dens i ty $m= (1-x)$ th er e is af ully po l a ri zedvacuum, which rem a in s invariant und er the e l ec t r on- r otated-ele ct ron u n itary t r an s f o rma t ion, $$\vert0_{\eta s}\ r ang le = \ ve rt0 _{\eta };N_{ a_ { \et a}}^D\rangl e\ti mes\vert0_{s}; N _{a_{s} } ^D\rang le
\ti mes \ve rt G S _c ;2 S_c \r a ngl e \ ,.\ lab el{vacuu m} $$ Here the $ \ e ta$- spi n $S U(2)$ vacuum $\ver t 0 _{\e t a}; N_{a_ {\eta }}^D \r angle $ asso ciate dwith $N_{a_{\et a}}^ D=[N_a^D- 2S_ c] $ i nd epend e nt $+1 /2$ $\ eta$-sp inons,t hesp i n $S U(2)$ vacuum $\ver t0 _ {s };N_{a_{ s}}^D\ r an gl e $ with $ N_ {a_ {s}} ^ D =2S_c $ in d ep endent $ +1/2$s pi no ns, and t he $c$ $ U(1 )$vacuu m $\v ert GS _c;2S_c\ rangl e $ with $N_c=2S _ c$ $c$ fermio n sr e ma i n in var iant underthee lect ron- r ota t ed-el ectro nu ni t ary transformation. I n turn , for states witha finite n u m b er of $s 1$ b o nd particles the$c$ f ermions ar e not inv arian t underthat tran s f ormation .
The wh ole p hy sics can be e x t ract ed from t hemodel ( \[H \]) in th eLWS-subsp ace refe rr in gto va lueso f $S_{\a lp ha} $and $S^z _ {\alph a}$ s uchth at $S_ {\alpha } =- S ^ z_{\ al ph a}$for $ \alph a =\ e ta, s$ [@co mpanion]. Wi t hinth eLWS rep resentation,th e $c$ ferm io n c reatio n operator can be expressed in te r ms of t herotat ed-e lectron o per atorsasf ollows , $$f_ {\vec {r }_j , c }^{\d a g }=
{ \t ilde{c}}_{ \ v ec{ r}_j, \u parr ow}^{\d ag}\,
(1-{\tilde{n } }_{ \vec{r}_j,\do wna rrow } )
+ e^ { i\ v ec{ \p i }\c d o t\vec{r}_j}\,{\ tilde{c}}_ {\ v ec {r}_j,\upa r row }\ ,
{\ | ite] are_constructed in_such a way that_$S_c$ is_a_good quantum_number_in the rotated-electron_picture. The subspaces_spanned by states whose_number $2S_c$ is_constant_play an important role. For hole concentrations $0\leq x<1$ and maximum spin density $m=(1-x)$_there_is a_fully_polarized_vacuum, which remains invariant under_the electron - rotated-electron unitary_transformation, $$\vert_0_{\eta s}\rangle = \vert 0_{\eta};N_{a_{\eta}}^D\rangle\times\vert 0_{s};N_{a_{s}}^D\rangle
\times\vert GS_c;2S_c\rangle \,.
\label{vacuum}$$_Here_the $\eta$-spin $SU(2)$_vacuum $\vert 0_{\eta};N_{a_{\eta}}^D\rangle$ associated with $N_{a_{\eta}}^D=[N_a^D-2S_c]$ independent $+1/2$ $\eta$-spinons,_the spin $SU(2)$ vacuum $\vert 0_{s};N_{a_{s}}^D\rangle$_with $N_{a_{s}}^D=2S_c$ independent_$+1/2$_spinons,_and the $c$ $U(1)$_vacuum $\vert GS_c;2S_c\rangle$ with $N_c=2S_c$ $c$_fermions remain invariant under the electron_- rotated-electron unitary transformation. In turn, for_states with a finite number of_$s1$ bond particles the $c$_fermions are_not invariant under that transformation.
The_whole physics can_be extracted_from the model_(\[H\]) in the LWS-subspace referring to_values of $S_{\alpha}$_and $S^z_{\alpha}$ such that $S_{\alpha}=-S^z_{\alpha}$ for_$\alpha_=\eta,s$ [@companion]. Within_the_LWS_representation, the_$c$ fermion creation_operator_can be_expressed_in terms of the rotated-electron operators_as_follows, $$f_{\vec{r}_j,c}^{\dag} =
{\tilde{c}}_{\vec{r}_j,\uparrow}^{\dag}\,
(1-{\tilde{n}}_{\vec{r}_j,\downarrow})
+ e^{i\vec{\pi}\cdot\vec{r}_j}\,{\tilde{c}}_{\vec{r}_j,\uparrow}\,
{\ |
practical implementation.
At the moment, the largest LPN instances with errors in ’s range $\tau \in [0.1, 0.15]$ are solved with variants of the low-memory algorithms <span style="font-variant:small-caps;">Pooled Gauss</span> and <span style="font-variant:small-caps;">Well-Pooled Gauss</span> of Esser, Kübler, May [@C:EssKubMay17]. We show that <span style="font-variant:small-caps;">Pooled Gauss</span> solves LSN for $\tau \leq 0.292$ faster than classical period finding algorithms. <span style="font-variant:small-caps;">Well-Pooled Gauss</span> even improves on any classical period finding algorithm for all errors $\tau < \frac 1 2$.
<span style="font-variant:small-caps;">Well-Pooled Gauss</span> is able to handle errors in time $2^{cn}$, where $c<\frac 1 2$ is constant for constant $\tau$. For $\tau=0$, we obtain polynomial time as predicted by Simon’s analysis. However, for $0 <\tau < \frac 1 2$ we achieve exponential run time, but still improve over the purely classical computation. This indicates that we achieve [*quantum supremacy*]{} for the period finding problem on sufficiently large computers, even in the presence of errors: Our quantum oracle helps us in speeding up computation! But as opposed to the exponential speedup from the (overly optimistic) error-free Simon setting $\tau =0$, we obtain in the (realistic) general error-prone Simon setting $0 <\tau < \frac 1 2$ only a [*polynomial speedup*]{} with a polynomial of degree $\frac{1}{2c} >1$.
Concerning quantum supremacy, assume that one could build a quantum device with $486$ qubits performing Simon’s circuit on a $243$-bit periodic function with error $\tau = \frac 1 8$. Then the error handling would translate into an -instance with $(n,\tau)=(243, \frac 1 8)$. Such an LPN instance was solved in [@C:EssKubMay17] on 64 threads in only $15$ days, whereas classically we would need $2^{121}$ steps for period finding.\
The paper is organized as follows. In \[sec:simon\] | practical implementation.
At the moment, the largest licensed practical nurse case with errors in ’s range $ \tau \in [ 0.1, 0.15]$ are clear with form of the low - memory algorithms < couple style="font - variant: small - caps;">Pooled Gauss</span > and < span style="font - variant: small - caps;">Well - pool Gauss</span > of Esser, Kübler, May [ @C: EssKubMay17 ]. We show that < span style="font - variant: small - caps;">Pooled Gauss</span > clear LSN for $ \tau \leq 0.292 $ faster than classical time period finding algorithms. < span style="font - variant: small - caps;">Well - Pooled Gauss</span > even better on any classical period finding algorithm for all error $ \tau < \frac 1 2$.
< span style="font - variant: small - caps;">Well - Pooled Gauss</span > is able to handle mistake in time $ 2^{cn}$, where $ c<\frac 1 2 $ is constant for constant $ \tau$. For $ \tau=0 $, we obtain polynomial time as predict by Simon ’s analysis. However, for $ 0 < \tau < \frac 1 2 $ we achieve exponential run time, but still improve over the purely classical computation. This indicates that we achieve [ * quantum supremacy * ] { } for the period finding problem on sufficiently large computers, even in the presence of errors: Our quantum oracle help us in speeding up computation! But as oppose to the exponential acceleration from the (overly optimistic) error - free Simon setting $ \tau = 0 $, we obtain in the (realistic) general mistake - prone Simon setting $ 0 < \tau < \frac 1 2 $ only a [ * polynomial speedup * ] { } with a polynomial of degree $ \frac{1}{2c } > 1$.
Concerning quantum supremacy, simulate that one could build a quantum device with $ 486 $ qubits performing Simon ’s circuit on a $ 243$-bit periodic function with mistake $ \tau = \frac 1 8$. Then the error treatment would understand into an -instance with $ (n,\tau)=(243, \frac 1 8)$. Such an LPN case was solved in [ @C: EssKubMay17 ] on 64 threads in only $ 15 $ day, whereas classically we would need $ 2^{121}$ steps for period finding.\
The newspaper is organized as follows. In \[sec: simon\ ] | prwctical implementation.
At the moment, the largesv LPN ihstances with errors in ’s range $\tau \mn [0.1, 0.15]$ are wolved with variants ow the low-lemory aogormthms <span style="hknt-variant:smamp-capv;">'ooled Gauss</spak> and <span vtyle="font-variatt:rmcll-caps;">Well-Pooled Gauss</span> of Esser, Kübler, Kaj [@C:EssKubMay17]. We shoe thaf <span style="font-variant:small-caps;">Pokled Galss</span> solves LSM for $\tau \leq 0.292$ faster than cladsical period findlng algorityms. <fpan style="fong-variant:small-caps;">Well-Pkoled Gauss</span> even improves ov any classical pwrilg finding aogorinhm for all evgors $\tag < \frac 1 2$.
<span style="fokt-varmant:wmall-caps;">Well-Pooled Geuss</span> is able to randle ersoxs in time $2^{cn}$, where $c<\drqc 1 2$ hs cmnstxbt wor cpnatant $\hau$. For $\tau=0$, ws obtain pooynomial time as prtdisnrd by Simon’s analyfif. However, for $0 <\tau < \frac 1 2$ we achieve eqponsntial run time, but stiol improve over the porely clasfical computation. This indicates that we achieve [*xuantnm sukrcixxy*]{} for the period finding problem on sufficientjg kagge computers, evek in the presence pf ettors: Our quantom oracle gelps us in speedijg up cjmputqtion! But as ppposed to the exponential wpeedup from rhe (overly optimiscic) error-frez Simom setying $\tau =0$, we obtain in che (rezlistic) genfral errod-orone Simon settkng $0 <\dau < \frac 1 2$ only a [*polynomyal speednp*]{} wich a polhnomoal of degree $\frwc{1}{2c} >1$.
Concerning quantum skpremccy, avsume that one could build a quantum devirx with $486$ qubitx [erxorming Fimon’x circuit on w $243$-bit periodic functijn wigh error $\tzu = \frar 1 8$. Then the error handlith would tranvlate ineo ab -inwtance dkth $(n,\tau)=(243, \frac 1 8)$. Such an LPN instabce was solved in [@G:EssKobJay17] on 64 threads in only $15$ days, whetear cjadsmcallr we would nead $2^{121}$ rteox for period finbivg.\
Thr paper is organized as rollows. In \[sec:simom\] | practical implementation. At the moment, the largest with in ’s $\tau \in [0.1, of low-memory algorithms <span Gauss</span> and <span Gauss</span> of Esser, Kübler, May [@C:EssKubMay17]. show that <span style="font-variant:small-caps;">Pooled Gauss</span> solves LSN for $\tau \leq 0.292$ faster than period finding algorithms. <span style="font-variant:small-caps;">Well-Pooled Gauss</span> even improves on any classical period finding for errors < 1 2$. <span style="font-variant:small-caps;">Well-Pooled Gauss</span> is able to handle errors in time $2^{cn}$, where $c<\frac 1 is constant for constant $\tau$. For $\tau=0$, we polynomial time as predicted Simon’s analysis. However, for $0 < 1 2$ achieve run but still improve the purely classical computation. This indicates that we achieve [*quantum supremacy*]{} for the period finding problem on large computers, the presence errors: quantum helps us in computation! But as opposed to the the (overly optimistic) error-free Simon setting $\tau =0$, obtain in (realistic) general error-prone Simon setting $0 < \frac 1 2$ only a [*polynomial speedup*]{} a polynomial of degree $\frac{1}{2c} >1$. Concerning quantum supremacy, assume that one could build a with $486$ qubits performing circuit on a periodic with $\tau \frac 1 Then the error handling would translate into an -instance with $(n,\tau)=(243, 1 8)$. Such an LPN instance was solved in [@C:EssKubMay17] threads only $15$ days, classically we would need steps period finding.\ The paper as In | practical implementation.
At The moment, tHe larGesT LPn iNstaNces With errors in ’s rANge $\tAu \in [0.1, 0.15]$ are solved with variaNts of ThE Low-mEMoRy algOrithms <SPaN STylE="fOnT-vaRiANt:Small-CapS;">Pooled gauss</span> aNd <sPaN style="font-vaRIaNt:small-capS;">WeLl-Pooled GausS</spAn> of EsSeR, KüBLer, MaY [@C:ESsKubmay17]. We sHOw that <Span style="FoNT-variaNT:small-cAPS;">POoleD Gauss</span> solves Lsn fOR $\tau \leq 0.292$ faster tHan claSsICaL PEriOd fInding algoRiThms. <sPAn style="FOnT-VARiaNT:small-caps;">WelL-Pooled GausS</SpaN> even iMpRovES on any ClassIcAL peRiod finding AlgoRithm for aLl erroRS $\tau < \fraC 1 2$.
<Span styLe="font-VarIanT:smaLL-cApS;">WeLl-pOolED GAusS</SpaN> is able tO hAnDle erRors IN TIMe $2^{cn}$, WheRe $c<\fRac 1 2$ is Constant for coNstAnt $\tAU$. FoR $\tau=0$, wE obtaIn poLyNomiaL time aS predIcTed by Simon’s analYsis. however, foR $0 <\taU < \fRac 1 2$ We AchieVE exponEntIal Run time, But stilL ImpRoVE OVeR the purely classicaL cOMPuTation. ThIs indiCAtEs THat we achIeVe [*qUantUM SupreMacy*]{} FOr The perioD findiNG pRoBlem on sUfFicienTlY laRge CompuTErs, eVen in tHe presenCe of eRRors: Our quantum ORacle helps us iN SpEEDiNG up cOmpUtation! But aS oppOSed tO the EXpOneNTial sPeeduP fROm THe (overly optimistic) eRrOr-free simon Setting $\tau =0$, we oBtain in the (REAListic) geNeraL ErROr-prone Simon seTting $0 <\Tau < \frac 1 2$ onlY A [*polynomIal spEedup*]{} witH a polynomIAL of degreE $\frAc{1}{2c} >1$.
conCerNINg Quantum supremACY, assUmE that onE coUld builD a qUanTum DevIcE with $486$ qubiTs perforMiNg siMoN’s cIrcuiT On a $243$-bit peRiOdiC fUncTion wITh erroR $\tau = \fRac 1 8$. THeN tHE erRor handLInG WOuld TrAnSlatE inTo An -insTancE WitH $(n,\tau)=(243, \frAc 1 8)$. Such an LpN iNStanCe WaS solved In [@C:EssKubMay17] oN 64 tHreads in onLy $15$ DayS, whereAS ClassicaLly we would need $2^{121}$ steps for pERiod finDinG.\
The pAper Is organizEd aS folloWs. IN \[Sec:simOn\] | practical implementation.
At the m oment , t hela rges t LP N instances wi t h er rors in ’s range $\tau \in[0 . 1, 0 . 15 ]$ ar e solve d w i t h v ar ia nts o f t he lo w-m emory a lgorithms<sp an style="font - va riant:smal l-c aps;">Pooled Ga uss</s pa n>a nd <s pan styl e="fon t -varia nt:small- ca p s;">We l l-Poole d Ga uss< /span> of Esser,K üb l er, May [@C:Es sKubMa y1 7 ]. W e s how that <spa nstyle = "font-v a ri a n t :sm a ll-caps;">Poo led Gauss</ s pan > solv es LS N for $ \tau\l e q 0 .292$ faste r th an classi cal pe r iod fin d ing alg orithm s.<sp an s t yl e= "fo nt - var i an t:s m all -caps;"> We ll -Pool ed G a u s s </sp an> eve n imp roves on anycla ssic a l p eriod find ingal gorit hm for aller rors $\tau < \f rac1 2$.
<s pan s tyl e= "font - varian t:s mal l-caps; ">Well- P ool ed G a us s</span> is able t oh a nd le error s in t i me $ 2 ^{cn}$,wh ere $c< \ f rac 1 2$i sconstant for c o ns ta nt $\ta u$ . For$\ tau =0$ , weo btai n poly nomial t ime a s predicted byS imon’s analys i s. H ow e ver, fo r $0 <\tau< \f r ac 1 2$w each i eve e xpone nt i al run time, but still i mprove over the purely c lassical c o m p utation. Thi s i n dicates that w e ach ieve [*qua n tum supr emacy *]{} for the peri o d finding pr obl emons u ff iciently larg e comp ut ers, ev enin thepre sen ceofer rors: Our quantum o ra cl ehel ps us in speed in g u pcom putat i on! Bu t asoppo se dt o t he expo n en t i al s pe ed up f rom t he (o verl y op timisti c) error- fre e Sim on s etting$\tau =0$, we o btain in t he (r ealist i c ) genera l error-prone Simon set t ing $0<\t au <\fra c 1 2$ on lya [*po lyn o mial s peedup *]{}wi tha polyn o m ia l o fdegree $\f r a c{1 }{2c} > 1$.
Concer ning quantum supre m acy , assume that on e co u l dbui l da qu an t umd e vice with $486$ qubits pe rf o rm ing Simon’ s ci rc uit ona $243$ -bitp eriodic function with err or $\t a u =\frac 1 8$ . Then t he errorh andli n gwould tr anslat eint o an-insta n cewith$(n,\t au )=(243 , \fr ac 1 8)$.Such an LPN instance wa s solv ed in [@ C:EssKubM ay1 7 ] o n 64 thre adsin only $1 5$day s, wh ere a s cla ssic a ll y w e woul d ne e d $2^{121 } $ste p s f or period f i n d ing .\
Th e p a per is org anized as follows . In \[sec:simo n\] | practical_implementation.
At the_moment, the largest LPN_instances with_errors_in ’s_range_$\tau \in [0.1,_0.15]$ are solved_with variants of the_low-memory algorithms <span_style="font-variant:small-caps;">Pooled_Gauss</span> and <span style="font-variant:small-caps;">Well-Pooled Gauss</span> of Esser, Kübler, May [@C:EssKubMay17]. We show that <span style="font-variant:small-caps;">Pooled_Gauss</span>_solves LSN_for_$\tau_\leq 0.292$ faster than classical_period finding algorithms. <span style="font-variant:small-caps;">Well-Pooled_Gauss</span> even_improves on any classical period finding algorithm for_all_errors $\tau <_\frac 1 2$.
<span style="font-variant:small-caps;">Well-Pooled Gauss</span> is able to handle_errors in time $2^{cn}$, where $c<\frac_1 2$ is_constant_for_constant $\tau$. For $\tau=0$,_we obtain polynomial time as predicted_by Simon’s analysis. However, for $0_<\tau < \frac 1 2$ we achieve_exponential run time, but still improve_over the purely classical computation._This indicates_that we achieve [*quantum supremacy*]{}_for the period_finding problem_on sufficiently large_computers, even in the presence of_errors: Our quantum_oracle helps us in speeding up_computation!_But as opposed_to_the_exponential speedup_from the (overly_optimistic)_error-free Simon_setting_$\tau =0$, we obtain in the_(realistic)_general error-prone Simon setting $0 <\tau <_\frac 1 2$ only_a_[*polynomial speedup*]{} with a_polynomial of degree $\frac{1}{2c} >1$.
Concerning_quantum supremacy, assume that one could_build a_quantum device_with $486$ qubits performing Simon’s circuit on a $243$-bit periodic function_with error $\tau = \frac 1_8$. Then the error_handling would_translate_into an -instance_with_$(n,\tau)=(243, \frac_1 8)$. Such an LPN instance was_solved in_[@C:EssKubMay17] on 64 threads in only_$15$ days, whereas classically_we_would need $2^{121}$ steps for period_finding.\
The paper is organized as follows._In \[sec:simon\] |
isation, and overall $U(1)'$ neutrality implies they can even be much larger than the galactic escape velocity. It has been pointed out [@foot14] that this effect can give rise to observable keV electron recoils in direct detection experiments. It might even be possible to explain the DAMA annual modulation signal [@dama1a; @dama1b; @dama2a; @dama2b; @dama2c] in this manner, since the constraints on electron recoils provided by other experiments are generally much weaker than those of nuclear recoils. However, a detailed description of the plasma dark matter density and velocity distribution in the vicinity of the Earth is required. This is a highly non-trivial problem. If the dark plasma has interactions with ordinary matter then it will by captured by the Earth, forming an approximate “dark sphere” within. Understanding the interaction of the dark plasma with this dark sphere as the Earth moves through the halo is therefore of primary importance. The aim of this paper is to provide a consistent description of this interaction in order to qualitatively understand the implications for direct detection.
The captured dark sphere within the Earth forms an obstacle to the dark plasma wind. Two limiting cases can be envisaged: (1) if the captured dark matter is largely neutral, i.e. poorly conducting, then the dark plasma wind will be absorbed by the dark sphere, or; (2) if the captured dark matter is largely ionised, then the obstacle forms a conducting sphere which effectively deflects the dark plasma wind. Interestingly, these limiting cases appear analogous to the solar wind interaction with the Moon and Venus, respectively (e.g. [@Spreiter70; @Luhmann04]); this is sketched in Figure \[FigMoonVenus\]. The Moon has no magnetic field and no atmosphere, so that the solar wind is largely absorbed at the lunar surface with very little upwind activity. Venus has no magnetic field, but forms an electrically conductive layer at the edge of its ionosphere, which at first approximation forms an impenetrable obstacle to the solar wind. These systems have been studied using magnetohydrodynamic (MHD) models, and this would seem to be an appropriate starting point for studying the dark plasma wind interaction with the captured dark matter within the Earth.
![[The solar wind interaction with Moon/Venus. These systems appear to represent useful analogues to the possible ways in which the dark plasma wind interacts with captured dark | isation, and overall $ U(1)'$ neutrality implies they can even be much larger than the astronomic evasion velocity. It has been pointed out [ @foot14 ] that this consequence can hold rise to observable keV electron recoil in lineal detection experiments. It might even be possible to explain the DAMA annual modulation signal [ @dama1a; @dama1b; @dama2a; @dama2b; @dama2c ] in this manner, since the constraint on electron recoils provided by early experiments are generally a lot weaker than those of nuclear recoil. However, a detailed description of the plasma dark topic density and velocity distribution in the vicinity of the Earth is required. This is a highly non - superficial problem. If the dark plasma hold interaction with ordinary topic then it will by captured by the Earth, forming an approximate “ dark sphere ” within. understand the interaction of the dark plasma with this dark sphere as the Earth moves through the halo is consequently of chief importance. The aim of this paper is to provide a consistent description of this interaction in order to qualitatively understand the implications for direct detection.
The capture dark sphere within the Earth form an obstacle to the dark plasma wind. Two limiting cases can be imagine: (1) if the captured blue matter is largely neutral, i.e. ill conducting, then the dark plasma wind will be absorbed by the dark sphere, or; (2) if the captured dark matter is largely ionised, then the obstacle forms a conducting sector which effectively parry the dark plasma wind. Interestingly, these limiting cases appear analogous to the solar fart interaction with the Moon and Venus, respectively (e.g. [ @Spreiter70; @Luhmann04 ]); this is sketched in Figure \[FigMoonVenus\ ]. The Moon take no charismatic field and no atmosphere, so that the solar wind is largely absorbed at the lunar surface with very fiddling upwind activity. Venus has no charismatic field, but forms an electrically conductive layer at the edge of its ionosphere, which at inaugural approximation form an impenetrable obstacle to the solar wind. These system have been studied using magnetohydrodynamic (MHD) models, and this would seem to be an appropriate starting point for studying the dark plasma wind interaction with the captured dark topic within the Earth.
! [ [ The solar wind interaction with Moon / Venus. These systems look to represent useful analogues to the potential ways in which the dark plasma wind interacts with captured dark | isahion, and overall $U(1)'$ neutrxlity implies tkwy can even ge much uarger than the galactic escepe celocuty. It has been pointea out [@foon14] that thus ehfect can give rmae to onfervznle kzV electron recolls in direwt detection efpdrnments. It might even be possible to qxplain tje DAMA annual modllwtioh signal [@dama1a; @dama1b; @dama2a; @dama2b; @dzma2c] in this manner, xince the constraints on epectgon recoils providfd by other expqeiments are eenerally much weaker jhan those of nuclear recoils. Hodever, a detailee eesftiption of tie plafma dark matbvr denshty and velocity distvibutmon un the vicinity of thx Earth is required. Jhis is a viyhly non-trivial problwm. If tve ddrk ooasoa gax jnteraftikns with oddinary matrer then it will by cwiyured by the Earth, fjrming an approximate “dark sphere” withit. Uhderstanding the interaxtion of the dark pladma with ehis dark sphere as the Earth moves through the hdlo ia thtrcnore if primary importance. The aim of this paper is ek kronide a consistent descriptiom lf jhis interactiun in ordsr to qualitativelj underftand the implycatoons for direct detection.
Thw captured dcrk sphere within the Earth formr an obstscle to the dark plasma wins. Two limitlng cases zan be envisaged: (1) in tve captured dark matter is largely ieutrcl, i.e. pourly condusting, then the dark plasma wind wipl be atsorbed by the dark sphere, or; (2) if the capvnred dark matjer is largely ionixed, then the jbstacle forms a condbcting sphere whpch effecvively deflests the dark [pasma wind. Iiterestindly, rhesw limitkvg cases appeat analogols to rhe solar wind intcractkkn with the Moou qnd Venus, respevtixelr (v.g. [@Xpwaiter70; @Luhmant04]); thks kx skegched in Finurd \[FigKoonVenus\]. The Moon hds nk magnetic field amd no atmowphere, sj that the sokar wind is largelj absmrbxd at yhe lunar surface with very littls upwind wctlvity. Venus hws ni magnetic fneld, but forms an electrically conductivx layer at the edge of uts ionosphere, whick st first ap'roximwtion forks an impenetrable ovstacle to the sokar wind. These systema have been studied using magnetohydrodynamic (MHD) models, and this would seem to be ab appropriate starfing poinn fjx studyyng vhx dark plasma winc interaction with the captured dark madtzr within the Earth.
![[The solar eivd interactiov with Moon/Venus. These sgstems alpear to represent useful analobues to the possible wayx in whixh ths dark plarma wind interacts with eapturwd dark | isation, and overall $U(1)'$ neutrality implies they be larger than galactic escape velocity. [@foot14] this effect can rise to observable electron recoils in direct detection experiments. might even be possible to explain the DAMA annual modulation signal [@dama1a; @dama1b; @dama2b; @dama2c] in this manner, since the constraints on electron recoils provided by experiments generally weaker those of nuclear recoils. However, a detailed description of the plasma dark matter density and velocity in the vicinity of the Earth is required. is a highly non-trivial If the dark plasma has with matter then will captured the Earth, forming approximate “dark sphere” within. Understanding the interaction of the dark plasma with this dark sphere as the moves through is therefore primary The of this paper provide a consistent description of this to qualitatively understand the implications for direct detection. captured dark within the Earth forms an obstacle the dark plasma wind. Two limiting cases can envisaged: (1) if the captured dark matter is largely neutral, i.e. poorly conducting, then the wind will be absorbed the dark sphere, (2) the dark is largely then the obstacle forms a conducting sphere which effectively deflects the plasma wind. Interestingly, these limiting cases appear analogous to the interaction the Moon and respectively (e.g. [@Spreiter70; @Luhmann04]); is in Figure \[FigMoonVenus\]. The no field so the wind is largely absorbed the lunar surface with very upwind activity. Venus has an electrically conductive layer at the edge of ionosphere, which at first approximation forms an obstacle to the solar wind. These systems have been studied using magnetohydrodynamic models, and seem to be an appropriate starting point for the dark plasma wind with the captured dark matter within the Earth. ![[The wind with Moon/Venus. systems appear to useful analogues to possible ways in dark plasma interacts captured | isation, and overall $U(1)'$ neutralIty implies They cAn eVen Be Much LargEr than the galacTIc esCape velocity. It has been pOinteD oUT [@fooT14] ThAt thiS effect CAn GIVe rIsE tO obSeRVaBle kev elEctron rEcoils in diRecT dEtection expeRImEnts. It mighT evEn be possible To eXplain ThE DAma annuAl mOdulaTion siGNal [@damA1a; @dama1b; @daMa2A; @Dama2b; @dAMa2c] in thIS MaNner, Since the constrainTS oN Electron recoilS proviDeD By OTHer ExpEriments arE gEneraLLy much wEAkER THan THose of nuclear Recoils. HoweVEr, a DetailEd DesCRiptioN of thE pLAsmA dark matter DensIty and velOcity dIStributIOn in the ViciniTy oF thE EarTH iS rEquIrED. ThIS iS a hIGhlY non-trivIaL pRobleM. If tHE DARk plAsmA has InterActions with orDinAry mATteR then It wilL by cApTured By the EArth, fOrMing an approximaTe “daRk sphere” wIthIn. undErStandINg the iNteRacTion of tHe dark pLAsmA wITH ThIs dark sphere as the EArTH MoVes throuGh the hALo Is THerefore Of PriMary IMPortaNce. THE aIm of this Paper iS To PrOvide a cOnSistenT dEscRipTion oF This InteraCtion in oRder tO Qualitatively uNDerstand the imPLiCATiONs foR diRect detectiOn.
ThE CaptUred DArK spHEre wiThin tHe eArTH forms an obstacle to tHe Dark plAsma wInd. Two limitinG cases can bE ENVisaged: (1) iF the CApTUred dark matter Is larGely neutraL, I.e. poorly ConduCting, theN the dark pLASma wind wIll Be aBsoRbeD BY tHe dark sphere, oR; (2) IF the CaPtured dArk Matter iS laRgeLy iOniSeD, then the oBstacle fOrMs A cOnDucTing sPHere whicH eFfeCtIveLy defLEcts thE dark PlasMa WiND. InTerestiNGlY, THese LiMiTing CasEs AppeaR anaLOgoUs to the Solar wind IntERactIoN wIth the MOon and Venus, reSpEctively (e.g. [@spReiTer70; @LuhMANn04]); this is Sketched in Figure \[FigMoonvEnus\]. The mooN has nO magNetic fielD anD no atmOspHEre, so tHat the Solar WiNd iS LArgelY ABsOrbEd At the lunar SURfaCe witH vEry lIttle upWind activity. Venus hAS no Magnetic field, But FormS AN eLecTRiCAllY cONduCTIve layer at the edGe of its ionOsPHeRe, which at fIRst ApProximaTion forMs an iMPenetraBle obstacLe to the soLaR winD. tHesE systems haVe been stUdied usinG MagneTOhYdrodYnaMic (MHD) MoDelS, and tHis wouLD seEm to bE an appRoPriate StartInG point foR studying the dark plasma wInd intEractIon With the caPtuREd dArk matter WithIn the Earth.
![[the SolAr winD inTEractIon wITh mooN/venus. thesE Systems apPEaR to REPrEsent useful ANALogUes to The POssiblE wayS in which the dark plASma wind interacTs wiTH CapTurED darK | isation, and overall $U(1) '$ neutral ity i mpl ies t heycaneven be much l a rger than the galactic esc ape v el o city . I t has been p o in t e d o ut [ @fo ot 1 4] that th is effe ct can giv e r is e to observa b le keV elect ron recoils indir ect de te cti o n exp eri ments . It m i ght ev en be pos si b le toe xplaint h eDAMA annual modulatio n s i gnal [@dama1a; @dama 1b ; @ d a ma2 a;@dama2b; @ da ma2c] in this ma n n e r,s ince the cons traints one lec tron r ec oil s provi ded b yo the r experimen ts a re genera lly mu c h weake r than t hose o f n ucl earr ec oi ls. H o wev e r, ad eta iled des cr ip tionof t h e p lasm a d arkmatte r density and ve loci t y d istri butio n in t he vi cinity of t he Earth is requi red. This isa h ig hly n on-tr i vial p rob lem . If th e darkp las ma h a sinteractions withor d i na ry matte r then it w i ll by ca pt ure d by t he Ea rth, fo rming an appro x im at e “dark s phere” w ith in. Unde r stan ding t he inter actio n of the dark p l asma with thi s d a r ks pher e a s the Earth mov e s th roug h t heh alo i s the re f or e of primary importa nc e. The aimof this paper is to pro v i d e a cons iste n td escription ofthisinteractio n in orde r toqualitat ively und e r stand th e i mpl ica tio n s f or direct det e c tion .
The ca ptu red dar k s phe rewit hi n the Ear th forms a nob st acl e tot he darkpl asm awin d. Tw o limit ing c ases c an beenvisag e d: ( 1) i fth e ca ptu re d dar k ma t ter is lar gely neut ral , i.e .po orly co nducting, the nthe dark p la sma windw i ll be ab sorbed by the dark sphe r e, or;(2) if t he c aptured d ark matte r i s large ly ion ised, t hen t he ob s t ac lefo rms a cond u c tin g sph er e wh ich eff ectively deflectst hedark plasma w ind . In t e re sti n gl y , t he s e l i m iting cases app ear analog ou s t o the sola r wi nd intera ction w ith t h e Moonand Venus , respect iv ely( e .g. [@Spreite r70; @Lu hmann04]) ; this is sket che d in F ig ure \[Fi gMoonV e nus \]. T he Moo nhas no magn et ic field and no atmosphere, sothat t he so lar wind islar g ely absorbed atthe lunarsur fac e wit h v e ry li ttle up win d acti vity . Venus ha s n o m a g ne tic field,b u t fo rms a n e l ectric ally conductive layer at the edge of its i ono sph e re,wh ich at first a ppr ox i m ation fo rm s an impene trable o bs t acleto the solar wind.T h es e syste ms h ave been stu die du sing ma gn et o hydrod ynam ic (MHD) model s , an d this would seemto be a n app r opr iatest artingp oint for study ing the dar k plas ma w ind i nteract io n with th ecaptured d a rk matter with in theEa rth.
! [[Thesola r windinte ra cti on with M o o n/ V en us . Th esesyste ms app ear to re p resent u sef u l analo gu est o the p o ss i b le ways in wh ich t h e dark plas m a wi n di ntera cts wi th cap tured d a rk | isation, and_overall $U(1)'$_neutrality implies they can_even be_much_larger than_the_galactic escape velocity._It has been_pointed out [@foot14] that_this effect can_give_rise to observable keV electron recoils in direct detection experiments. It might even be_possible_to explain_the_DAMA_annual modulation signal [@dama1a; @dama1b;_@dama2a; @dama2b; @dama2c] in this_manner, since_the constraints on electron recoils provided by other_experiments_are generally much_weaker than those of nuclear recoils. However, a detailed_description of the plasma dark matter_density and velocity_distribution_in_the vicinity of the_Earth is required. This is a_highly non-trivial problem. If the dark_plasma has interactions with ordinary matter then_it will by captured by the_Earth, forming an approximate “dark_sphere” within._Understanding the interaction of the_dark plasma with_this dark_sphere as the_Earth moves through the halo is_therefore of primary_importance. The aim of this paper_is_to provide a_consistent_description_of this_interaction in order_to_qualitatively understand_the_implications for direct detection.
The captured dark_sphere_within the Earth forms an obstacle to_the dark plasma wind._Two_limiting cases can be_envisaged: (1) if the captured_dark matter is largely neutral, i.e._poorly conducting,_then the_dark plasma wind will be absorbed by the dark sphere, or;_(2) if the captured dark matter_is largely ionised, then_the obstacle_forms_a conducting sphere_which_effectively deflects_the dark plasma wind. Interestingly, these limiting_cases appear_analogous to the solar wind interaction_with the Moon and_Venus,_respectively (e.g. [@Spreiter70; @Luhmann04]); this is_sketched in Figure \[FigMoonVenus\]. The Moon has_no magnetic field and no_atmosphere,_so_that the solar wind is_largely absorbed at the lunar surface_with very little_upwind activity. Venus has no magnetic field,_but_forms an electrically conductive layer at_the_edge of its ionosphere, which at_first_approximation_forms an impenetrable obstacle to_the solar wind. These systems have_been studied using magnetohydrodynamic (MHD) models, and this would_seem to be_an appropriate starting point for_studying_the_dark plasma wind interaction with the captured dark matter within_the Earth.
![[The_solar wind interaction_with Moon/Venus. These systems appear to represent useful analogues to_the possible ways in which the dark_plasma wind interacts with captured dark |
k})\\
h(\mathbf{x})
\end{array}\right.\label{eq:Relation I and h}$$
Equation (\[eq:Relation I and h\]) allows to postulate an equivalent condition which constitutes an upper limit for the expected value of the potential energy given the expected value for differential entropies
$$\left\langle V\right\rangle <\left\{ \begin{array}{l}
\left\langle h(t_{k})\right\rangle \\
\left\langle h(\mathbf{x})\right\rangle
\end{array}\right.\label{eq:Relation V and h}$$
where $\left\langle h(t_{k})\right\rangle $ and $\left\langle h(\mathbf{x})\right\rangle $ are given by $$\left\langle h(t_{k})\right\rangle =\iint\Psi^{*}h(t_{k})\Psi\,dt_{k}d\mathbf{x}$$ $$\left\langle h(\mathbf{x})\right\rangle =\int\Psi^{*}h(\mathbf{x})\Psi\,d\mathbf{x}$$
Expressions (\[eq:Relation I and h\]) and (\[eq:Relation V and h\]) establish that the mutual information and ultimately the expected value of the potential energy are upper limited by a maximum value given by the entropy of the distribution of inputs or targets. As we have already discussed, to choose this maximum value as an arbitrary reference point it is desirable to identify the differential entropies from probability densities independent of the problem under consideration. Adequate choices can be the entropy of the uniform distribution or the entropy of the normal distribution with the same variance of the dataset. To analyze more deeply the physical nature of the probability density which gives rise to the maximum entropy, we will now consider the question from a point of view closer to quantum mechanics.
Consider a system defined by a state function $\Psi_{0}$ in which inputs and targets are independent $$p(\mathbf{t},\mathbf{x})=p(\mathbf{t})p(\mathbf{\mathbf{x}})$$ Such a system has zero mutual information. Assuming valid the interpretation we gave in the Section \[subsec:Interpretation-of-the state function\] of the square of the state function as conditional density $p(\mathbf{x}|\mathbf{t})$, together with the equation (\[eq:I and h bis\]), calling $ | k})\\
h(\mathbf{x })
\end{array}\right.\label{eq: Relation I and h}$$
Equation (\[eq: Relation I and h\ ]) allows to postulate an equivalent condition which constitutes an upper terminus ad quem for the expect value of the potential energy give the ask value for differential entropies
$ $ \left\langle V\right\rangle < \left\ { \begin{array}{l }
\left\langle h(t_{k})\right\rangle \\
\left\langle h(\mathbf{x})\right\rangle
\end{array}\right.\label{eq: Relation V and h}$$
where $ \left\langle h(t_{k})\right\rangle $ and $ \left\langle h(\mathbf{x})\right\rangle $ are give by $ $ \left\langle h(t_{k})\right\rangle = \iint\Psi^{*}h(t_{k})\Psi\,dt_{k}d\mathbf{x}$$ $ $ \left\langle h(\mathbf{x})\right\rangle = \int\Psi^{*}h(\mathbf{x})\Psi\,d\mathbf{x}$$
Expressions (\[eq: Relation I and h\ ]) and (\[eq: Relation V and h\ ]) establish that the mutual data and ultimately the expect value of the potential department of energy are upper limited by a maximum value given by the entropy of the distribution of stimulation or targets. As we have already discussed, to choose this maximum value as an arbitrary reference point it is desirable to name the differential entropies from probability densities independent of the trouble under consideration. Adequate choices can be the entropy of the uniform distribution or the entropy of the normal distribution with the like variance of the dataset. To analyze more profoundly the physical nature of the probability density which gives rise to the maximal entropy, we will now consider the question from a point of view closer to quantum mechanics.
Consider a system defined by a state function $ \Psi_{0}$ in which inputs and targets are independent $ $ p(\mathbf{t},\mathbf{x})=p(\mathbf{t})p(\mathbf{\mathbf{x}})$$ Such a system have zero mutual information. Assuming valid the rendition we give in the Section \[subsec: Interpretation - of - the state function\ ] of the public square of the state affair as conditional density $ p(\mathbf{x}|\mathbf{t})$, together with the equality (\[eq: I and h bis\ ]), calling $ | k})\\
h(\mwthbf{x})
\end{array}\right.\label{tq:Relation I and k}$$
Wquatimn (\[eq:Rslation K and h\]) allows to postulate en ewuivaoent condition which cunstituted an uppwr lmmit for the expxdted value of bhe pmvential energy niven the efpected value xof bifferential entropies
$$\left\langle V\ridht\rangke <\left\{ \begin{arrwy}{l}
\ltft\jangmv m(t_{k})\right\rangle \\
\left\langle h(\mathbr{x})\right\gangle
\end{array}\ribht.\label{eq:Relation V and h}$$
aherf $\left\langle h(t_{k})\rihht\rangle $ qnd $\jwft\langle h(\mxthbf{x})\righu\rcngle $ are fiven by $$\left\langle h(t_{k})\right\ranele =\innt\Psi^{*}h(t_{k})\Psu\,dr_{k}d\ldthbf{x}$$ $$\left\oanglv h(\mathbf{x})\rigmn\rangle =\int\Psi^{*}n(\mathbf{x})\Psi\,d\mabhbf{x}$$
Xxprwssions (\[eq:Relation I end h\]) and (\[eq:Relation V and h\]) ascablish that the mutuql infotmatimn ave uutijavelg the fxpxcted value of the potwntial energy are ukpew limited by a maximtm value given by the entropy of the distgibufion of inputs or targers. As we have already discusseq, to choose this maximum value as an arbitrary rexerenre ponkb it us desirable to identify the differential entro[jex nrom probability densities onfelgndent of the kroblem unser consideration. Wdequatg choixes can bt the entropy of the uniform disrribution or rhe entropy of the normal discributoon woth the same variance oy the sataset. To wnalyze mkfe deeply the phhsibal tature of the probability qensity wiich yives rire tp the iaximum enhropy, we will now considfr thg quesdion from w point of view closer to quantnk mechanics.
Cpnvidvr a systzm deflned by a state function $\Psi_{0}$ nn which inpugs and tarfets arx independene $$p(\mathbf{t},\matvhf{x})=p(\mathbf{t})p(\kathbf{\maehbf{z}})$$ Suxh a syrgem has zero mitual information. Awsuming valid the lnterkrstation we gave in the Section \[sunsez:Ineegprxtatijt-of-the state funztium\] of ghe square iw thr state function as wondjtional density $p(\mstmbf{x}|\mathbd{t})$, togetrer with the rquation (\[eq:I and h bis\]), rallinj $ | k})\\ h(\mathbf{x}) \end{array}\right.\label{eq:Relation I and h}$$ Equation and allows to an equivalent condition for expected value of potential energy given expected value for differential entropies $$\left\langle <\left\{ \begin{array}{l} \left\langle h(t_{k})\right\rangle \\ \left\langle h(\mathbf{x})\right\rangle \end{array}\right.\label{eq:Relation V and h}$$ where $\left\langle $ and $\left\langle h(\mathbf{x})\right\rangle $ are given by $$\left\langle h(t_{k})\right\rangle =\iint\Psi^{*}h(t_{k})\Psi\,dt_{k}d\mathbf{x}$$ $$\left\langle h(\mathbf{x})\right\rangle Expressions I h\]) (\[eq:Relation V and h\]) establish that the mutual information and ultimately the expected value of the energy are upper limited by a maximum value by the entropy of distribution of inputs or targets. we already discussed, choose maximum as an arbitrary point it is desirable to identify the differential entropies from probability densities independent of the problem under Adequate choices the entropy the distribution the entropy of distribution with the same variance of analyze more deeply the physical nature of the density which rise to the maximum entropy, we now consider the question from a point of closer to quantum mechanics. Consider a system defined by a state function $\Psi_{0}$ in which targets are independent $$p(\mathbf{t},\mathbf{x})=p(\mathbf{t})p(\mathbf{\mathbf{x}})$$ a system has mutual Assuming the we gave the Section \[subsec:Interpretation-of-the state function\] of the square of the state as conditional density $p(\mathbf{x}|\mathbf{t})$, together with the equation (\[eq:I and calling | k})\\
h(\mathbf{x})
\end{array}\right.\labEl{eq:RelatiOn I anD h}$$
EQuaTiOn (\[eq:relaTion I and h\]) allowS To poStulate an equivalent conDitioN wHIch cONsTitutEs an uppER lIMIt fOr ThE exPeCTeD valuE of The poteNtial energY giVeN the expected VAlUe for diffeRenTial entropieS
$$\leFt\langLe v\riGHt\ranGle <\Left\{ \bEgin{arRAy}{l}
\lefT\langle h(t_{K})\rIGht\ranGLe \\
\left\lANGlE h(\maThbf{x})\right\rangle
\eND{aRRay}\right.\label{eQ:RelatIoN v aND H}$$
whEre $\Left\langle H(t_{K})\righT\Rangle $ aND $\lEFT\LanGLe h(\mathbf{x})\rigHt\rangle $ are GIveN by $$\lefT\lAngLE h(t_{k})\riGht\raNgLE =\iiNt\Psi^{*}h(t_{k})\Psi\,Dt_{k}d\Mathbf{x}$$ $$\leFt\langLE h(\mathbF{X})\right\rAngle =\iNt\PSi^{*}h(\MathBF{x})\psI\,d\mAtHBf{x}$$
eXpResSIonS (\[eq:RelatIoN I And h\]) aNd (\[eq:rELATion v anD h\]) esTabliSh that the mutuAl iNforMAtiOn and UltimAtelY tHe expEcted vAlue oF tHe potential enerGy arE upper limIteD bY a mAxImum vALue givEn bY thE entropY of the dIStrIbUTIOn Of inputs or targets. AS wE HAvE already DiscusSEd, To CHoose thiS mAxiMum vALUe as aN arbITrAry referEnce poINt It Is desirAbLe to idEnTifY thE diffERentIal entRopies frOm proBAbility densitiES independent oF ThE PRoBLem uNdeR consideratIon. ADEquaTe chOIcEs cAN be thE entrOpY Of THe uniform distributiOn Or the eNtropY of the normal dIstributioN WITh the samE varIAnCE of the dataset. TO analYze more deePLy the phySical Nature of The probabILIty densiTy wHicH giVes RISe To the maximum eNTRopy, We Will now ConSider thE quEstIon FroM a Point of viEw closer To QuAnTuM meChaniCS.
ConsideR a SysTeM deFined BY a statE funcTion $\psI_{0}$ iN WhiCh inputS AnD TArgeTs ArE indEpeNdEnt $$p(\mAthbF{T},\maThbf{x})=p(\mAthbf{t})p(\maThbF{\MathBf{X}})$$ SUch a sysTem has zero mutUaL informatiOn. assUming vALId the intErpretation we gave in the SECtion \[suBseC:InteRpreTation-of-tHe sTate fuNctIOn\] of thE squarE of thE sTatE FUnctiON As ConDiTional densITY $p(\mAthbf{X}|\mAthbF{t})$, togetHer with the equation (\[EQ:I aNd h bis\]), calling $ | k})\\
h(\mathbf{x})
\end{a rray}\righ t.\la bel {eq :R elat ionI and h}$$
Eq u atio n (\[eq:Relation I and h\]) a l lows to post ulate a n e q u iva le nt co nd i ti on wh ich consti tutes an u ppe rlimit for th e e xpected va lue of the pote nti al ene rg y g i ven t heexpec ted va l ue for differen ti a l entr o pies
$ $ \ le ft\l angle V\right\ran g le <\left\{ \begi n{arra y} { l} \ lef t\l angle h(t_ {k })\ri g ht\rang l e\ \ \le f t\langle h(\m athbf{x})\r i ght \rangl e
\e n d{arra y}\ri gh t .\l abel{eq:Rel atio n V and h }$$
w h ere $\l e ft\lang le h(t _{k })\ righ t \r an gle $ and $\ lef t \la ngle h(\ ma th bf{x} )\ri g h t \ rang le$ ar e giv en by $$\left \la ngle h(t _{k}) \righ t\ra ng le =\ iint\P si^{* }h (t_{k})\Psi\,dt _{k} d\mathbf{ x}$ $$$\ le ft\la n gle h( \ma thb f{x})\r ight\ra n gle = \ i n t\ Psi^{*}h(\mathbf{x }) \ P si \,d\math bf{x}$ $
Ex p ressions ( \[e q:Re l a tionI an d h \]) and(\[eq: R el at ion V a nd h\])es tab lis h tha t the mutua l inform ation and ultimately the expectedv al u e o f the po tential ene rgya re u pper li mit e d bya max im u mv alue given by the e nt ropy o f the distribution of inputs o r targets . As we have already d iscus sed, to ch o ose this maxi mum valu e as an a r b itrary r efe ren cepoi n t i t is desirabl e to i de ntify t hediffere nti alent rop ie s from pr obabilit yde ns it ies inde p endent o fthe p rob lem u n der co nside rati on .A deq uate ch o ic e s can b etheent ro py of the uni form di stributio n o r the e nt ropy of the normal d is tributionwi ththe sa m e varianc e of the dataset. To an a lyze mo redeepl y th e physica l n atureoft he pro babili ty de ns ity w hichg i ve s r is e to the m a x imu m ent ro py,we will now consider theq ues tion from a p oin t of v ie w c l os e r t oq uan t u m mechanics.
C onsider asy s te m definedb y a s tate fu nction$\Psi _ {0}$ in which in puts andta rget s are independe nt $$p(\ mathbf{t} , \math b f{ x})=p (\m athbf{ t} )p( \math bf{\ma t hbf {x}}) $$ Suc ha syst em ha szero mut ual information. Assumi ng val id th e i nterpreta tio n we gave intheSection \[ sub sec :Inte rpr e tatio n-of - th e s t ate f unct i on\] of t h esqu a r eof the stat e f unc tionasc onditi onal density $p(\math b f{x}|\mathbf{t })$, t oge the r wit hthe equation ( \[e q: I and h bi s\ ]), calling $ | k})\\
h(\mathbf{x})
\end{array}\right.\label{eq:Relation I_and h}$$
Equation_(\[eq:Relation I and h\])_allows to_postulate_an equivalent_condition_which constitutes an_upper limit for_the expected value of_the potential energy_given_the expected value for differential entropies
$$\left\langle V\right\rangle <\left\{ \begin{array}{l}
\left\langle h(t_{k})\right\rangle \\
\left\langle h(\mathbf{x})\right\rangle
\end{array}\right.\label{eq:Relation V_and_h}$$
where $\left\langle_h(t_{k})\right\rangle_$_and $\left\langle h(\mathbf{x})\right\rangle $ are_given by $$\left\langle h(t_{k})\right\rangle =\iint\Psi^{*}h(t_{k})\Psi\,dt_{k}d\mathbf{x}$$_$$\left\langle h(\mathbf{x})\right\rangle_=\int\Psi^{*}h(\mathbf{x})\Psi\,d\mathbf{x}$$
Expressions (\[eq:Relation I and h\]) and (\[eq:Relation V_and_h\]) establish that_the mutual information and ultimately the expected value of_the potential energy are upper limited_by a maximum_value_given_by the entropy of_the distribution of inputs or targets._As we have already discussed, to_choose this maximum value as an arbitrary_reference point it is desirable to_identify the differential entropies from_probability densities_independent of the problem under_consideration. Adequate choices_can be_the entropy of_the uniform distribution or the entropy_of the normal_distribution with the same variance of_the_dataset. To analyze_more_deeply_the physical_nature of the_probability_density which_gives_rise to the maximum entropy, we_will_now consider the question from a point_of view closer to_quantum_mechanics.
Consider a system defined_by a state function $\Psi_{0}$_in which inputs and targets are_independent $$p(\mathbf{t},\mathbf{x})=p(\mathbf{t})p(\mathbf{\mathbf{x}})$$_Such a_system has zero mutual information. Assuming valid the interpretation we gave_in the Section \[subsec:Interpretation-of-the state function\]_of the square of_the state_function_as conditional density_$p(\mathbf{x}|\mathbf{t})$,_together with_the equation (\[eq:I and h bis\]), calling_$ |
A_1}\|_{L^p(X)} \|a_1\|_{W_{A_1}^{1,p}(X)},$$ which again yields for this case.
An argument identical to that for gives $$\|\nabla_{A_1}\Phi\times a_1\|_{L^{r'}(X)}
\leq
z\|\nabla_{A_1}\Phi\|_{L^p(X)} \|a_1\|_{W_{A_1}^{1,p}(X)}.$$ proving. This completes Step \[step:Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\].
\[step:Lrprime\_estimate\_Phi\_times\_nablaA1Phi\_and\_two\_similar\_terms\] We claim that $$\begin{aligned}
\label{eq:Lrprime_bound_Phi_times_nablaA1Phi}
\|\nabla_{A_1}\Phi\times \Phi\|_{L^{r'}(X)}
&\leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}^2,
\\
\label{eq:Lrprime_bound_Phi_times_nablaA1a1}
\|\nabla_{A_1}a_1\times \Phi\|_{L^{r'}(X)}
&\leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}\|a_1\|_{W_{A_1}^{1,p}(X)},
\\
\label{eq:Lrprime_bound_a1_times_nablaA1a1}
\|\nabla_{A_1}a_1\times a_1\|_{L^{r'}(X)}
&\leq z\|a_1\|_{W_{A_1}^{1,p}(X)}^2\end{aligned}$$ where $z = z(g,p) \in [1,\infty)$.
From the proof of in Step \[step:Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\], we have $$\|\Phi\times \nabla_{A_1}\Phi\|_{L^{r'}(X)} \leq z\|\Phi\|_{W_{A_1}^{1,p}(X)} \|\nabla_{A_1}\Phi\|_{L^p(X)},$$ and this gives ; identical arguments give | A_1}\|_{L^p(X) } \|a_1\|_{W_{A_1}^{1,p}(X)},$$ which again yields for this case.
An argument identical to that for gives $ $ \|\nabla_{A_1}\Phi\times a_1\|_{L^{r'}(X) }
\leq
z\|\nabla_{A_1}\Phi\|_{L^p(X) } \|a_1\|_{W_{A_1}^{1,p}(X)}.$$ proving. This dispatch Step \[step: Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\ ].
\[step: Lrprime\_estimate\_Phi\_times\_nablaA1Phi\_and\_two\_similar\_terms\ ] We claim that $ $ \begin{aligned }
\label{eq: Lrprime_bound_Phi_times_nablaA1Phi }
\|\nabla_{A_1}\Phi\times \Phi\|_{L^{r'}(X) }
& \leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}^2,
\\
\label{eq: Lrprime_bound_Phi_times_nablaA1a1 }
\|\nabla_{A_1}a_1\times \Phi\|_{L^{r'}(X) }
& \leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}\|a_1\|_{W_{A_1}^{1,p}(X) },
\\
\label{eq: Lrprime_bound_a1_times_nablaA1a1 }
\|\nabla_{A_1}a_1\times a_1\|_{L^{r'}(X) }
& \leq z\|a_1\|_{W_{A_1}^{1,p}(X)}^2\end{aligned}$$ where $ omega = z(g, p) \in [ 1,\infty)$.
From the proof of in Step \[step: Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\ ], we have $ $ \|\Phi\times \nabla_{A_1}\Phi\|_{L^{r'}(X) } \leq z\|\Phi\|_{W_{A_1}^{1,p}(X) } \|\nabla_{A_1}\Phi\|_{L^p(X)},$$ and this give; identical controversy give | A_1}\|_{L^p(D)} \|a_1\|_{W_{A_1}^{1,p}(X)},$$ which again yielas for this casg.
Ab argukent isentical to that for gives $$\|\nabla_{A_1}\Phi\vimew a_1\|_{L^{r'}(Z)}
\leq
z\|\nabla_{A_1}\Phi\|_{L^p(X)} \|a_1\|_{W_{A_1}^{1,p}(B)}.$$ proving. This conpleues Step \[step:Lrprmje\_estimate\_FA1\_tjles\_a1\_cnv\_nablaA1Phi\_times\_s1\].
\[step:Lrprike\_estimate\_Phi\_thmds\_uablaA1Phi\_and\_two\_similar\_terms\] We claim that $$\brgln{aligned}
\label{gq:Lrpgiie_bohnd_Phi_times_nablaA1Phi}
\|\nabla_{A_1}\Phi\times \Phi\|_{L^{r'}(X)}
&\neq z\|\Phi\|_{W_{A_1}^{1,p}(X)}^2,
\\
\lsbel{eq:Lrprime_bound_Phi_times_jablwA1a1}
\|\nabla_{A_1}a_1\times \Phl\|_{L^{r'}(X)}
&\leq z\|\Phu\|_{W_{A_1}^{1,p}(V)}\|q_1\|_{W_{A_1}^{1,p}(X)},
\\
\label{eq:Urprime_bound_a1_times_nablzA1a1}
\|\nabla_{A_1}a_1\times a_1\|_{L^{r'}(X)}
&\leq z\|a_1\|_{W_{A_1}^{1,p}(X)}^2\evd{aliyned}$$ where $e = z(h,k) \in [1,\infty)$.
Frim thv proof of in Step \[stap:Lrprike\_estimate\_FA1\_tikes\_e1\_and\_bablaA1Phi\_times\_a1\], we hate $$\|\Phi\times \nabla_{A_1}\Phy\|_{L^{r'}(X)} \leq s\|\Pki\|_{W_{A_1}^{1,p}(X)} \|\nabla_{A_1}\Phi\|_{L^p(X)},$$ abd this givas ; keengiczl adgumenhs jive | A_1}\|_{L^p(X)} \|a_1\|_{W_{A_1}^{1,p}(X)},$$ which again yields for this argument to that gives $$\|\nabla_{A_1}\Phi\times a_1\|_{L^{r'}(X)} completes \[step:Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\]. \[step:Lrprime\_estimate\_Phi\_times\_nablaA1Phi\_and\_two\_similar\_terms\] We that $$\begin{aligned} \label{eq:Lrprime_bound_Phi_times_nablaA1Phi} \Phi\|_{L^{r'}(X)} &\leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}^2, \\ \label{eq:Lrprime_bound_Phi_times_nablaA1a1} \|\nabla_{A_1}a_1\times &\leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}\|a_1\|_{W_{A_1}^{1,p}(X)}, \\ \label{eq:Lrprime_bound_a1_times_nablaA1a1} \|\nabla_{A_1}a_1\times a_1\|_{L^{r'}(X)} &\leq z\|a_1\|_{W_{A_1}^{1,p}(X)}^2\end{aligned}$$ where $z = z(g,p) \in From the proof of in Step \[step:Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\], we have $$\|\Phi\times \nabla_{A_1}\Phi\|_{L^{r'}(X)} \leq z\|\Phi\|_{W_{A_1}^{1,p}(X)} and gives identical give | A_1}\|_{L^p(X)} \|a_1\|_{W_{A_1}^{1,p}(X)},$$ which again yields For this casE.
An arGumEnt IdEntiCal tO that for gives $$\|\nABla_{A_1}\phi\times a_1\|_{L^{r'}(X)}
\leq
z\|\nabla_{A_1}\phi\|_{L^p(x)} \|a_1\|_{w_{a_1}^{1,p}(X)}.$$ pROvIng. ThIs complETeS sTep \[StEp:lrpRiME\_eStimaTe\_Fa1\_times\_a1\_And\_nablaA1PHi\_tImEs\_a1\].
\[step:LrpriME\_eStimate\_Phi\_TimEs\_nablaA1Phi\_aNd\_tWo\_simiLaR\_teRMs\] We cLaiM that $$\Begin{aLIgned}
\lAbel{eq:LrpRiME_bound_pHi_times_NABlAA1PhI}
\|\nabla_{A_1}\Phi\times \PhI\|_{l^{r'}(x)}
&\Leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}^2,
\\
\labEl{eq:LrPrIMe_BOUnd_phi_Times_nablaa1a1}
\|\Nabla_{a_1}A_1\times \PHI\|_{L^{R'}(x)}
&\LEq z\|\pHi\|_{W_{A_1}^{1,p}(X)}\|a_1\|_{W_{A_1}^{1,p}(X)},
\\
\laBel{eq:LrprimE_BouNd_a1_timEs_NabLAA1a1}
\|\nabLa_{A_1}a_1\tImES a_1\|_{L^{R'}(X)}
&\leq z\|a_1\|_{W_{A_1}^{1,p}(X)}^2\End{aLigned}$$ wheRe $z = z(g,p) \IN [1,\infty)$.
FROm the prOof of iN StEp \[sTep:LRPrImE\_esTiMAte\_fa1\_tImeS\_A1\_anD\_nablaA1PHi\_TiMes\_a1\], wE havE $$\|\pHI\TimeS \naBla_{A_1}\phi\|_{L^{r'}(x)} \leq z\|\Phi\|_{W_{A_1}^{1,p}(X)} \|\nAblA_{A_1}\PhI\|_{l^p(X)},$$ And thIs givEs ; idEnTical ArgumeNts giVe | A_1}\|_{L^p(X)} \|a_1\|_{W _{A_1}^{1, p}(X) },$ $ w hi ch a gain yields for th i s ca se.
An argument ident icalto that fo r giv es $$\| \ na b l a_{ A_ 1} \Ph i\ t im es a_ 1\| _{L^{r' }(X)}
\leq
z\ |\ nabla_{A_1}\ P hi \|_{L^p(X) } \ |a_1\|_{W_{A _1} ^{1,p} (X )}. $ $ pro vin g. Th is com p letesStep \[st ep : Lrprim e \_estim a t e\ _FA1 \_times\_a1\_and\ _ na b laA1Phi\_times \_a1\] .\[ s t ep: Lrp rime\_esti ma te\_P h i\_time s \_ n a b laA 1 Phi\_and\_two \_similar\_ t erm s\] We c lai m that$$\be gi n {al igned}
\lab el{e q:Lrprime _bound _ Phi_tim e s_nabla A1Phi}
\| \na bla_ { A_ 1} \Ph i\ t ime s \ Phi \ |_{ L^{r'}(X )}
& \leqz\|\ P h i \ |_{W _{A _1}^ {1,p} (X)}^2,
\\
\l abe l{eq : Lrp rime_ bound _Phi _t imes_ nablaA 1a1}\| \nabla_{A_1}a_1 \tim es \Phi\| _{L ^{ r'} (X )}
&\ l eq z\| \Ph i\| _{W_{A_ 1}^{1,p } (X) }\ | a _ 1\ |_{W_{A_1}^{1,p}(X )} , \\
\label{ eq:Lrp r im e_ b ound_a1_ ti mes _nab l a A1a1}
\|\ n ab la_{A_1} a_1\ti m es a _1\|_{L ^{ r'}(X) }&\l eqz\|a_ 1 \|_{ W_{A_1 }^{1,p}( X)}^2 \ end{aligned}$$ where $z = z( g ,p ) \i n [1, \in fty)$.
Fro m th e pro of o f i n S t ep \[ step: Lr p ri m e\_estimate\_FA1\_t im es\_a1 \_and \_nablaA1Phi\ _times\_a1 \ ] , we have $$\ | \P h i\times \nabla _{A_1 }\Phi\|_{L ^ {r'}(X)} \leq z\|\Phi \|_{W_{A_ 1 } ^{1,p}(X )}\|\ nab la_ { A _1 }\Phi\|_{L^p( X ) },$$ a nd this gi ves ; i den tic alarg um ents give | A_1}\|_{L^p(X)} \|a_1\|_{W_{A_1}^{1,p}(X)},$$_which again_yields for this case.
An_argument identical_to_that for_gives_$$\|\nabla_{A_1}\Phi\times a_1\|_{L^{r'}(X)}
\leq
z\|\nabla_{A_1}\Phi\|_{L^p(X)} \|a_1\|_{W_{A_1}^{1,p}(X)}.$$_proving. This completes_Step \[step:Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\].
\[step:Lrprime\_estimate\_Phi\_times\_nablaA1Phi\_and\_two\_similar\_terms\] We claim_that $$\begin{aligned}
\label{eq:Lrprime_bound_Phi_times_nablaA1Phi}
\|\nabla_{A_1}\Phi\times \Phi\|_{L^{r'}(X)}
&\leq_z\|\Phi\|_{W_{A_1}^{1,p}(X)}^2,
\\
\label{eq:Lrprime_bound_Phi_times_nablaA1a1}
\|\nabla_{A_1}a_1\times_\Phi\|_{L^{r'}(X)}
&\leq z\|\Phi\|_{W_{A_1}^{1,p}(X)}\|a_1\|_{W_{A_1}^{1,p}(X)},
\\
\label{eq:Lrprime_bound_a1_times_nablaA1a1}
\|\nabla_{A_1}a_1\times a_1\|_{L^{r'}(X)}
&\leq z\|a_1\|_{W_{A_1}^{1,p}(X)}^2\end{aligned}$$ where $z = z(g,p) \in [1,\infty)$.
From the proof of in_Step_\[step:Lrprime\_estimate\_FA1\_times\_a1\_and\_nablaA1Phi\_times\_a1\], we_have_$$\|\Phi\times_\nabla_{A_1}\Phi\|_{L^{r'}(X)} \leq z\|\Phi\|_{W_{A_1}^{1,p}(X)} \|\nabla_{A_1}\Phi\|_{L^p(X)},$$ and_this gives ; identical arguments_give |
*$ stands for the convolution product), where $\theta$ is a radially symmetrical function, such that $\int_{\mathbb R ^3} \theta (\textbf{x})d\textbf{x}=1$ and $ \theta_\epsilon (\textbf{x})= \frac{1}{\epsilon^3}\theta (\textbf{x}/\epsilon)$[@RobVar08].
Numerical study
===============
We would like now to study the statistical properties of the velocity field defined by Eq. (\[eq:K41Field\]). Analytical formulas are difficult to obtain, thus we will focus on numerical simulations. To do so, one has to choose the short time scale $t=\tau$. It is easy to check that the variance of the matrix $\textbf{S}_0$ (defined as $\langle \mbox{tr } \textbf{S}_0^2\rangle$) goes to infinity as the small scale parameter $\epsilon$ goes to zero. So we take for $\tau$ the local normalizing value $\tau = (\mbox{tr } \textbf{S}_0^2)^{-1/2}$, in the spirit of the RFD closures provided in Ref. [@CheA2].
The simulation is performed in a 1-periodic box with $N^3$ collocation points. The infinitesimal volume is given by $dV=dx^3$, with $dx=1/N$. We choose as a regularizing function and large-scale cut-off the isotropic normalized Gaussian function $\varphi_{L} (\textbf{x}) = \theta_{L}(\textbf{x}) = \left(\frac{6}{\pi L^2}\right)^{3/2}\exp (-6|\textbf{x}|^2/L^2)$. This allows to compute analytically the regularized norm of a vector $\textbf{x}$, useful for numerical purposes: $|\textbf{x}|_\epsilon = \frac{\epsilon}{\sqrt{6\pi}}e^{-6\frac{|\textbf{x}|^2}{\epsilon^2}} + \left(|\textbf{x}| + \frac{\epsilon^2}{12|\textbf{x}|}\right)\mbox{erf}\left( \frac{\sqrt{6}|\textbf{x}|}{\epsilon}\right)\mbox{,}$ with erf the error function. The small-scale cut-off is chosen as $\epsilon=2dx$ and the large one as $L=1/2$. The kernels | * $ stands for the convolution product), where $ \theta$ is a radially symmetrical affair, such that $ \int_{\mathbb radius ^3 } \theta (\textbf{x})d\textbf{x}=1 $ and $ \theta_\epsilon (\textbf{x})= \frac{1}{\epsilon^3}\theta (\textbf{x}/\epsilon)$[@RobVar08 ].
Numerical study
= = = = = = = = = = = = = = =
We would like now to learn the statistical properties of the velocity playing field define by Eq. (\[eq: K41Field\ ]). Analytical formula are difficult to obtain, therefore we will focus on numerical simulations. To do therefore, one has to choose the short time scale $ t=\tau$. It is easy to see that the variance of the matrix $ \textbf{S}_0 $ (defined as $ \langle \mbox{tr } \textbf{S}_0 ^ 2\rangle$) goes to eternity as the small scale argument $ \epsilon$ goes to zero. then we take for $ \tau$ the local normalizing value $ \tau = (\mbox{tr } \textbf{S}_0 ^ 2)^{-1/2}$, in the heart of the RFD closures provided in Ref. [ @CheA2 ].
The simulation is perform in a 1 - periodic box with $ N^3 $ collocation points. The infinitesimal volume is given by $ dV = dx^3 $, with $ dx=1 / N$. We choose as a regularizing function and large - scale cut - off the isotropic normalized Gaussian function $ \varphi_{L } (\textbf{x }) = \theta_{L}(\textbf{x }) = \left(\frac{6}{\pi L^2}\right)^{3/2}\exp (-6|\textbf{x}|^2 / L^2)$. This allows to compute analytically the regularized norm of a vector $ \textbf{x}$, useful for numerical purpose: $ |\textbf{x}|_\epsilon = \frac{\epsilon}{\sqrt{6\pi}}e^{-6\frac{|\textbf{x}|^2}{\epsilon^2 } } + \left(|\textbf{x}| + \frac{\epsilon^2}{12|\textbf{x}|}\right)\mbox{erf}\left (\frac{\sqrt{6}|\textbf{x}|}{\epsilon}\right)\mbox{,}$ with erf the error function. The belittled - scale cut - off is chosen as $ \epsilon=2dx$ and the large one as $ L=1/2$. The kernel | *$ stwnds for the convolution product), where $\jhwta$ is a radjally syometrical function, such that $\ibt_{\matybb R ^3} \theta (\textbf{x})d\tdxtbf{x}=1$ anf $ \theta_\wpsioin (\textbf{x})= \frac{1}{\epsljon^3}\tgcta (\tzxvbf{x}/\epsilon)$[@RobVsr08].
Numerican study
===============
We woulg uine now to study the statistical pro[erties ov the velocity fiekq derpntd by Eq. (\[eq:K41Field\]). Analytical formhlas art difficult to obtsin, thus we will focus on jumegical simulations. Ho do so, ong haf to choose tfe short tpke scale $t=\jau$. It is easy to check that the varicnce of the mqtrlf $\textbf{S}_0$ (dxfined as $\langle \mnpx{tr } \dextbf{S}_0^2\tangle$) goes to inhiniry as the small scale parameter $\epsilon$ goes to sexo. So we take for $\tau$ tye lowal tormxoizkng velus $\tau = (\muox{tr } \textgf{S}_0^2)^{-1/2}$, in the wpirit of the RFD ckofltes provided in Res. [@SheA2].
The simulation is performed in a 1-pegiodjc box with $N^3$ collocatiin points. The infinitgsimal voltme is given by $dV=dx^3$, with $dx=1/N$. We choose as a regunarizmne fbkgtiov ajd large-scale cut-off the isotropic normalized Faisxian function $\yarphi_{L} (\textbf{x}) = \tneha_{K}(\jextbf{x}) = \left(\ftac{6}{\pi L^2}\xjggt)^{3/2}\exp (-6|\textbf{x}|^2/L^2)$. Thid allowf to xompute agalyyically the regularized norn of a vectog $\teztbf{x}$, useful for nbmerical pur'oses: $|\jextbf{c}|_\epsilon = \frac{\epsilon}{\sqxt{6\pi}}e^{-6\fdac{|\textbf{x}|^2}{\eosilon^2}} + \lswt(|\textbf{x}| + \frac{\eosikot^2}{12|\textbf{x}|}\right)\mbox{erf}\left( \fwac{\sqrt{6}|\teetbf{x}|}{\zpsilon}\rkght)\kbox{,}$ wyth erf thf error function. The smapl-scape cut-off is chosen as $\epsilon=2dx$ and the lacje one as $L=1/2$. Tne kegnels | *$ stands for the convolution product), where a symmetrical function, that $\int_{\mathbb R \theta_\epsilon \frac{1}{\epsilon^3}\theta (\textbf{x}/\epsilon)$[@RobVar08]. Numerical =============== We would now to study the statistical properties the velocity field defined by Eq. (\[eq:K41Field\]). Analytical formulas are difficult to obtain, we will focus on numerical simulations. To do so, one has to choose short scale It easy to check that the variance of the matrix $\textbf{S}_0$ (defined as $\langle \mbox{tr } \textbf{S}_0^2\rangle$) to infinity as the small scale parameter $\epsilon$ to zero. So we for $\tau$ the local normalizing $\tau (\mbox{tr } in spirit the RFD closures in Ref. [@CheA2]. The simulation is performed in a 1-periodic box with $N^3$ collocation points. The infinitesimal is given with $dx=1/N$. choose a function and large-scale isotropic normalized Gaussian function $\varphi_{L} (\textbf{x}) \left(\frac{6}{\pi L^2}\right)^{3/2}\exp (-6|\textbf{x}|^2/L^2)$. This allows to compute analytically regularized norm a vector $\textbf{x}$, useful for numerical $|\textbf{x}|_\epsilon = \frac{\epsilon}{\sqrt{6\pi}}e^{-6\frac{|\textbf{x}|^2}{\epsilon^2}} + \left(|\textbf{x}| + \frac{\epsilon^2}{12|\textbf{x}|}\right)\mbox{erf}\left( \frac{\sqrt{6}|\textbf{x}|}{\epsilon}\right)\mbox{,}$ erf the error function. The small-scale cut-off is chosen as $\epsilon=2dx$ and the large one The kernels | *$ stands for the convolution prOduct), where $\Theta$ Is a RadIaLly sYmmeTrical function, SUch tHat $\int_{\mathbb R ^3} \theta (\textBf{x})d\tExTBf{x}=1$ aND $ \tHeta_\ePsilon (\tEXtBF{X})= \frAc{1}{\EpSilOn^3}\THeTa (\texTbf{X}/\epsiloN)$[@RobVar08].
NumEriCaL study
===============
We woulD LiKe now to stuDy tHe statisticaL prOpertiEs Of tHE veloCitY fielD definED by Eq. (\[eQ:K41Field\]). AnAlYTical fORmulas aRE DiFficUlt to obtain, thus we WIlL Focus on numericAl simuLaTIoNS. to dO so, One has to chOoSe the SHort timE ScALE $T=\taU$. it is easy to cheCk that the vaRIanCe of thE mAtrIX $\textbF{S}_0$ (defInED as $\Langle \mbox{tR } \texTbf{S}_0^2\ranglE$) goes tO InfinitY As the smAll scaLe pAraMeteR $\EpSiLon$ GoES to ZErO. So WE taKe for $\tau$ ThE lOcal nOrmaLIZINg vaLue $\Tau = (\mBox{tr } \Textbf{S}_0^2)^{-1/2}$, in the sPirIt of THe RfD cloSures ProvIdEd in REf. [@CheA2].
the siMuLation is performEd in A 1-periodic Box WiTh $N^3$ CoLlocaTIon poiNts. the InfinitEsimal vOLumE iS GIVeN by $dV=dx^3$, with $dx=1/N$. We chOoSE As A regularIzing fUNcTiON and largE-sCalE cut-OFF the iSotrOPiC normaliZed GauSSiAn FunctioN $\vArphi_{L} (\TeXtbF{x}) = \tHeta_{L}(\TExtbF{x}) = \left(\Frac{6}{\pi L^2}\rIght)^{3/2}\eXP (-6|\textbf{x}|^2/L^2)$. This aLLows to compute ANaLYTiCAlly The Regularized Norm OF a veCtor $\TExTbf{X}$, UsefuL for nUmERiCAl purposes: $|\textbf{x}|_\epSiLon = \fraC{\epsiLon}{\sqrt{6\pi}}e^{-6\fraC{|\textbf{x}|^2}{\epSILOn^2}} + \left(|\teXtbf{X}| + \FrAC{\epsilon^2}{12|\textbf{X}|}\righT)\mbox{erf}\leFT( \frac{\sqrT{6}|\textBf{x}|}{\epsilOn}\right)\mbOX{,}$ With erf tHe eRroR fuNctION. THe small-scale cUT-Off iS cHosen as $\EpsIlon=2dx$ aNd tHe lArgE onE aS $L=1/2$. The kernEls | *$ stands for the convolut ion produc t), w her e $ \t heta $ is a radially sy m metr ical function, such th at $\ in t _{\m a th bb R^3} \th e ta ( \te xt bf {x} )d \ te xtbf{ x}= 1$ and$ \theta_\ eps il on (\textbf{ x }) = \frac{1} {\e psilon^3}\th eta (\tex tb f{x } /\eps ilo n)$[@ RobVar 0 8].
N umericalst u dy
=== = ======= = = ==
We would like now t o s t udy the statis ticalpr o pe r t ies of the veloc it y fie l d defin e db y Eq. (\[eq:K41Fiel d\]). Analy t ica l form ul asa re dif ficul tt o o btain, thus wewill focu s on n u merical simulat ions.Todoso,o ne h asto cho o se th e sh ort time s ca le $t =\ta u $ . It i s e asyto ch eck that thevar ianc e of thematri x $\ te xtbf{ S}_0$(defi ne d as $\langle \ mbox {tr } \te xtb f{ S}_ 0^ 2\ran g le$) g oes to infini ty as t h e s ma l l sc ale parameter $\ep si l o n$ goes to zero. So w e take fo r$\t au$t h e loc al n o rm alizingvalue$ \t au = (\m bo x{tr } \ tex tbf {S}_0 ^ 2)^{ -1/2}$ , in the spir i t of the RFD c l osures provid e di n R e f. [ @Ch eA2].
Thesimu l atio n is pe rfo r med i n a 1 -p e ri o dic box with $N^3$co llocat ion p oints. The in finitesima l v olume is giv e nb y $dV=dx^3$, w ith $ dx=1/N$. W e chooseas aregulari zing func t i on and l arg e-s cal e c u t -o ff the isotro p i c no rm alizedGau ssian f unc tio n $ \va rp hi_{L} (\ textbf{x }) = \ th eta _{L}( \ textbf{x }) =\l eft (\fra c {6}{\p i L^2 }\ri gh t) ^ {3/ 2}\exp( -6 | \ text bf {x }|^2 /L^ 2) $. Th is a l low s to co mpute ana lyt i call yth e regul arized norm o fa vector $ \t ext bf{x}$ , useful f or numerical purposes:$ |\textb f{x }|_\e psil on = \fra c{\ epsilo n}{ \ sqrt{6 \pi}}e ^{-6\ fr ac{ | \ textb f { x} |^2 }{ \epsilon^2 } } +\left (| \tex tbf{x}| + \frac{\epsilon^ 2 }{1 2|\textbf{x}| }\r ight ) \ mb ox{ e rf } \le ft ( \f r a c{\sqrt{6}|\tex tbf{x}|}{\ ep s il on}\right) \ mbo x{ ,}$ wit h erf t he er r or func tion. The small-sc al e cu t - off is chosen as $\ep silon=2dx $ andt he larg e o ne as$L =1/ 2$. T he ker n els | *$ stands_for the_convolution product), where $\theta$_is a_radially_symmetrical function,_such_that $\int_{\mathbb R_^3} \theta (\textbf{x})d\textbf{x}=1$_and $ \theta_\epsilon (\textbf{x})=_\frac{1}{\epsilon^3}\theta (\textbf{x}/\epsilon)$[@RobVar08].
Numerical study
===============
We_would_like now to study the statistical properties of the velocity field defined by Eq._(\[eq:K41Field\])._Analytical formulas_are_difficult_to obtain, thus we will_focus on numerical simulations. To_do so,_one has to choose the short time scale_$t=\tau$._It is easy_to check that the variance of the matrix $\textbf{S}_0$_(defined as $\langle \mbox{tr } \textbf{S}_0^2\rangle$)_goes to infinity_as_the_small scale parameter $\epsilon$_goes to zero. So we take_for $\tau$ the local normalizing value_$\tau = (\mbox{tr } \textbf{S}_0^2)^{-1/2}$, in_the spirit of the RFD closures_provided in Ref. [@CheA2].
The simulation_is performed_in a 1-periodic box with_$N^3$ collocation points._The infinitesimal_volume is given_by $dV=dx^3$, with $dx=1/N$. We choose_as a regularizing_function and large-scale cut-off the isotropic_normalized_Gaussian function $\varphi_{L}_(\textbf{x})_=_\theta_{L}(\textbf{x}) =_\left(\frac{6}{\pi L^2}\right)^{3/2}\exp (-6|\textbf{x}|^2/L^2)$._This_allows to_compute_analytically the regularized norm of a_vector_$\textbf{x}$, useful for numerical purposes: $|\textbf{x}|_\epsilon =_\frac{\epsilon}{\sqrt{6\pi}}e^{-6\frac{|\textbf{x}|^2}{\epsilon^2}} + \left(|\textbf{x}| +_\frac{\epsilon^2}{12|\textbf{x}|}\right)\mbox{erf}\left(_\frac{\sqrt{6}|\textbf{x}|}{\epsilon}\right)\mbox{,}$ with erf the_error function. The small-scale cut-off_is chosen as $\epsilon=2dx$ and the_large one_as $L=1/2$._The kernels |
$)$.$ Let $A$ be the operator in $l_{q}\left( N\right) $ defined by$$\text{ }A=\left[ a_{mj}\right] \text{, }a_{mj}=g_{m}2^{sj},\text{ }m,j=1,2,...,N,\text{ }D\left( A\right) =\text{ }l_{q}^{s}\left( N\right) =$$
$$\left\{ \text{ }u=\left\{ u_{j}\right\},\text{ }j=1,2,...N,\left\Vert
u\right\Vert _{l_{q}^{s}\left( N\right) }=\left(
\sum\limits_{j=1}^{N}2^{sj}u_{j}^{q}\right) ^{\frac{1}{q}}<\infty \right\}.$$
From Theorem 3.1 we obtain the following result
**Theorem 4.1.** Assume:
\(1) the Condition 4.1 hold and $0\leq \gamma <\frac{1}{2}$.
\(2) the function $u\rightarrow $ $F\left( u\right) $: $R^{n}\times \left[ 0,T\right] \times E_{0q}\rightarrow l_{q}$ is a measurable function in $\left(
x,t\right) \in R^{n}\times \left[ 0,T\right] $ for $u\in E_{0q};$
\(3) the function $u\rightarrow $ $F\left( u\right) $ is continuous in $u\in
E_{0q}$ for $x,$ $t\in R^{n}\times \left[ 0,T\right] ;$ moreover $F\left(
u\right) \in C^{\left( 1\right) }\left( E_{0q};l_{q}\right) $.
Then problem $\left( 4.1\right) -\left( 4.2\right) $ has a unique local strange solution$$u\in C^{\left( 2\right) }\left( \left[ 0\right.,\left. T_{0}\right)
;Y_{\infty }^{s,p,q,\sigma }\right),$$where $T_{0 | $) $ .$ Let $ A$ be the operator in $ l_{q}\left (N\right) $ defined by$$\text { } A=\left [ a_{mj}\right ] \text {, } a_{mj}=g_{m}2^{sj},\text { } m, j=1,2,... ,N,\text { } D\left (A\right) = \text { } l_{q}^{s}\left (N\right) = $ $
$ $ \left\ { \text { } u=\left\ { u_{j}\right\},\text { } j=1,2,... N,\left\Vert
u\right\Vert _ { l_{q}^{s}\left (N\right) } = \left (
\sum\limits_{j=1}^{N}2^{sj}u_{j}^{q}\right) ^{\frac{1}{q}}<\infty \right\}.$$
From Theorem 3.1 we obtain the following resultant role
* * Theorem 4.1. * * Assume:
\(1) the Condition 4.1 delay and $ 0\leq \gamma < \frac{1}{2}$.
\(2) the function $ u\rightarrow $ $ F\left (u\right) $: $ R^{n}\times \left [ 0,T\right ] \times E_{0q}\rightarrow l_{q}$ is a measurable function in $ \left (
x, t\right) \in R^{n}\times \left [ 0,T\right ] $ for $ u\in E_{0q};$
\(3) the function $ u\rightarrow $ $ F\left (u\right) $ is continuous in $ u\in
E_{0q}$ for $ x,$ $ t\in R^{n}\times \left [ 0,T\right ]; $ furthermore $ F\left (
u\right) \in C^{\left (1\right) } \left (E_{0q};l_{q}\right) $.
Then problem $ \left (4.1\right) -\left (4.2\right) $ has a unique local strange solution$$u\in C^{\left (2\right) } \left (\left [ 0\right. ,\left. T_{0}\right)
; Y_{\infty } ^{s, phosphorus, q,\sigma } \right),$$where $ T_{0 | $)$.$ Lft $A$ be the operator in $u_{q}\left( N\right) $ bwfined by$$\texf{ }A=\left[ x_{mj}\right] \text{, }a_{mj}=g_{m}2^{sj},\text{ }m,j=1,2,...,I,\texr{ }D\ledt( A\right) =\text{ }l_{q}^{s}\left( N\right) =$$
$$\lvft\{ \text{ }y=\lefu\{ u_{j}\right\},\text{ }j=1,2,...N,\lxrt\Vert
u\vnght\Vsvt _{l_{q}^{v}\oeft( N\right) }=\lent(
\sum\limits_{b=1}^{N}2^{sj}u_{j}^{q}\right) ^{\fsaz{1}{q}}<\nnfty \right\}.$$
From Theorem 3.1 we obtain tre follpwlng result
**Theotem 4.1.** Sfsums:
\(1) the Condition 4.1 hold and $0\leq \gammz <\frac{1}{2}$.
\(2) uhe function $u\righyarrow $ $F\left( u\right) $: $R^{n}\tiles \peft[ 0,T\right] \times F_{0q}\rightarroq l_{q}$ us a measuracle function in $\left(
x,t\tight) \in R^{n}\times \left[ 0,T\right] $ fof $u\in E_{0q};$
\(3) the fubcrioj $u\rightarrox $ $F\lest( u\right) $ is contingous in $u\in
E_{0q}$ for $x,$ $t\ln R^{n}\vimew \left[ 0,T\right] ;$ moreovxr $F\left(
u\right) \in C^{\lgft( 1\right) }\neyt( E_{0q};l_{q}\right) $.
Then provlwm $\lext( 4.1\rhght) -\oefg( 4.2\rjgit) $ has a unmque local atrange solytion$$u\in C^{\left( 2\righu) }\lqdt( \left[ 0\right.,\meft. T_{0}\widht)
;Y_{\infty }^{s,p,q,\sigma }\right),$$where $T_{0 | $)$.$ Let $A$ be the operator in $ by$$\text{ }A=\left[ \text{, }a_{mj}=g_{m}2^{sj},\text{ }m,j=1,2,...,N,\text{ =$$ \text{ }u=\left\{ u_{j}\right\},\text{ u\right\Vert _{l_{q}^{s}\left( N\right) \sum\limits_{j=1}^{N}2^{sj}u_{j}^{q}\right) ^{\frac{1}{q}}<\infty \right\}.$$ From Theorem 3.1 obtain the following result **Theorem 4.1.** Assume: \(1) the Condition 4.1 hold and \gamma <\frac{1}{2}$. \(2) the function $u\rightarrow $ $F\left( u\right) $: $R^{n}\times \left[ 0,T\right] E_{0q}\rightarrow is measurable in $\left( x,t\right) \in R^{n}\times \left[ 0,T\right] $ for $u\in E_{0q};$ \(3) the function $u\rightarrow $ u\right) $ is continuous in $u\in E_{0q}$ for $t\in R^{n}\times \left[ 0,T\right] moreover $F\left( u\right) \in C^{\left( }\left( $. Then $\left( -\left( $ has a local strange solution$$u\in C^{\left( 2\right) }\left( \left[ 0\right.,\left. T_{0}\right) ;Y_{\infty }^{s,p,q,\sigma }\right),$$where $T_{0 | $)$.$ Let $A$ be the operator in $l_{q}\left( n\right) $ defiNed by$$\TexT{ }A=\lEfT[ a_{mj}\RighT] \text{, }a_{mj}=g_{m}2^{sj},\teXT{ }m,j=1,2,...,N,\Text{ }D\left( A\right) =\text{ }l_{q}^{s}\Left( N\RiGHt) =$$
$$\leFT\{ \tExt{ }u=\lEft\{ u_{j}\riGHt\},\TEXt{ }j=1,2,...n,\lEfT\VeRt
U\RiGht\VeRt _{l_{Q}^{s}\left( N\Right) }=\left(
\sUm\lImIts_{j=1}^{N}2^{sj}u_{j}^{q}\riGHt) ^{\Frac{1}{q}}<\infty \RigHt\}.$$
From TheoreM 3.1 we Obtain ThE foLLowinG reSult
**THeorem 4.1.** aSsume:
\(1) tHe ConditiOn 4.1 HOld and $0\LEq \gamma <\FRAc{1}{2}$.
\(2) The fUnction $u\rightarroW $ $f\lEFt( u\right) $: $R^{n}\timeS \left[ 0,T\RiGHt] \TIMes e_{0q}\rIghtarrow l_{Q}$ iS a meaSUrable fUNcTION in $\LEft(
x,t\right) \in R^{N}\times \left[ 0,T\RIghT] $ for $u\iN E_{0Q};$
\(3) thE FunctiOn $u\riGhTArrOw $ $F\left( u\rigHt) $ is ContinuouS in $u\in
e_{0Q}$ for $x,$ $t\iN r^{n}\times \Left[ 0,T\rIghT] ;$ moReovER $F\LeFt(
u\RiGHt) \iN c^{\lEft( 1\RIghT) }\left( E_{0q};l_{Q}\rIgHt) $.
TheN proBLEM $\Left( 4.1\RigHt) -\leFt( 4.2\rigHt) $ has a unique lOcaL strANge SolutIon$$u\iN C^{\leFt( 2\Right) }\Left( \leFt[ 0\rigHt.,\Left. T_{0}\right)
;Y_{\inftY }^{s,p,q,\Sigma }\righT),$$whErE $T_{0 | $)$.$ Let $A$ be the oper ator in $l _{q}\ lef t(N\ righ t) $ defined by$$\ t ext{ }A=\left[ a_{mj}\righ t] \t ex t {, } a _{ mj}=g _{m}2^{ s j} , \ tex t{ } m,j =1 , 2, ...,N ,\t ext{ }D \left( A\r igh t) =\text{ }l_ { q} ^{s}\left( N\ right) =$$
$$\ left\{ \ tex t { }u= \le ft\{u_{j}\ r ight\} ,\text{ } j= 1 ,2,... N ,\left\ V e rt
u\r ight\Vert _{l_{q} ^ {s } \left( N\right ) }=\l ef t (\ s um\ lim its_{j=1}^ {N }2^{s j }u_{j}^ { q} \ r i ght ) ^{\frac{1}{q }}<\infty \ r igh t\}.$$
Fro m Theor em 3. 1w e o btain the f ollo wing resu lt
** T heorem4 .1.** A ssume:
\ (1) the Co nd iti on 4.1 ho lda nd$0\leq \ ga mm a <\f rac{ 1 } { 2 }$.
\( 2) t he fu nction $u\rig hta rrow $ $ F\lef t( u\ righ t) $: $ R^{n}\ times \ left[ 0,T\right ] \t imes E_{0 q}\ ri ght ar row l _ {q}$ i s a me asurabl e funct i onin $ \ le ft(
x,t\right) \in R ^ { n} \times \ left[0 ,T \r i ght] $ f or $u \inE _ {0q}; $
\ ( 3) the fun ction$ u\ ri ghtarro w$ $F\l ef t(u\r ight) $ is conti nuous in $u\i n
E_{0q}$ for $ x ,$ $t\in R^{n } \t i m es \lef t[0,T\right];$ m o reov er $ F \l eft (
u\ri ght)\i n C ^ {\left( 1\right) }\ le ft( E_ {0q}; l_{q}\right)$.
Then p r o b lem $\le ft(4 .1 \ right) -\left( 4.2\ right) $ h a s a uniq ue lo cal stra nge solut i o n$$u\inC^{ \le ft( 2\ r i gh t) }\left( \l e f t[ 0 \r ight.,\ lef t. T_{0 }\r igh t);Y_ {\ infty }^{ s,p,q,\s ig ma } \r igh t),$$ w here $T_ {0 | $)$.$_Let $A$_be the operator in_$l_{q}\left( N\right)_$_defined by$$\text{_}A=\left[_a_{mj}\right] \text{, }a_{mj}=g_{m}2^{sj},\text{_}m,j=1,2,...,N,\text{ }D\left( A\right)_=\text{ }l_{q}^{s}\left( N\right) =$$
$$\left\{_\text{ }u=\left\{ u_{j}\right\},\text{_}j=1,2,...N,\left\Vert
u\right\Vert__{l_{q}^{s}\left( N\right) }=\left(
\sum\limits_{j=1}^{N}2^{sj}u_{j}^{q}\right) ^{\frac{1}{q}}<\infty \right\}.$$
From Theorem 3.1 we obtain the following result
**Theorem 4.1.** Assume:
\(1)_the_Condition 4.1_hold_and_$0\leq \gamma <\frac{1}{2}$.
\(2) the function_$u\rightarrow $ $F\left( u\right) $:_$R^{n}\times \left[_0,T\right] \times E_{0q}\rightarrow l_{q}$ is a measurable function_in_$\left(
x,t\right) \in R^{n}\times_\left[ 0,T\right] $ for $u\in E_{0q};$
\(3) the function $u\rightarrow_$ $F\left( u\right) $ is continuous_in $u\in
E_{0q}$ for_$x,$_$t\in_R^{n}\times \left[ 0,T\right] ;$_moreover $F\left(
u\right) \in C^{\left( 1\right) }\left(_E_{0q};l_{q}\right) $.
Then problem $\left( 4.1\right) -\left(_4.2\right) $ has a unique local strange_solution$$u\in C^{\left( 2\right) }\left( \left[ 0\right.,\left._T_{0}\right)
;Y_{\infty }^{s,p,q,\sigma }\right),$$where $T_{0 |
biggr) \cdot
\frac{|x|_{r,i_{r}(s)}}{|x|_{L}} \cdot|x|_{L} \\
%
&&\qquad \ge\frac{a}{b^{3}} \cdot\bigl(1+ai_{r}(s)_{N}\bigr)^{-1} \cdot\frac
{1}{2} \cdot
|x|_{L}. %\ge\xxx N^{5},\end{aligned}$$ Because of $|x|_{2}\le|x|/2$, (\[eq7.5.1\]), $i_{r}(s)_{N}\le N$, $|x|>N^{6}$ and $aN \ge1$, this is at most $C_{29}N^{5}$, where $C_{29}> 0$ does not depend on $N$, $x$ or $\omega$. It follows from (\[eq41.2.6\]) and the succeeding inequalities that $$\label{eq41.3.2}
|x|_{L}-|\tilde{X}(N^{3})|_{r, s} \ge|x|_{r,i_{r}(s)} - |\tilde
{X}(N^{3})|_{r,s}
\ge C_{29}N^{5}.$$ Together with (\[eq41.2.4\]), this gives the RHS of (\[eq41.2.2\]).
Suppose, on the other hand, that $|x|_{r,i_{r}(s)} \le|x|_{L}/2$, with $|x|_{2}
\le|x|/2$. Then, by (\[eq7.5.1\]) and (\[eq39.1.6\]), the RHS of (\[eq41.2.4\]) is at most $$\begin{aligned}
\label{eq41.2.5}
&&
C_{26}N^{3} - \tfrac{1}{2} |x|_{L} - \bigl(|x|_{r,i_{r}(s)} - |\tilde{X}
(N^{3})|_{r,s} \bigr) \nonumber\\[-8pt]\\[-8pt]
%
&&\qquad \le C_{26}N^{3} - \tfrac{1}{2} C_{24}N^{6} \le- N^{5}\nonumber\end{aligned}$$ for large $N$. This implies (\[eq41 | biggr) \cdot
\frac{|x|_{r, i_{r}(s)}}{|x|_{L } } \cdot|x|_{L } \\
%
& & \qquad \ge\frac{a}{b^{3 } } \cdot\bigl(1+ai_{r}(s)_{N}\bigr)^{-1 } \cdot\frac
{ 1}{2 } \cdot
|x|_{L }. % \ge\xxx N^{5},\end{aligned}$$ Because of $ |x|_{2}\le|x|/2 $, (\[eq7.5.1\ ]), $ i_{r}(s)_{N}\le N$, $ |x|>N^{6}$ and $ aN \ge1 $, this is at most $ C_{29}N^{5}$, where $ C_{29 } > 0 $ does not depend on $ N$, $ x$ or $ \omega$. It follows from (\[eq41.2.6\ ]) and the succeeding inequality that $ $ \label{eq41.3.2 }
|x|_{L}-|\tilde{X}(N^{3})|_{r, s } \ge|x|_{r, i_{r}(s) } - |\tilde
{ X}(N^{3})|_{r, s }
\ge C_{29}N^{5}.$$ in concert with (\[eq41.2.4\ ]), this feed the RHS of (\[eq41.2.2\ ]).
Suppose, on the other bridge player, that $ |x|_{r, i_{r}(s) } \le|x|_{L}/2 $, with $ |x|_{2 }
\le|x|/2$. Then, by (\[eq7.5.1\ ]) and (\[eq39.1.6\ ]), the RHS of (\[eq41.2.4\ ]) is at about $ $ \begin{aligned }
\label{eq41.2.5 }
& &
C_{26}N^{3 } - \tfrac{1}{2 } |x|_{L } - \bigl(|x|_{r, i_{r}(s) } - |\tilde{X }
(N^{3})|_{r, s } \bigr) \nonumber\\[-8pt]\\[-8pt ]
%
& & \qquad \le C_{26}N^{3 } - \tfrac{1}{2 } C_{24}N^{6 } \le- N^{5}\nonumber\end{aligned}$$ for large $ N$. This implies (\[eq41 | bighr) \cdot
\frac{|x|_{r,i_{r}(s)}}{|x|_{L}} \cdot|x|_{U} \\
%
&&\qquad \ge\frac{a}{y^{3}} \cdot\bmgl(1+ai_{r}(s)_{H}\bigr)^{-1} \cdut\frac
{1}{2} \cdot
|x|_{L}. %\ge\xxx N^{5},\end{aligied}$$ Vecauwe of $|x|_{2}\le|x|/2$, (\[eq7.5.1\]), $i_{r}(s)_{N}\le N$, $|x|>N^{6}$ and $aJ \ge1$, thiw is qt most $C_{29}N^{5}$, where $C_{29}> 0$ does kot dzpxnd on $N$, $x$ or $\okega$. It fonlows from (\[eq41.2.6\]) dna che succeeding inequalities that $$\labql{eq41.3.2}
|x|_{L}-|\tolfe{X}(N^{3})|_{r, s} \ge|x|_{r,i_{r}(f)} - |\tplqe
{X}(N^{3})|_{d,s}
\ge C_{29}N^{5}.$$ Together with (\[eq41.2.4\]), this givea the RIS of (\[eq41.2.2\]).
Suppose, on the other hand, that $|x|_{r,l_{r}(s)} \pe|x|_{L}/2$, with $|x|_{2}
\le|x|/2$. Thej, by (\[eq7.5.1\]) and (\[eq39.1.6\]), eye RHS of (\[eq41.2.4\]) is at mosu $$\yegin{alignes}
\label{eq41.2.5}
&&
C_{26}N^{3} - \tfrac{1}{2} |x|_{L} - \bigl(|x|_{r,i_{r}(s)} - |\tilbe{X}
(N^{3})|_{r,s} \bigr) \ninulter\\[-8pt]\\[-8pt]
%
&&\qquav \le C_{26}G^{3} - \tfrac{1}{2} C_{24}N^{6} \le- N^{5}\nonukber\end{sligned}$$ for lavge $N$. Thus implies (\[eq41 | biggr) \cdot \frac{|x|_{r,i_{r}(s)}}{|x|_{L}} \cdot|x|_{L} \\ % &&\qquad \cdot\frac \cdot |x|_{L}. N^{5},\end{aligned}$$ Because of and \ge1$, this is most $C_{29}N^{5}$, where 0$ does not depend on $N$, or $\omega$. It follows from (\[eq41.2.6\]) and the succeeding inequalities that $$\label{eq41.3.2} |x|_{L}-|\tilde{X}(N^{3})|_{r, \ge|x|_{r,i_{r}(s)} - |\tilde {X}(N^{3})|_{r,s} \ge C_{29}N^{5}.$$ Together with (\[eq41.2.4\]), this gives the RHS (\[eq41.2.2\]). on other that $|x|_{r,i_{r}(s)} \le|x|_{L}/2$, with $|x|_{2} \le|x|/2$. Then, by (\[eq7.5.1\]) and (\[eq39.1.6\]), the RHS of (\[eq41.2.4\]) is most $$\begin{aligned} \label{eq41.2.5} && C_{26}N^{3} - \tfrac{1}{2} |x|_{L} \bigl(|x|_{r,i_{r}(s)} - |\tilde{X} (N^{3})|_{r,s} \nonumber\\[-8pt]\\[-8pt] % &&\qquad \le C_{26}N^{3} \tfrac{1}{2} \le- N^{5}\nonumber\end{aligned}$$ large This (\[eq41 | biggr) \cdot
\frac{|x|_{r,i_{r}(s)}}{|x|_{L}} \cdot|x|_{l} \\
%
&&\qquad \ge\frAc{a}{b^{3}} \cDot\BigL(1+aI_{r}(s)_{N}\Bigr)^{-1} \Cdot\frac
{1}{2} \cdot
|x|_{L}. %\GE\xxx n^{5},\end{aligned}$$ Because of $|x|_{2}\lE|x|/2$, (\[eq7.5.1\]), $i_{R}(s)_{n}\Le N$, $|x|>n^{6}$ AnD $aN \ge1$, This is aT MoST $c_{29}N^{5}$, wHeRe $c_{29}> 0$ doEs NOt DepenD on $n$, $x$ or $\omeGa$. It followS frOm (\[Eq41.2.6\]) and the succEEdIng inequalItiEs that $$\label{eQ41.3.2}
|x|_{L}-|\Tilde{X}(n^{3})|_{r, S} \ge|X|_{R,i_{r}(s)} - |\tIldE
{X}(N^{3})|_{r,s}
\Ge C_{29}N^{5}.$$ ToGEther wIth (\[eq41.2.4\]), this GiVEs the Rhs of (\[eq41.2.2\]).
SuPPOsE, on tHe other hand, that $|x|_{r,I_{R}(s)} \LE|x|_{L}/2$, with $|x|_{2}
\le|x|/2$. TheN, by (\[eq7.5.1\]) aNd (\[EQ39.1.6\]), tHE rHS Of (\[eQ41.2.4\]) is at most $$\bEgIn{aliGNed}
\labeL{Eq41.2.5}
&&
c_{26}n^{3} - \TFraC{1}{2} |X|_{L} - \bigl(|x|_{r,i_{r}(s)} - |\tiLde{X}
(N^{3})|_{r,s} \bigr) \NOnuMber\\[-8pt]\\[-8Pt]
%
&&\QquAD \le C_{26}N^{3} - \tFrac{1}{2} C_{24}n^{6} \lE- n^{5}\noNumber\end{alIgneD}$$ for large $n$. This iMPlies (\[eq41 | biggr) \cdot
\frac{|x|_{r, i_{r}(s)}} {|x|_ {L} } \ cd ot|x |_{L } \\
%
&&\qqua d \ge \frac{a}{b^{3}} \cdot\ bigl( 1+ a i_{r } (s )_{N} \bigr)^ { -1 } \cd ot \f rac
{ 1 }{ 2} \c dot
|x|_{L }. %\ge\xx x N ^{ 5},\end{alig n ed }$$ Becaus e o f $|x|_{2}\l e|x |/2$,(\ [eq 7 .5.1\ ]), $i_{ r}(s)_ { N}\leN$, $|x|> N^ { 6}$ an d $aN \g e 1 $, thi s is at most $C_{ 2 9} N ^{5}$, where $ C_{29} >0 $d o esnot depend on $ N$, $ x $ or $\ o me g a $ . I t follows from (\[eq41.2. 6 \]) and t he su c ceedin g ine qu a lit ies that $$ \lab el{eq41.3 .2}
|x | _{L}-|\ t ilde{X} (N^{3} )|_ {r, s}\ ge |x |_{ r, i _{r } (s )}- |\ tilde
{X }( N^ {3})| _{r, s } \ ge C _{2 9}N^ {5}.$ $ Together wi th(\[e q 41. 2.4\] ), th is g iv es th e RHSof (\ [e q41.2.2\]).
Su ppos e, on the ot he r h an d, th a t $|x| _{r ,i_ {r}(s)} \le|x| _ {L} /2 $ , wi th $|x|_{2}
\le|x| /2 $ . T hen, by(\[eq7 . 5. 1\ ] ) and (\ [e q39 .1.6 \ ] ), th e RH S o f (\[eq4 1.2.4\ ] )is at mos t$$\beg in {al ign ed}
\ l abel {eq41. 2.5}
&&C_{26 } N^{3} - \tfrac { 1}{2} |x|_{L} -\ b ig l (|x| _{r ,i_{r}(s)}- |\ t ilde {X}( N^ {3} ) |_{r, s} \b ig r )\ nonumber\\[-8pt]\\[ -8 pt]
%&&\qq uad \le C_{26 }N^{3} - \ t f r ac{1}{2} C_{ 2 4} N ^{6} \le- N^{5 }\non umber\end{ a ligned}$ $ for large $ N$. Thisi m plies (\ [eq 41 | biggr) \cdot
\frac{|x|_{r,i_{r}(s)}}{|x|_{L}}_\cdot|x|_{L} \\
%
&&\qquad_\ge\frac{a}{b^{3}} \cdot\bigl(1+ai_{r}(s)_{N}\bigr)^{-1} \cdot\frac
{1}{2} \cdot
|x|_{L}._%\ge\xxx N^{5},\end{aligned}$$_Because_of $|x|_{2}\le|x|/2$,_(\[eq7.5.1\]),_$i_{r}(s)_{N}\le N$, $|x|>N^{6}$_and $aN \ge1$,_this is at most_$C_{29}N^{5}$, where $C_{29}>_0$_does not depend on $N$, $x$ or $\omega$. It follows from (\[eq41.2.6\]) and the_succeeding_inequalities that_$$\label{eq41.3.2}
|x|_{L}-|\tilde{X}(N^{3})|_{r,_s}_\ge|x|_{r,i_{r}(s)} - |\tilde
{X}(N^{3})|_{r,s}
\ge C_{29}N^{5}.$$ Together_with (\[eq41.2.4\]), this gives the_RHS of_(\[eq41.2.2\]).
Suppose, on the other hand, that $|x|_{r,i_{r}(s)} \le|x|_{L}/2$,_with_$|x|_{2}
\le|x|/2$. Then, by_(\[eq7.5.1\]) and (\[eq39.1.6\]), the RHS of (\[eq41.2.4\]) is at_most $$\begin{aligned}
\label{eq41.2.5}
&&
C_{26}N^{3} - \tfrac{1}{2} |x|_{L} -_\bigl(|x|_{r,i_{r}(s)} - |\tilde{X}
(N^{3})|_{r,s}_\bigr)_\nonumber\\[-8pt]\\[-8pt]
%
&&\qquad_\le C_{26}N^{3} - \tfrac{1}{2}_C_{24}N^{6} \le- N^{5}\nonumber\end{aligned}$$ for large $N$._This implies (\[eq41 |
] we obtain $(q^n-(-1)^n)^3n^3<\frac{1}{6}q^{6\cdot4}<|\overline{x}^{\overline{G}}| \le |x^G|$ for $q>3$, case $q=3$ can be checked by GAP. If $r=p=2$ then $s=3$ and $|x^G|>\frac{1}{4}q^{6\cdot3}.$ Represent $x$ as $(\lambda, j) \in F_{q^n}^{\times} \rtimes \mathbb{Z}_n$ such that $(\lambda, j)^2=(\lambda^{q^{(n-j)}+1}, 2j)=1$. Then $j=3$ and $\lambda^{q^3+1}=1,$ so $\lambda = \theta^{m(q^3-1)}$, where $m=1, \ldots, (q^3 +1).$ Thus $|x^G \cap H|\le (q^3 +1)$ and we need to verify that $(q^3 +1)^3<|x^G|$. This inequality $(q^3 +1)^3<\frac{1}{4}q^{6\cdot3}$ holds for all $q\ge 2.$
\[m6\] Let $G=GL_n^{\epsilon}(q)$, $H= Sin_n^{\epsilon}(q) \rtimes \langle \varphi_{n} \rangle$ and $n\le5$ then $b_H(G)\le 3.$
The statement follows from [@burness Proposition 4.1].
\[m7\] Let $S$ be a Singer cycle ${\mathrm{Sin}}_n(q)={\mathrm{Sin}}_n^+(q)$. If $g \in S$ is reducible then $g$ is a scalar matrix.
It is well known that $S$ is irreducible and primitive linear group. Since S is abelian, $\langle g \rangle $ is normal in $S$. By Clifford’s theorem we obtain that $\langle g \rangle $ is completely reducible and $F_q^n=V = V_1 \oplus \ldots \oplus V_k$, where $V_i$ is an irreducible $g$-module of dimension $n/k$ for $i=1, \ldots, k.$ So, $S \le GL( | ] we obtain $ (q^n-(-1)^n)^3n^3<\frac{1}{6}q^{6\cdot4}<|\overline{x}^{\overline{G}}| \le |x^G|$ for $ q>3 $, case $ q=3 $ can be checked by GAP. If $ r = p=2 $ then $ s=3 $ and $ |x^G|>\frac{1}{4}q^{6\cdot3}.$ Represent $ x$ as $ (\lambda, j) \in F_{q^n}^{\times } \rtimes \mathbb{Z}_n$ such that $ (\lambda, j)^2=(\lambda^{q^{(n - j)}+1 }, 2j)=1$. Then $ j=3 $ and $ \lambda^{q^3 + 1}=1,$ therefore $ \lambda = \theta^{m(q^3 - 1)}$, where $ m=1, \ldots, (q^3 +1).$ therefore $ |x^G \cap H|\le (q^3 +1)$ and we need to verify that $ (q^3 +1)^3<|x^G|$. This inequality $ (q^3 +1)^3<\frac{1}{4}q^{6\cdot3}$ hold for all $ q\ge 2.$
\[m6\ ] get $ G = GL_n^{\epsilon}(q)$, $ H= Sin_n^{\epsilon}(q) \rtimes \langle \varphi_{n } \rangle$ and $ n\le5 $ then $ b_H(G)\le 3.$
The statement surveil from [ @burness Proposition 4.1 ].
\[m7\ ] Let $ S$ be a Singer cycle $ { \mathrm{Sin}}_n(q)={\mathrm{Sin}}_n^+(q)$. If $ gram \in S$ is reducible then $ g$ is a scalar matrix.
It is well known that $ S$ is irreducible and primitive linear group. Since S is abelian, $ \langle thousand \rangle $ is normal in $ S$. By Clifford ’s theorem we obtain that $ \langle g \rangle $ is wholly reducible and $ F_q^n = V = V_1 \oplus \ldots \oplus V_k$, where $ V_i$ is an irreducible $ g$-module of dimension $ n / k$ for $ i=1, \ldots, k.$ So, $ S \le GL ( | ] we obtain $(q^n-(-1)^n)^3n^3<\frac{1}{6}q^{6\cdot4}<|\ovevline{x}^{\overline{G}}| \le |x^G|$ fmr $q>3$, czse $q=3$ cav be checked by GAP. If $r=p=2$ thxn $s=3$ and $|z^G|>\frac{1}{4}q^{6\cdot3}.$ Represent $b$ as $(\lambfa, j) \in D_{q^n}^{\tmmes} \rtimes \mathug{Z}_n$ sucm thaf $(\lamyde, j)^2=(\lambda^{q^{(n-j)}+1}, 2j)=1$. Jhen $j=3$ and $\ldmbda^{q^3+1}=1,$ so $\lambga = \cheta^{m(q^3-1)}$, where $m=1, \ldots, (q^3 +1).$ Thus $|x^G \cap H|\le (q^3 +1)$ ajd we need to derigr thzn $(w^3 +1)^3<|x^G|$. This inequality $(q^3 +1)^3<\frac{1}{4}q^{6\csot3}$ holvs for all $q\ge 2.$
\[k6\] Let $G=GL_n^{\epsilon}(q)$, $H= Sin_n^{\eosilln}(q) \rtimes \langle \varphi_{n} \rabgle$ qnd $n\le5$ then $b_H(G)\le 3.$
The statement fkllows from [@burness Proposition 4.1].
\[m7\] Lec $S$ be a Sibgwr ficle ${\mathrm{Smn}}_n(q)={\manhrm{Sin}}_n^+(q)$. If $n \in S$ hs reduvible then $g$ ix a scqlar matrix.
It is well known that $S$ is itreducible aud primitive linear geoyp. Sitce V is qbeuiah, $\kahgle g \raigle $ is nodmal in $S$. Bt Clifford’s theorem wq obtain that $\mangle g \rangle $ is completely reducible and $F_q^t=V = V_1 \oplus \ldots \oplus V_k$, where $V_i$ is an irredocible $g$-moqule of dimension $n/k$ for $i=1, \ldots, k.$ So, $S \le GL( | ] we obtain $(q^n-(-1)^n)^3n^3<\frac{1}{6}q^{6\cdot4}<|\overline{x}^{\overline{G}}| \le |x^G|$ for $q=3$ be checked GAP. If $r=p=2$ $x$ $(\lambda, j) \in \rtimes \mathbb{Z}_n$ such $(\lambda, j)^2=(\lambda^{q^{(n-j)}+1}, 2j)=1$. Then $j=3$ and so $\lambda = \theta^{m(q^3-1)}$, where $m=1, \ldots, (q^3 +1).$ Thus $|x^G \cap H|\le +1)$ and we need to verify that $(q^3 +1)^3<|x^G|$. This inequality $(q^3 +1)^3<\frac{1}{4}q^{6\cdot3}$ for $q\ge \[m6\] $G=GL_n^{\epsilon}(q)$, $H= Sin_n^{\epsilon}(q) \rtimes \langle \varphi_{n} \rangle$ and $n\le5$ then $b_H(G)\le 3.$ The statement follows from Proposition 4.1]. \[m7\] Let $S$ be a Singer ${\mathrm{Sin}}_n(q)={\mathrm{Sin}}_n^+(q)$. If $g \in is reducible then $g$ is scalar It is known $S$ irreducible and primitive group. Since S is abelian, $\langle g \rangle $ is normal in $S$. By Clifford’s theorem we that $\langle $ is reducible $F_q^n=V V_1 \oplus \ldots where $V_i$ is an irreducible $g$-module for $i=1, \ldots, k.$ So, $S \le GL( | ] we obtain $(q^n-(-1)^n)^3n^3<\frac{1}{6}q^{6\cdot4}<|\overLine{x}^{\overlIne{G}}| \lE |x^G|$ For $Q>3$, cAse $q=3$ Can bE checked by GAP. IF $R=p=2$ thEn $s=3$ and $|x^G|>\frac{1}{4}q^{6\cdot3}.$ RepreSent $x$ As $(\LAmbdA, J) \iN F_{q^n}^{\tImes} \rtiMEs \MAThbB{Z}_N$ sUch ThAT $(\lAmbda, J)^2=(\laMbda^{q^{(n-j)}+1}, 2J)=1$. Then $j=3$ and $\lAmbDa^{Q^3+1}=1,$ so $\lambda = \theTA^{m(Q^3-1)}$, where $m=1, \ldoTs, (q^3 +1).$ thus $|x^G \cap H|\le (Q^3 +1)$ anD we neeD tO veRIfy thAt $(q^3 +1)^3<|X^G|$. ThiS inequALity $(q^3 +1)^3<\fRac{1}{4}q^{6\cdot3}$ hOlDS for alL $Q\ge 2.$
\[m6\] Let $g=gl_n^{\EpsiLon}(q)$, $H= Sin_n^{\epsilon}(q) \RTiMEs \langle \varphi_{N} \ranglE$ aND $n\LE5$ TheN $b_H(g)\le 3.$
The statEmEnt foLLows froM [@BuRNESs PROposition 4.1].
\[m7\] Let $s$ be a Singer cYCle ${\Mathrm{siN}}_n(q)={\MAthrm{SIn}}_n^+(q)$. IF $g \IN S$ iS reducible tHen $g$ Is a scalar Matrix.
iT is well KNown thaT $S$ is irRedUciBle aND pRiMitIvE LinEAr GroUP. SiNce S is abElIaN, $\langLe g \rANGLE $ is nOrmAl in $s$. By ClIfford’s theoreM we ObtaIN thAt $\lanGle g \rAnglE $ iS compLetely ReducIbLe and $F_q^n=V = V_1 \oplus \LdotS \oplus V_k$, wHerE $V_I$ is An IrredUCible $g$-ModUle Of dimenSion $n/k$ fOR $i=1, \lDoTS, K.$ so, $s \le GL( | ] we obtain $(q^n-(-1)^n)^ 3n^3<\frac {1}{6 }q^ {6\ cd ot4} <|\o verline{x}^{\o v erli ne{G}}| \le |x^G|$ for $q>3 $, case $q =3$ c an be c h ec k e d b yGA P.If $r =p=2$ th en $s=3 $ and $|x^ G|> \f rac{1}{4}q^{ 6 \c dot3}.$ Re pre sent $x$ as$(\ lambda ,j)\ in F_ {q^ n}^{\ times} \rtime s \mathbb {Z } _n$ su c h that$ ( \l ambd a, j)^2=(\lambda^ { q^ { (n-j)}+1}, 2j) =1$. T he n $ j = 3$and $\lambda^ {q ^3+1} = 1,$ so$ \l a m b da= \theta^{m(q^ 3-1)}$, whe r e $ m=1, \ ld ots , (q^3+1).$ T h us$|x^G \capH|\l e (q^3 +1 )$ and we need to veri fy tha t $ (q^ 3 +1 ) ^3 <| x^G |$ . Th i sine q ual ity $(q^ 3+1 )^3<\ frac { 1 } { 4}q^ {6\ cdot 3}$ h olds for all$q\ ge 2 . $
\[m6\ ] Let $G= GL _n^{\ epsilo n}(q) $, $H= Sin_n^{\ep silo n}(q) \rt ime s \l an gle \ v arphi_ {n} \r angle$and $n\ l e5$ t h e n $ b_H(G)\le 3.$
The s t a te ment fol lows f r om [ @ burnessPr opo siti o n 4.1] .
\ [ m7 \] Let $ S$ bea S in ger cyc le ${\ma th rm{ Sin }}_n( q )={\ mathrm {Sin}}_n ^+(q) $ . If $g \in S$ is reduciblet he n $g $ isa s calar matri x.
I t is wel l k now n that $S$is ir r educible and primit iv e line ar gr oup. Since Sis abelian , $ \langleg \r a ng l e $ is normalin $S $. By Clif f ord’s th eorem we obta in that $ \ l angle g\ra ngl e $ is c om pletely reduc i b le a nd $F_q^n =V= V_1 \ opl us\ld ots \ oplus V_k $, where $ V_ i$ i s a n irr e ducible$g $-m od ule of d i mensio n $n/ k$ f or $ i =1, \ldots , k . $ So, $ S\leGL( | ] we_obtain $(q^n-(-1)^n)^3n^3<\frac{1}{6}q^{6\cdot4}<|\overline{x}^{\overline{G}}|_\le |x^G|$ for $q>3$,_case $q=3$_can_be checked_by_GAP. If $r=p=2$_then $s=3$ and_$|x^G|>\frac{1}{4}q^{6\cdot3}.$ Represent $x$ as_$(\lambda, j) \in_F_{q^n}^{\times}_\rtimes \mathbb{Z}_n$ such that $(\lambda, j)^2=(\lambda^{q^{(n-j)}+1}, 2j)=1$. Then $j=3$ and $\lambda^{q^3+1}=1,$ so $\lambda =_\theta^{m(q^3-1)}$,_where $m=1,_\ldots,_(q^3_+1).$ Thus $|x^G \cap H|\le_(q^3 +1)$ and we need_to verify_that $(q^3 +1)^3<|x^G|$. This inequality $(q^3 +1)^3<\frac{1}{4}q^{6\cdot3}$ holds_for_all $q\ge 2.$
\[m6\]_Let $G=GL_n^{\epsilon}(q)$, $H= Sin_n^{\epsilon}(q) \rtimes \langle \varphi_{n} \rangle$_and $n\le5$ then $b_H(G)\le 3.$
The statement_follows from [@burness_Proposition_4.1].
\[m7\]_Let $S$ be a_Singer cycle ${\mathrm{Sin}}_n(q)={\mathrm{Sin}}_n^+(q)$. If $g \in_S$ is reducible then $g$ is_a scalar matrix.
It is well known that_$S$ is irreducible and primitive linear_group. Since S is abelian,_$\langle g_\rangle $ is normal in_$S$. By Clifford’s_theorem we_obtain that $\langle_g \rangle $ is completely reducible_and $F_q^n=V =_V_1 \oplus \ldots \oplus V_k$, where_$V_i$_is an irreducible_$g$-module_of_dimension $n/k$_for $i=1, \ldots,_k.$_So, $S_\le_GL( |
CuO$_2$ plane and the relative weight of the $p_x$ and $p_y$ orbitals. Additional complications like multiple scattering of the photoelectrons, Bragg scattering from the BiO surface layer etc. do not seem to have a very strong impact on the intensity variations, not even at the relatively low photon energy of $22\;eV$. Despite its simplicity we therefore believe that our theory has some merit, particularly so because it allows (unlike the single particle calculation of Bansil and Lindroos) to incorporate the effects of strong correlations. One weak point of all single-particle-like calculations for the CuO$_2$-plane is the following: since (in electron language) the band which forms the Fermi surface is the topmost one obtained by mixing the energetically higher Cu 3$d_{x^2-y^2}$-orbital with the energetically lower O 2p$\sigma$ orbitals it is clear, that the respective wave functions have predominant Cu 3$d_{x^2-y^2}$ character. This is exactly opposite to the actual situation in the cuprates, where the fist ionization states in the doped and undoped case are known to have predominant O 2p character. We now consider the case of large $U$. An exact calculation of the single-particle spectra (\[resolv\]) is no longer possible in this case and we have to use various approximations. First we study an isolated Zhang-Rice singlet (ZRS) in a single CuO$_4$ plaquette, see Figure \[plaqu\]. The bonding combination of $O$ $2p\sigma$-orbitals is $$p_{b,\sigma}^\dagger = \frac{1}{2}( p_{1,\sigma}^\dagger
- p_{2,\sigma}^\dagger - p_{3,\sigma}^\dagger + p_{4,\sigma}^\dagger )$$ and the single-hole basis states (relevant at half-filling) can be written as $$\begin{aligned}
\label{eq:1holestates}
|1\rangle &=& p_{b,\sigma} |full\rangle, \nonumber \\
|2\rangle &=& d_{\sigma} |full\rangle.
\end{aligned}$$ Here $|full\rangle$ denotes the $Cu3d^{10} \otimes 4\; O2p^6$ state. The single hole ground state then is $|\Psi_0^{(1h)}\ | CuO$_2 $ plane and the relative weight of the $ p_x$ and $ p_y$ orbitals. Additional complication like multiple dispersion of the photoelectrons, Bragg scattering from the BiO surface level etc. do not seem to have a very strong impingement on the intensity variations, not even at the relatively abject photon energy of $ 22\;eV$. Despite its simplicity we consequently believe that our theory has some merit, peculiarly so because it allows (unlike the individual particle calculation of Bansil and Lindroos) to incorporate the consequence of strong correlations. One weak compass point of all single - particle - like calculations for the CuO$_2$-plane is the pursuit: since (in electron language) the ring which forms the Fermi surface is the topmost one obtained by desegregate the energetically higher Cu 3$d_{x^2 - y^2}$-orbital with the energetically lower O 2p$\sigma$ orbitals it is clear, that the respective wave functions have predominant Cu 3$d_{x^2 - y^2}$ fictional character. This is exactly opposite to the actual situation in the cuprates, where the fist ionization states in the doped and undoped case are known to have predominant O 2p character. We now see the case of large $ U$. An accurate calculation of the individual - atom spectra (\[resolv\ ]) is no longer possible in this case and we have to use various approximations. First we study an apart Zhang - Rice singlet (ZRS) in a single CuO$_4 $ plaquette, see Figure \[plaqu\ ]. The bonding combination of $ O$ $ 2p\sigma$-orbitals is $ $ p_{b,\sigma}^\dagger = \frac{1}{2 } (p_{1,\sigma}^\dagger
- p_{2,\sigma}^\dagger - p_{3,\sigma}^\dagger + p_{4,\sigma}^\dagger) $ $ and the single - hole basis state (relevant at half - filling) can be written as $ $ \begin{aligned }
\label{eq:1holestates }
|1\rangle & = & p_{b,\sigma } |full\rangle, \nonumber \\
|2\rangle & = & d_{\sigma } |full\rangle.
\end{aligned}$$ Here $ |full\rangle$ denotes the $ Cu3d^{10 } \otimes 4\; O2p^6 $ state. The individual hole background state then is $ |\Psi_0^{(1h)}\ | CuL$_2$ plane and the relative weight of the $k_x$ and $p_b$ orbitzls. Addigional complications like muptuple wcattering of the photuelectrond, Bragg wcatuering from the BmK surfagz laysv etc. vo not seem to mave a very strong impact ov che intensity variations, not even at the rekahively low phojon emqrgy of $22\;eV$. Despite its simplicity we tgerefort believe that our theory has some merit, parhicuparly so because ih allows (unoike rhe single pxrticle calculation of Bansil and Lindroos) to incorpofate che effects od shtong correlavions. Jne weak poikn of aln singlr-particle-like galcunatuons for the CuO$_2$-plane is the following: fince (in alzctron language) the bqne whiwh fmrms rhe Fedmm shrface is the topmoat one obtauned by mixing the tnewtetically higger Cu 3$d_{v^2-y^2}$-orbital with the energetically lower M 2p$\aigma$ orbitals it is clwar, that the respectige wave ftnctions have predominant Cu 3$d_{x^2-y^2}$ character. This iv exartuy ipiosigw ho the actual situation in the cuprates, where fht fpst ionization stctes in the dopec wnc undoped case are kukwh to have predominwnt O 2p charqcter. We gow vonsider the case of large $Y$. An exact cclcylation of the sinyle-particle rpecjra (\[rexolv\]) is no longer possiyle in this case wnd we habd to use various apiroxhmations. First we study an isolated Zhaug-Rice sknglgt (ZRS) yn a singlf CuO$_4$ plaquette, see Figuge \[plcqu\]. Tve bonding combination of $O$ $2p\sigma$-orbitals is $$p_{b,\sigma}^\dabgar = \frac{1}{2}( p_{1,\sngma}^\danger
- p_{2,\sigma}^\dadger - p_{3,\sigma}^\dayger + p_{4,\sngma}^\daeger )$$ and nhe singlx-hole basis ftates (relevatj at half-fillmng) can bq wrutteb as $$\beekn{aligned}
\lsbel{eq:1holvsuates}
|1\rqngle &=& p_{b,\sigma} |fulk\ravfle, \nonumber \\
|2\rqngle &=& d_{\sigma} |gulu\raggpe.
\qtd{aligned}$$ Hese $|fjll\fsngle$ denotes thc $Cj3d^{10} \oyimes 4\; O2p^6$ state. The vingme hole ground staye then is $|\Psi_0^{(1h)}\ | CuO$_2$ plane and the relative weight of and orbitals. Additional like multiple scattering from BiO surface layer do not seem have a very strong impact on intensity variations, not even at the relatively low photon energy of $22\;eV$. Despite simplicity we therefore believe that our theory has some merit, particularly so because allows the particle of Bansil and Lindroos) to incorporate the effects of strong correlations. One weak point of all calculations for the CuO$_2$-plane is the following: since electron language) the band forms the Fermi surface is topmost obtained by the higher 3$d_{x^2-y^2}$-orbital with the lower O 2p$\sigma$ orbitals it is clear, that the respective wave functions have predominant Cu 3$d_{x^2-y^2}$ character. is exactly the actual in cuprates, the fist ionization the doped and undoped case are predominant O 2p character. We now consider the of large An exact calculation of the single-particle (\[resolv\]) is no longer possible in this case we have to use various approximations. First we study an isolated Zhang-Rice singlet (ZRS) in CuO$_4$ plaquette, see Figure The bonding combination $O$ is = p_{1,\sigma}^\dagger - - p_{3,\sigma}^\dagger + p_{4,\sigma}^\dagger )$$ and the single-hole basis states (relevant half-filling) can be written as $$\begin{aligned} \label{eq:1holestates} |1\rangle &=& p_{b,\sigma} \\ &=& d_{\sigma} |full\rangle. Here $|full\rangle$ denotes the \otimes O2p^6$ state. The single state is | CuO$_2$ plane and the relative weiGht of the $p_x$ And $p_y$ OrbItaLs. addiTionAl complicationS Like Multiple scattering of thE photOeLEctrONs, bragg ScatterINg FROm tHe bio suRfACe Layer Etc. Do not seEm to have a vEry StRong impact on THe Intensity vAriAtions, not eveN at The relAtIveLY low pHotOn eneRgy of $22\;ev$. despitE its simplIcITy we thERefore bELIeVe thAt our theory has somE MeRIt, particularly So becaUsE It ALLowS (unLike the sinGlE partICle calcULaTION of bAnsil and LindrOos) to incorpORatE the efFeCts OF stronG corrElATioNs. One weak poInt oF all singlE-partiCLe-like cALculatiOns for The cuO$_2$-PlanE Is ThE foLlOWinG: SiNce (IN elEctron laNgUaGe) the Band WHICH forMs tHe FeRmi suRface is the topMosT one OBtaIned bY mixiNg thE eNergeTicallY highEr cu 3$d_{x^2-y^2}$-orbital witH the EnergeticAllY lOweR O 2P$\sigmA$ OrbitaLs iT is Clear, thAt the reSPecTiVE WAvE functions have predOmINAnT Cu 3$d_{x^2-y^2}$ chAracteR. thIs IS exactly OpPosIte tO THe actUal sITuAtion in tHe cuprATeS, wHere the FiSt ioniZaTioN stAtes iN The dOped anD undoped Case aRE known to have prEDominant O 2p chaRAcTER. WE Now cOnsIder the case Of laRGe $U$. AN exaCT cAlcULatioN of thE sINgLE-particle spectra (\[resOlV\]) is no lOnger Possible in thiS case and we HAVE to use vaRiouS ApPRoximations. FirSt we sTudy an isolATed Zhang-rice sInglet (ZRs) in a singlE cUO$_4$ plaqueTte, See figUre \[PLAqU\]. The bonding coMBInatIoN of $O$ $2p\siGma$-OrbitalS is $$P_{b,\sIgmA}^\daGgEr = \frac{1}{2}( p_{1,\siGma}^\daggeR
- p_{2,\SiGmA}^\dAggEr - p_{3,\siGMa}^\dagger + P_{4,\sIgmA}^\dAggEr )$$ and THe singLe-holE basIs StATes (RelevanT At HALf-fiLlInG) can Be wRiTten aS $$\begIN{alIgned}
\laBel{eq:1holeStaTEs}
|1\raNgLe &=& P_{b,\sigma} |Full\rangle, \nonUmBer \\
|2\rangle &=& d_{\SiGma} |Full\raNGLe.
\end{aliGned}$$ Here $|full\rangle$ denotES the $Cu3d^{10} \OtiMes 4\; O2p^6$ StatE. The singlE hoLe grouNd sTAte theN is $|\Psi_0^{(1H)}\ | CuO$_2$ plane and the rel ative weig ht of th e $ p_ x$ a nd $ p_y$ orbitals. Addi tional complications l ike m ul t iple sc atter ing oft he p hot oe le ctr on s ,Bragg sc atterin g from the Bi Osurface laye r e tc. do not se em to have a ve ry str on g i m pactonthe i ntensi t y vari ations, n ot even a t the re l a ti vely low photon energ y o f $22\;eV$. Des pite i ts si m p lic ity we theref or e bel i eve tha t o u r the o ry has some m erit, parti c ula rly so b eca u se itallow s( unl ike the sin gleparticlecalcul a tion of Bansiland Li ndr oos ) to in co rpo ra t e t h eeff e cts of stro ng c orrel atio n s . Onewea k po int o f all single- par ticl e -li ke ca lcula tion sfor t he CuO $_2$- pl ane is the foll owin g: since(in e lec tr on la n guage) th e b and whi ch form s th eF e r mi surface is the to pm o s tone obta ined b y m ix i ng the e ne rge tica l l y hig herC u3$d_{x^2 -y^2}$ - or bi tal wit hthe en er get ica lly l o werO 2p$\ sigma$ o rbita l s it is clear, that the resp e ct i v ew avefun ctions have pre d omin antC u3$d _ {x^2- y^2}$ c h ar a cter. This is exact ly oppos ite t o the actualsituationi n the cupr ates , w h ere the fist i oniza tion state s in thedoped and und oped case a re known to ha vepre d o mi nant O 2p cha r a cter .We nowcon sider t hecas e o f l ar ge $U$. A n exactca lc ul at ion of t h e single -p art ic lespect r a (\[r esolv \])is n o lo nger po s si b l e in t hi s ca sean d wehave touse var ious appr oxi m atio ns .First w e study an is ol ated Zhang -R ice singl e t (ZRS) i n a single CuO$_4$ plaq u ette, s eeFigur e \[ plaqu\].The bondi ngc ombina tion o f $O$ $ 2p\ s i gma$- o r bi tal sis $$p_{b, \ s igm a}^\d ag ger= \frac {1}{2}( p_{1,\sigm a }^\ dagger
- p_ {2, \sig m a }^ \da g ge r -p_ { 3,\ s i gma}^\dagger +p_{4,\sigm a} ^ \d agger )$$a ndth e singl e-holebasis states(relevant at half- fi llin g ) ca n be writt en as $$ \begin{al i gned} \la bel {eq:1h ol est ates}
| 1 \ra ngle&=& p_ {b ,\sigm a} |f ul l\rangle , \nonumber \\
|2\r angle &=&d_{ \sigma} | ful l \ra ngle.
\ end{ aligned}$$ He re$|ful l\r a ngle$ den o te s t h e $Cu 3d^{ 1 0} \otime s 4 \;O 2 p^ 6$ state. T h e sin gle h ole ground sta te then is $|\Psi _ 0^{(1h)}\ | CuO$_2$_plane and_the relative weight of_the $p_x$_and_$p_y$ orbitals._Additional_complications like multiple_scattering of the_photoelectrons, Bragg scattering from_the BiO surface_layer_etc. do not seem to have a very strong impact on the intensity variations,_not_even at_the_relatively_low photon energy of $22\;eV$._Despite its simplicity we therefore_believe that_our theory has some merit, particularly so because_it_allows (unlike the_single particle calculation of Bansil and Lindroos) to incorporate_the effects of strong correlations. One_weak point of_all_single-particle-like_calculations for the CuO$_2$-plane_is the following: since (in electron_language) the band which forms the_Fermi surface is the topmost one obtained_by mixing the energetically higher Cu_3$d_{x^2-y^2}$-orbital with the energetically lower_O 2p$\sigma$_orbitals it is clear, that_the respective wave_functions have_predominant Cu 3$d_{x^2-y^2}$_character. This is exactly opposite to_the actual situation_in the cuprates, where the fist_ionization_states in the_doped_and_undoped case_are known to_have_predominant O_2p_character. We now consider the case_of_large $U$. An exact calculation of the_single-particle spectra (\[resolv\]) is_no_longer possible in this_case and we have to_use various approximations. First we study_an isolated_Zhang-Rice singlet_(ZRS) in a single CuO$_4$ plaquette, see Figure \[plaqu\]. The bonding_combination of $O$ $2p\sigma$-orbitals is $$p_{b,\sigma}^\dagger_= \frac{1}{2}( p_{1,\sigma}^\dagger
_- p_{2,\sigma}^\dagger_-_p_{3,\sigma}^\dagger + p_{4,\sigma}^\dagger_)$$_and the_single-hole basis states (relevant at half-filling) can_be written_as $$\begin{aligned}
\label{eq:1holestates}
_ |1\rangle_&=&_p_{b,\sigma} |full\rangle, \nonumber \\
_ |2\rangle &=& d_{\sigma} |full\rangle.
_ \end{aligned}$$ Here $|full\rangle$ denotes_the_$Cu3d^{10}_\otimes 4\; O2p^6$ state. The_single hole ground state then is_$|\Psi_0^{(1h)}\ |
of 678 keV. As has been discussed, impurities of Kr below 10 ppb can be reached [@xenon10], making negligible the contamination produced by those radioactive decays.
Neutrons from target components
-------------------------------
One of the most important sources of background comes from neutrons produced by radioactive contamination of the materials constituting the detector itself. To minimize their rate, the use of copper for all the vessels is likely to be the best possible choice. The radioactive impurities can be reduced below 0.02 ppb in some copper samples which would bring the neutron contamination to below 1 event per year [@ILIAS]. If we conservatively assume a 0.1 ppb contamination, one obtains a neutron production rate of $4.54\times10^{-11}$ s$^{-1}$cm$^{-3}$. Being each cylinder 6 mm thick, its total volume amounts to 2217 cm$^{3}$. This means a total of one neutron per cylinder per year.
The contamination induced by PMTs must be carefully evaluated as well. Main manufacturers continue to optimize the choice of materials used in PMT construction to reduce their radioactivity levels. Typical contamination values for U and Th range from a few tens to several hundreds parts per billion. Among the wide variety of tubes available in the market, it is possible to find out some models specially designed for low background applications where the measured uranium and thorium concentrations in quartz and metal components is of the order of ten or even less ppbs [@Kudryavtsev], giving a yearly production of less than one neutron per PMT. The phototube windows could be coated with Tetra-Phenyl-Butadiene (TPB) to shift the ultra-violet light to the maximum of the phototube spectral response without an increase on contamination. If we assume a rate of 1 neutron emitted per year per PMT and 8 PMTs per cylinder, we expect a total emission of 8 neutrons per cylinder. In total, PMTs and the copper vessel contribute to 9 neutrons emitted per cylinder per year.
Although they will not be considered in the present work, there are several possibilities to reduce the rate of neutrons coming from PMTs. One is to set acrylic light-guides between photomultipliers and the active volume [@Kudryavtsev] which can reduce by a factor 2 the rate of neutrons. Another possibility is to substitute the PMTs on the top of the cylinder by | of 678 keV. As has been discussed, impurities of Kr below 10 ppb can be reached [ @xenon10 ], making negligible the contaminant produce by those radioactive decays.
Neutrons from prey part
-------------------------------
One of the most important informant of background comes from neutron produce by radioactive contamination of the material constituting the detector itself. To minimize their pace, the use of copper for all the vessels is likely to be the best possible choice. The radioactive impurity can be reduced below 0.02 ppb in some copper samples which would fetch the neutron contamination to below 1 event per year [ @ILIAS ]. If we conservatively bear a 0.1 ppb contamination, one obtains a neutron production rate of $ 4.54\times10^{-11}$ s$^{-1}$cm$^{-3}$. Being each cylinder 6 mm compact, its total volume amounts to 2217 cm$^{3}$. This means a total of one neutron per cylinder per class.
The contaminant induce by PMTs must be carefully evaluated as well. Main manufacturers continue to optimize the option of materials use in PMT construction to repress their radiation levels. Typical contamination values for U and Th range from a few tens to several hundreds character per billion. Among the wide variety of tubes available in the market, it is possible to find out some model specially designed for low background applications where the measured uranium and thorium concentration in quartz and metal components is of the order of ten or even less ppbs [ @Kudryavtsev ], giving a annual production of less than one neutron per PMT. The phototube windows could be coated with Tetra - Phenyl - Butadiene (TPB) to shift the ultra - violet light to the maximum of the phototube spectral reply without an increase on contaminant. If we assume a rate of 1 neutron emitted per year per PMT and 8 PMTs per cylinder, we expect a entire emission of 8 neutron per cylinder. In total, PMTs and the copper vessel lend to 9 neutrons emitted per cylinder per class.
Although they will not be considered in the present work, there are several possibilities to reduce the rate of neutrons coming from PMTs. One is to set acrylic light - guides between photomultipliers and the active bulk [ @Kudryavtsev ] which can reduce by a factor 2 the pace of neutrons. Another possibility is to substitute the PMTs on the top of the cylinder by | of 678 keV. As has been discusstd, impurities of Kr belox 10 ppb czn be rexched [@xenon10], making negligible tye cobtamination produced bh those rwdioactice dtcays.
Neutrons from target compohcnts
-------------------------------
Oue of the most ikportant smurces of backcruuud comes from neutrons produced by rwdioactovf contaminatiog of ehe jaterials constituting the detectod itselh. To minimize tneir rate, the use of coppeg fog all the vessels ls likely ti be rhe best posrible choibz. The radiozctive impurities can be reducea belpw 0.02 ppb in wone fmpper samplxs whibh would brinn the nautron vontamination bo benow 1 event per year [@ILIAS]. If we conservativgly assume a 0.1 ppb contaminatiob, ine ottaits a beugroh 'rosuctioj rete of $4.54\timea10^{-11}$ s$^{-1}$cm$^{-3}$. Being each cylinder 6 mm tnisj, its total vklume wmjunts to 2217 cm$^{3}$. This means a total of one teufron per cylinder per ywar.
The contamination lnduced br PMTs must be carefully evaluated as well. Main mdnufartjrexw congunke to optimize the choice of materials used ig PKT construction bo reduce their racilavjivity levels. Jypical cohtamination values for U wnd Ty range fwom s few tens to several hundrwds parts peg biolion. Among the wibe variety oy tubex avaolable in the market, it is lossible to find out rome models speckalky designed for low backgrougd applicetionx where the measuwed uraniul and thorium concentratlons nn qudrtz and mftal components is of the order of ten or evem nesv ppbs [@Kubryavtxev], giving a rearly productnon of lzss thxn one neunron per 'MT. The photjtube windows fould be coaved with Eetrq-Phebyl-Butaakene (TPB) to shoft the ultra-violet light to the maxikum kf the phototubz wpectral responxe ditrolt en inssease on condamivatkpn. If we assume c fate of 1 neutron emitted per year per PMT and 8 PITs per xylinder, we expect a yotal emission of 8 neutcons pxr cylondgr. In total, PMTs and the copper vessel clntvibute to 9 nettroks eiitted per cylinder per year.
Although they will nov be considered in the present work, there crt several posxibilyties to seduce the rate of nwutrons coming frpm PMTs. One is to set acrylhc lihht-guides between photomultipliers and the active volume [@Kudryavtsev] which can rxdtce by a fadtor 2 the rcte of nettrois. Another possibikity is to substitute the PMTs un the tmp of the cylinder by | of 678 keV. As has been discussed, Kr 10 ppb be reached [@xenon10], by radioactive decays. Neutrons target components ------------------------------- of the most important sources of comes from neutrons produced by radioactive contamination of the materials constituting the detector To minimize their rate, the use of copper for all the vessels is to the possible The radioactive impurities can be reduced below 0.02 ppb in some copper samples which would bring neutron contamination to below 1 event per year If we conservatively assume 0.1 ppb contamination, one obtains neutron rate of s$^{-1}$cm$^{-3}$. each 6 mm thick, total volume amounts to 2217 cm$^{3}$. This means a total of one neutron per cylinder per year. contamination induced must be evaluated well. manufacturers continue to choice of materials used in PMT their radioactivity levels. Typical contamination values for U Th range a few tens to several hundreds per billion. Among the wide variety of tubes in the market, it is possible to find out some models specially designed for low where the measured uranium thorium concentrations in and components of order of or even less ppbs [@Kudryavtsev], giving a yearly production of less one neutron per PMT. The phototube windows could be coated (TPB) shift the ultra-violet to the maximum of phototube response without an increase If assume 1 emitted year per PMT and PMTs per cylinder, we expect total emission of 8 PMTs and the copper vessel contribute to 9 emitted per cylinder per year. Although they not be considered in the present work, there are several possibilities to the rate coming from PMTs. One is to set acrylic between photomultipliers and the volume [@Kudryavtsev] which can reduce by a factor 2 rate neutrons. Another is to substitute PMTs on the of the cylinder | of 678 keV. As has been discussed, imPurities of kr belOw 10 pPb cAn Be reAcheD [@xenon10], making neGLigiBle the contamination proDuced By THose RAdIoactIve decaYS.
NEUTroNs FrOm tArGEt CompoNenTs
-------------------------------
One of The most impOrtAnT sources of baCKgRound comes FroM neutrons proDucEd by raDiOacTIve coNtaMinatIon of tHE materIals constItUTing thE DetectoR ITsElf. TO minimize their ratE, ThE Use of copper for All the VeSSeLS Is lIkeLy to be the bEsT possIBle choiCE. THE RAdiOActive impuritIes can be redUCed Below 0.02 pPb In sOMe coppEr samPlES whIch would briNg thE neutron cOntamiNAtion to BElow 1 eveNt per yEar [@iLIaS]. If WE cOnSerVaTIveLY aSsuME a 0.1 pPb contamInAtIon, onE obtAINS A neuTroN proDuctiOn rate of $4.54\times10^{-11}$ S$^{-1}$cm$^{-3}$. beinG EacH cyliNder 6 mM thiCk, Its toTal volUme amOuNts to 2217 cm$^{3}$. This meanS a toTal of one nEutRoN peR cYlindER per yeAr.
THe cOntaminAtion inDUceD bY pmts Must be carefully evaLuATEd As well. MaIn manuFAcTuRErs contiNuE to OptiMIZe the ChoiCE oF materiaLs used IN PmT ConstruCtIon to rEdUce TheIr radIOactIvity lEvels. TypIcal cONtamination valUEs for U and Th raNGe FROm A Few tEns To several huNdreDS parTs peR BiLliON. AmonG the wIdE VaRIety of tubes availablE iN the maRket, iT is possible to Find out somE MODels specIallY DeSIgned for low bacKgrouNd applicatIOns where The meAsured urAnium and tHORium concEntRatIonS in QUArTz and metal comPONentS iS of the oRdeR of ten oR evEn lEss PpbS [@KUdryavtseV], giving a YeArLy PrOduCtion OF less thaN oNe nEuTroN per Pmt. The phOtotuBe wiNdOwS CouLd be coaTEd WITh TeTrA-PHenyL-BuTaDiene (tPB) tO ShiFt the ulTra-violet LigHT to tHe MaXimum of The phototube sPeCtral respoNsE wiThout aN INcrease oN contamination. If we assumE A rate of 1 NeuTron eMittEd per year Per pMT and 8 pMTS Per cylInder, wE expeCt A toTAL emisSIOn Of 8 nEuTrons per cyLINdeR. In toTaL, PMTS and the Copper vessel contriBUte To 9 neutrons emiTteD per CYLiNdeR PeR YeaR.
ALThoUGH they will not be cOnsidered iN tHE pResent work, THerE aRe severAl possiBilitIEs to redUce the ratE of neutroNs ComiNG FroM PMTs. One is To set acrYlic light-GUides BEtWeen pHotOmultiPlIerS and tHe actiVE voLume [@KUdryavTsEv] whicH can rEdUce by a faCtor 2 the rate of neutrons. AnOther pOssibIliTy is to subStiTUte The PMTs on The tOp of the cylIndEr bY | of 678 keV. As has been d iscussed,impur iti esof Krbelo w 10 ppb can b e rea ched [@xenon10], makin g neg li g ible th e con taminat i on p rod uc ed by t h os e rad ioa ctive d ecays.
Ne utr on s from targe t c omponents--- ------------ --- ------ -- --- - -
On e o f the mosti mporta nt source so f back g round c o m es fro m neutrons produc e db y radioactivecontam in a ti o n of th e material sconst i tutingt he d e tec t or itself. To minimize t h eir rate, t heu se ofcoppe rf orall the ves sels is likel y to b e the be s t possi ble ch oic e.Ther ad io act iv e im p ur iti e s c an be re du ce d bel ow 0 . 0 2 ppbinsome copp er samples wh ich wou l d b ringthe n eutr on cont aminat ion t obelow 1 event p er y ear [@ILI AS] .Ifwe cons e rvativ ely as sume a0.1 ppb con ta m i n at ion, one obtains a n e u tr on produ ctionr at eo f $4.54\ ti mes 10^{ - 1 1}$ s $^{- 1 }$ cm$^{-3} $. Bei n gea ch cyli nd er 6 m mthi ck, itst otal volum e amount s to2 217 cm$^{3}$.T his means a t o ta l of oneneu tron per cy lind e r pe r ye a r.
T h e con tamin at i on induced by PMTs mus tbe car efull y evaluated a s well. Ma i n manufact urer s c o ntinue to opti mizethe choice of mater ialsused inPMT const r u ction to re duc e t hei r ra dioactivity l e v els. T ypicalcon taminat ion va lue s f or U and Th range f ro mafe w t ens t o several h und re dsparts per bi llion . Am on gt hewide va r ie t y oftu be s av ail ab le in the mar ket, it is possi ble to f in dout som e models spec ia lly design ed fo r lowb a ckground applications where the measure d u raniu m an d thorium co ncentr ati o ns inquartz andme tal c ompon e n ts is o f the orde r often o reven less p pbs [@Kudryavtsev] , gi ving a yearly pr oduc t i on of le s s t ha n on e neutron per PMT . The phot ot u be windows c o uld b e coate d withTetra - Phenyl- Butadiene (TPB) to s hift t heultra-viol et light to the m a ximum of thepho totube s pec tralrespon s e w ithou t an i nc reaseon co nt aminatio n. If we assume a rateof 1 n eutro n e mitted pe r y e arper PMT a nd 8 PMTs percyl ind er, w e e x pecta to t al em i ssion of8 neutrons pe r c y l in der. In tot a l , PM Ts an d t h e copp er v essel contributet o 9 neutrons e mitt e d pe r c y lind er per year.
Al tho ug h they wil lnot be cons idered i nt he pr esentwork,there a r e s e veralposs ibi lities to re du c e the r at eo f neut rons c omingfrom P M Ts.O n e is to set acry lic l i g ht-gu i des betw ee n photo m ulti pliers and the active volum e [@ Kudry avtsev] w hich c anre duce by af actor 2 t he ra te of n eu tron s.Anothe r po s s ibili ty i stosubstitut e th e P MT s on the topof the cylinder by | of_678 keV. As_has been discussed, impurities_of Kr_below_10 ppb can_be_reached [@xenon10], making negligible_the contamination produced_by those radioactive decays.
Neutrons_from target components
-------------------------------
One_of_the most important sources of background comes from neutrons produced by radioactive contamination of_the_materials constituting_the_detector_itself. To minimize their rate,_the use of copper for_all the_vessels is likely to be the best possible_choice._The radioactive impurities_can be reduced below 0.02 ppb in some copper samples_which would bring the neutron contamination_to below 1_event_per_year [@ILIAS]. If we conservatively_assume a 0.1 ppb contamination, one obtains_a neutron production rate of $4.54\times10^{-11}$_s$^{-1}$cm$^{-3}$. Being each cylinder 6 mm thick, its_total volume amounts to 2217 cm$^{3}$._This means a total of_one neutron_per cylinder per year.
The contamination_induced by PMTs_must be_carefully evaluated as_well. Main manufacturers continue to optimize_the choice of_materials used in PMT construction to_reduce_their radioactivity levels._Typical_contamination_values for_U and Th_range_from a_few_tens to several hundreds parts per_billion._Among the wide variety of tubes available_in the market, it_is_possible to find out_some models specially designed for_low background applications where the measured_uranium and_thorium concentrations_in quartz and metal components is of the order of ten_or even less ppbs [@Kudryavtsev], giving a_yearly production of less_than one_neutron_per PMT. The_phototube_windows could_be coated with Tetra-Phenyl-Butadiene (TPB) to shift_the ultra-violet_light to the maximum of the_phototube spectral response without_an_increase on contamination. If we assume_a rate of 1 neutron emitted_per year per PMT and_8_PMTs_per cylinder, we expect a_total emission of 8 neutrons per_cylinder. In total,_PMTs and the copper vessel contribute to_9_neutrons emitted per cylinder per year.
Although_they_will not be considered in the_present_work,_there are several possibilities to_reduce the rate of neutrons coming_from PMTs. One is to set acrylic light-guides between_photomultipliers and the_active volume [@Kudryavtsev] which can reduce_by_a_factor 2 the rate of neutrons. Another possibility is to_substitute the_PMTs on the_top of the cylinder by |
Neufeld & Hollenbach 1994, 1996), and thus deep observations are needed to detect them.
Young protostars exhibit a positive or almost flat ($>-0.1$) spectral index (defined as $\alpha$, where $S_\nu\propto \nu^a$) between e.g. 3.6 cm and 6 cm (André et al.1987; Anglada et al. 1998; Beltrán et al. 2001). This behaviour of the radio spectral index is characteristic of free-free thermal emission from ionised gas. This is most likely produced by a partially ionised jet that propagates into the collapsing envelope producing shocks (e.g. Curiel et al. 1987). It could also be produced by an ionised disc wind (e.g. Martin 1996) or by the accretion shock that develops on the surface of the protostar, and/or on the surface of the disc that surrounds the protostar, which heats and ionises the infalling gas (Winkler & Newman 1980; Cassen & Moosman 1981). However, the latter two mechanisms produce radio emission that is at too low a level to be readily detected.
A negative ($<-0.1$) radio spectral index indicative of non-thermal gyro-synchrotron emission has also been observed in some T Tauri stars (e.g. André et al. 1988; André 1996; Rodríguez et al. 1999), in a possible Class I object (Feigelson et al. 1998), and even in a possible Class 0 object, the triple radio continuum source in Serpens (Rodríguez et al. 1989a; Curiel et al. 1993; Curiel 1995; Raga et al. 2000). The non-thermal emission of more evolved protostars (Class II, III) is believed to be associated with the magnetic fields around these objects and the star-disc interaction region (e.g. André 1996). The non-thermal emission from younger protostars is thought to be the result of particle acceleration behind a diffuse shock wave (Rodríguez et al. 1989a; Curiel et al. 1993 and references therein).
Stamatellos et al. (2005a,b) performed 3-dimensional radiative transfer calculations on simulations of the collapse of a star-forming core and studied the transition from a starless core to a core with an embedded young protostar. These simulations suggest (i | Neufeld & Hollenbach 1994, 1996), and thus deep observations are needed to detect them.
Young protostars parade a incontrovertible or almost flat ($ > -0.1 $) spectral exponent (defined as $ \alpha$, where $ S_\nu\propto \nu^a$) between e.g. 3.6 cm and 6 centimeter (André et al.1987; Anglada et al. 1998; Beltrán et al. 2001). This behaviour of the radio apparitional exponent is characteristic of free - free thermal discharge from ionised gas. This is most probably produced by a partially ionised k that propagates into the collapsing envelope produce shocks (e.g. Curiel et al. 1987). It could also be produce by an ionised disc wind (e.g. Martin 1996) or by the accretion daze that develops on the surface of the protostar, and/or on the surface of the disc that surrounds the protostar, which inflame and ionises the infalling gas (Winkler & Newman 1980; Cassen & Moosman 1981). However, the latter two mechanisms produce radio emission that is at too low a level to be readily detected.
A negative ($ < -0.1 $) radio spectral index indicative of non - thermal gyro - synchrotron emission has besides been observe in some T Tauri stars (for example André et al. 1988; André 1996; Rodríguez et al. 1999), in a potential Class I object (Feigelson et al. 1998), and even in a possible Class 0 object, the triple radio receiver continuum source in Serpens (Rodríguez et al. 1989a; Curiel et al. 1993; Curiel 1995; Raga et al. 2000). The non - thermal discharge of more evolved protostars (Class II, III) is believed to be associated with the charismatic fields around these objects and the star - disc interaction region (e.g. André 1996). The non - thermal emission from younger protostars is thought to be the result of particle acceleration behind a diffuse shock wave (Rodríguez et al. 1989a; Curiel et al. 1993 and reference therein).
Stamatellos et al. (2005a, b) perform 3 - dimensional radiative transfer calculations on simulations of the collapse of a star - form core and studied the transition from a starless effect to a core with an embedded young protostar. These simulations hint (i | Nekfeld & Hollenbach 1994, 1996), and uhus deep observajiins arx needes to detdct them.
Young protostars exhmbit a powitive or almost flat ($>-0.1$) spectral index (dwfintd as $\alpha$, where $S_\nu\propbj \nu^z$) betcexn e.g. 3.6 cm and 6 gm (André et dl.1987; Anglada et dl. 1998; Yeltrán et al. 2001). This behaviour of the wadio sleftral index is chagasterjstic of free-free thermal emission from imnised gas. Thos is most likely produced by w partially ionisef jet that kdopwtates into tfe collapspug envelope producing shocks (e.g. Curiel et xl. 1987). Ic could alsi ve otoduced by ai ionifed disc wind (e.g. Mardin 1996) or by the accretlon siock that develops on the surface of the prjtostar, atd/kr on the surface od the disw thxr sjrrkuids the pgotkstar, whicg heats and ionises the infallond gas (Winkler & Newmag 1980; Cassen & Moosman 1981). However, the latter twm mschanisms produce radio emission that is at joo low a jevel to be readily detected.
A negative ($<-0.1$) radio spewtral kndtx indizqtlve of non-thermal gyro-synchrotron emission haf aksp been observeb in some T Tauro dtsts (e.g. André et xl. 1988; Anbdé 1996; Rodríguez et al. 1999), ij a posfible Class I jbjevt (Feigelson et al. 1998), and eveb in a possiyle Class 0 object, the triple radno conjinuum source in Serpens (Rodríyuez ef al. 1989a; Curifl et al. 1993; Zuriel 1995; Raga et xl. 2000). Tve non-thtfmal emission of iore evolted pxotostarr (Clsss II, III) is bepieved to be associated aith jhe macnetic fiepds around these objects and thx star-disc injerdctpon regiou (e.g. Akdré 1996). The non-thewmal emission yrom youuger pfotostars ps thoughv to be the wesult of pardlcle acceleretion behynd q didfuse sfuck wave (Rodrígiez et al. 1989a; Curiel wt al. 1993 and referenges tfsrein).
Stamatellow eu ql. (2005a,b) performec 3-dkmegsponel raqhative transxer zalzilatiuns on simuoxtioms of the collapse ox a atar-forming core amd studied the tragsition from s starless core to a coce witi an ekbeqded young protostar. These simhlations dugnest (i | Neufeld & Hollenbach 1994, 1996), and thus are to detect Young protostars exhibit ($>-0.1$) index (defined as where $S_\nu\propto \nu^a$) e.g. 3.6 cm and 6 cm et al.1987; Anglada et al. 1998; Beltrán et al. 2001). This behaviour of radio spectral index is characteristic of free-free thermal emission from ionised gas. This most produced a ionised jet that propagates into the collapsing envelope producing shocks (e.g. Curiel et al. 1987). It also be produced by an ionised disc wind Martin 1996) or by accretion shock that develops on surface the protostar, on surface the disc that the protostar, which heats and ionises the infalling gas (Winkler & Newman 1980; Cassen & Moosman 1981). the latter produce radio that at low a level readily detected. A negative ($<-0.1$) radio of non-thermal gyro-synchrotron emission has also been observed some T stars (e.g. André et al. 1988; 1996; Rodríguez et al. 1999), in a possible I object (Feigelson et al. 1998), and even in a possible Class 0 object, the continuum source in Serpens et al. 1989a; et 1993; 1995; et al. The non-thermal emission of more evolved protostars (Class II, III) is to be associated with the magnetic fields around these objects star-disc region (e.g. André The non-thermal emission from protostars thought to be the particle behind wave et 1989a; Curiel et al. and references therein). Stamatellos et (2005a,b) performed 3-dimensional radiative the collapse of a star-forming core and studied transition from a starless core to a with an embedded young protostar. These simulations suggest (i | Neufeld & Hollenbach 1994, 1996), and thus dEep observaTions Are NeeDeD to dEtecT them.
Young protOStarS exhibit a positive or almOst flAt ($>-0.1$) SPectRAl Index (Defined AS $\aLPHa$, wHeRe $s_\nu\PrOPtO \nu^a$) bEtwEen e.g. 3.6 cm And 6 cm (André Et aL.1987; ANglada et al. 1998; BeLTrÁn et al. 2001). This BehAviour of the rAdiO spectRaL inDEx is cHarActerIstic oF Free-frEe thermal EmISsion fROm ionisED GaS. ThiS is most likely prodUCeD By a partially ioNised jEt THaT PRopAgaTes into the CoLlapsINg envelOPe PRODucINg shocks (e.g. CurIel et al. 1987). It coULd aLso be pRoDucED by an iOniseD dISc wInd (e.g. Martin 1996) Or by The accretIon shoCK that deVElops on The surFacE of The pROtOsTar, AnD/Or oN ThE suRFacE of the diSc ThAt surRounDS THE proTosTar, wHich hEats and ioniseS thE infALliNg gas (winklEr & NeWmAn 1980; CasSen & MooSman 1981). HOwEver, the latter twO mecHanisms prOduCe RadIo EmissIOn that Is aT toO low a leVel to be REadIlY DETeCted.
A negative ($<-0.1$) radio SpECTrAl index iNdicatIVe Of NOn-thermaL gYro-SyncHROtron EmisSIoN has also Been obSErVeD in some t TAuri stArS (e.g. andRé et aL. 1988; andrÉ 1996; RodríGuez et al. 1999), In a poSSible Class I objECt (Feigelson et AL. 1998), aND EvEN in a PosSible Class 0 oBjecT, The tRiplE RaDio COntinUum soUrCE iN serpens (Rodríguez et aL. 1989a; curiel Et al. 1993; CUriel 1995; Raga et al. 2000). the non-therMAL Emission Of moRE eVOlved protostarS (ClasS II, III) is beLIeved to bE assoCiated wiTh the magnETIc fields AroUnd TheSe oBJEcTs and the star-dISC intErAction rEgiOn (e.g. AndRé 1996). THe nOn-tHerMaL emission From younGeR pRoToStaRs is tHOught to bE tHe rEsUlt Of parTIcle acCelerAtioN bEhINd a Diffuse SHoCK Wave (roDrÍgueZ et Al. 1989A; CuriEl et AL. 1993 anD refereNces thereIn).
STAmatElLoS et al. (2005a,b) Performed 3-dimeNsIonal radiaTiVe tRansfeR CAlculatiOns on simulations of the coLLapse of A stAr-forMing Core and stUdiEd the tRanSItion fRom a stArlesS cOre TO A core WITh An eMbEdded young PROtoStar. THeSe siMulatioNs suggest (i | Neufeld & Hollenbach 1994 , 1996), a nd th usdee pobse rvat ions are neede d todetect them.
Young pr otost ar s exh i bi t a p ositive or a lmo st f lat ( $ >- 0.1$) sp ectralindex (def ine das $\alpha$, wh ere $S_\nu \pr opto \nu^a$) be tweene. g.3 .6 cm an d 6 c m (And r é et a l.1987; A ng l ada et al. 199 8 ; B eltr án et al. 2001).T hi s behaviour ofthe ra di o s p e ctr alindex is c ha racte r istic o f f r e e -fr e e thermal emi ssion fromi oni sed ga s. Th i s is m ost l ik e lyproduced by a p artiallyionise d jet th a t propa gatesint o t he c o ll ap sin ge nve l op e p r odu cing sho ck s(e.g. Cur i e l et a l.1987 ). It could also b e p rodu c edby an ioni seddi sc wi nd (e. g. Ma rt in 1996) or bytheaccretion sh oc k t ha t dev e lops o n t hesurface of the pro to s t a r, and/or on the sur fa c e o f the di sc tha t s ur r ounds th epro tost a r , whi ch h e at s and io nisest he i nfallin ggas (W in kle r & Newm a n 19 80; Ca ssen & M oosma n 1981). Howeve r , the lattert wo m ec h anis msproduce rad io e m issi on t h at is at to o low a le v el to be readily de te cted.
A ne gative ($<-0. 1$) radios p e ctral in dexi nd i cative of non- therm al gyro-sy n chrotron emis sion has also bee n observed in so meT T a u ri stars (e.g.A n dréet al. 19 88; André199 6;Rod ríg ue z et al.1999), i napo ss ibl e Cla s s I obje ct (F ei gel son e t al. 1 998), and e ve n in a poss i bl e Clas s0obje ct, t he tr iple rad io cont inuum sou rce in S er pe ns (Rod ríguez et al. 1 989a; Curi el et al. 1 9 9 3; Curie l 1995; Raga et al. 200 0 ). Thenon -ther malemissionofmore e vol v ed pro tostar s (Cl as s I I , III) i sbel ie ved to bea s soc iated w iththe mag netic fields aroun d th ese objects a ndthes t ar -di s ci nte ra c tio n region (e.g. An dré 1996). T h enon-therma l em is sion fr om youn ger p r otostar s is thou ght to be t he r e s ult of partic le accel eration b e hinda d iffus e s hock w av e ( Rodrí guez e t al . 198 9a; Cu ri el etal. 1 99 3 and re ferences therein).
Sta matell os et al . (2005a, b)p erf ormed 3-d imen sional rad iat ive tran sfe r calc ulat i on s o n simu lati o ns of the co lla p s eof a star-f o r m ing core an d studi ed t he transition fro m a starless co re t o a c ore with a n embedded you ngpr o t ostar. T he se simulati ons sugg es t (i | Neufeld_& Hollenbach_1994, 1996), and thus_deep observations_are_needed to_detect_them.
Young protostars exhibit_a positive or_almost flat ($>-0.1$) spectral_index (defined as_$\alpha$,_where $S_\nu\propto \nu^a$) between e.g. 3.6 cm and 6 cm (André et al.1987; Anglada_et_al. 1998;_Beltrán_et_al. 2001). This behaviour of_the radio spectral index is_characteristic of_free-free thermal emission from ionised gas. This is_most_likely produced by_a partially ionised jet that propagates into the collapsing_envelope producing shocks (e.g. Curiel et_al. 1987). It_could_also_be produced by an_ionised disc wind (e.g. Martin 1996)_or by the accretion shock that_develops on the surface of the protostar,_and/or on the surface of the_disc that surrounds the protostar,_which heats_and ionises the infalling gas_(Winkler & Newman_1980; Cassen_& Moosman 1981)._However, the latter two mechanisms produce_radio emission that_is at too low a level_to_be readily detected.
A_negative_($<-0.1$)_radio spectral_index indicative of_non-thermal_gyro-synchrotron emission_has_also been observed in some T_Tauri_stars (e.g. André et al. 1988; André_1996; Rodríguez et al._1999),_in a possible Class_I object (Feigelson et al._1998), and even in a possible_Class 0_object, the_triple radio continuum source in Serpens (Rodríguez et al. 1989a; Curiel_et al. 1993; Curiel 1995; Raga_et al. 2000). The_non-thermal emission_of_more evolved protostars_(Class_II, III)_is believed to be associated with the_magnetic fields_around these objects and the star-disc_interaction region (e.g. André_1996)._The non-thermal emission from younger protostars_is thought to be the result_of particle acceleration behind a_diffuse_shock_wave (Rodríguez et al. 1989a;_Curiel et al. 1993 and references_therein).
Stamatellos et al._(2005a,b) performed 3-dimensional radiative transfer calculations on_simulations_of the collapse of a star-forming_core_and studied the transition from a_starless_core_to a core with an_embedded young protostar. These simulations suggest_(i |
OTWY Example 2.2]).
Let ${\mathfrak{q}}= (a)$ and set $f = (a,0) \in I$. Then, $I^2 = fI$, since $I^2 = (a^2) \times (aR + \alpha R) = (a^2) \times aR=fI$. Note that $I/fA = [(a) \oplus R]/[(a) \oplus (a)] \cong R/(a)$ and $A/I = [R \oplus R]/[(a) \oplus R] \cong R/(a)$ as $R$-modules. We then have $\ell_A(A/I) =\ell_A(I/fA) = \ell_R(R/(a))$. Hence, $A/I \cong I/fA$ as an $A$-module, because $I/fA$ is a cyclic $A$-module. Thus, $I \in {\mathcal{X}}_A$ with $\mu_A(I)=2$.
Two-generated Ulrich ideals are totally reflexive $R$-modules (Proposition \[Ulrich\] (3)), possessing minimal free resolutions of a very restricted form. Let us note the following, which we need to prove Theorem \[2.3\]. We include a brief proof for the sake of completeness.
\[2.2\] Suppose that $I$ is an Ulrich ideal of $R$ and assume that $\mu_R(I)=2$. We write $I=(a, b)$ with $(a)$ a reduction of $I$. Therefore, $b^2 = ac$ for some $c \in I$. With this notation, the minimal free resolution of $I$ is given by $$\Bbb F : \ \ \cdots \to R^2 \overset{
\begin{pmatrix}
-b & -c\\
a & b
\end{pmatrix}}{\longrightarrow}
R^2 \overset{
\begin{pmatrix}
-b & -c\\
a & b
\end{pmatrix}}{\longrightarrow}R^2 \overset{\begin{pmatrix}
a & b
\end{pmatrix}}{\longrightarrow} I \to 0,$$ Hence $\pd_RI= \infty$. The ideal $I$ is so called a totally reflexive $R$-module, because $I$ is reflexive, $\Ext_ | OTWY Example 2.2 ]).
Let $ { \mathfrak{q}}= (a)$ and set $ f = (a,0) \in I$. Then, $ I^2 = fI$, since $ I^2 = (a^2) \times (aR + \alpha R) = (a^2) \times aR = fI$. Note that $ I / fA = [ (a) \oplus R]/[(a) \oplus (a) ] \cong R/(a)$ and $ deoxyadenosine monophosphate / I = [ R \oplus R]/[(a) \oplus radius ] \cong R/(a)$ as $ R$-modules. We then have $ \ell_A(A / I) = \ell_A(I / fA) = \ell_R(R/(a))$. Hence, $ vitamin a / I \cong I / fA$ as an $ A$ -module, because $ I / fA$ is a cyclic $ A$ -module. Thus, $ I \in { \mathcal{X}}_A$ with $ \mu_A(I)=2$.
Two - generated Ulrich ideal are wholly reflexive $ R$-modules (Proposition \[Ulrich\ ] (3) ), possessing minimal free resolutions of a very restricted form. lease us note the following, which we necessitate to prove Theorem \[2.3\ ]. We include a brief proof for the sake of completeness.
\[2.2\ ] Suppose that $ I$ is an Ulrich ideal of $ R$ and bear that $ \mu_R(I)=2$. We write $ I=(a, b)$ with $ (a)$ a reduction of $ I$. Therefore, $ b^2 = ac$ for some $ deoxycytidine monophosphate \in I$. With this notation, the minimal free resolution of $ I$ is given by $ $ \Bbb degree fahrenheit: \ \ \cdots \to R^2 \overset {
\begin{pmatrix }
-b & -c\\
a & b
\end{pmatrix}}{\longrightarrow }
R^2 \overset {
\begin{pmatrix }
-b & -c\\
a & b
\end{pmatrix}}{\longrightarrow}R^2 \overset{\begin{pmatrix }
a & b
\end{pmatrix}}{\longrightarrow } I \to 0,$$ Hence $ \pd_RI= \infty$. The ideal $ I$ is so predict a totally reflexive $ R$-module, because $ I$ is reflexive, $ \Ext _ | OTWJ Example 2.2]).
Let ${\mathfrak{q}}= (x)$ and set $f = (a,0) \nb I$. Thxn, $I^2 = fJ$, since $K^2 = (a^2) \times (aR + \alpha R) = (a^2) \tilew aR=fU$. Note that $I/fA = [(a) \opljs R]/[(a) \oplls (a)] \cong R/(a)$ end $A/I = [R \oplus C]/[(z) \oplus R] \conf R/(a)$ cs $R$-modules. We tmen have $\eln_A(A/I) =\ell_A(I/fA) = \alu_R(X/(a))$. Hence, $A/I \cong I/fA$ as an $A$-module, bqcause $O/fW$ is a cyclic $W$-modllq. Thhs, $I \in {\mathcal{X}}_A$ with $\mu_A(I)=2$.
Two-genedated Unrich ideals sre totally reflexive $R$-modkles (Proposition \[Ulricj\] (3)), possessibg mybimal free rdsolutions of a very rgstricted form. Let us note the fullownng, which wg hefg to prove Vheorei \[2.3\]. We includc a briaf proog for the sake of conpleteness.
\[2.2\] Suppose thet $I$ is an Ulrich idgal of $R$ atd assume that $\mu_R(I)=2$. Ww wrije $I=(a, b)$ wkrh $(x)$ a rxdudtion lf $M$. Therefore, $b^2 = ac$ for wome $c \in I$. With thos botation, the jinimaj sree resolution of $I$ is given by $$\Bbb F : \ \ \ddots \to R^2 \overset{
\begin{pmatrix}
-b & -c\\
a & b
\end{pmajrix}}{\longridhtarrow}
R^2 \overset{
\begin{pmatrix}
-b & -c\\
a & b
\end{pmatrix}}{\lotgrigitxrriw}V^2 \ovdesft{\begin{pmatrix}
a & b
\end{pmatrix}}{\longrightarrow} I \ek 0,$$ Hvnce $\pd_RI= \infty$. Tme ideal $I$ is so cslpec a totally rewlexivz $R$-jodule, because $I$ id reflevive, $\Wxt_ | OTWY Example 2.2]). Let ${\mathfrak{q}}= (a)$ and = \in I$. $I^2 = fI$, (aR \alpha R) = \times aR=fI$. Note $I/fA = [(a) \oplus R]/[(a) \oplus \cong R/(a)$ and $A/I = [R \oplus R]/[(a) \oplus R] \cong R/(a)$ as We then have $\ell_A(A/I) =\ell_A(I/fA) = \ell_R(R/(a))$. Hence, $A/I \cong I/fA$ as an because is cyclic Thus, $I \in {\mathcal{X}}_A$ with $\mu_A(I)=2$. Two-generated Ulrich ideals are totally reflexive $R$-modules (Proposition \[Ulrich\] (3)), minimal free resolutions of a very restricted form. us note the following, we need to prove Theorem We a brief for sake completeness. \[2.2\] Suppose $I$ is an Ulrich ideal of $R$ and assume that $\mu_R(I)=2$. We write $I=(a, b)$ with $(a)$ reduction of $b^2 = for $c I$. With this minimal free resolution of $I$ is F : \ \ \cdots \to R^2 \overset{ -b & a & b \end{pmatrix}}{\longrightarrow} R^2 \overset{ -b & -c\\ a & b \end{pmatrix}}{\longrightarrow}R^2 \overset{\begin{pmatrix} & b \end{pmatrix}}{\longrightarrow} I \to 0,$$ Hence $\pd_RI= \infty$. The ideal $I$ is so called reflexive $R$-module, because $I$ reflexive, $\Ext_ | OTWY Example 2.2]).
Let ${\mathfrak{q}}= (a)$ aNd set $f = (a,0) \in I$. then, $I^2 = FI$, sIncE $I^2 = (A^2) \timEs (aR + \Alpha R) = (a^2) \times aR=Fi$. NotE that $I/fA = [(a) \oplus R]/[(a) \oplus (a)] \Cong R/(A)$ aND $A/I = [R \OPlUs R]/[(a) \oPlus R] \coNG R/(A)$ AS $R$-mOdUlEs. WE tHEn Have $\eLl_A(a/I) =\ell_A(I/FA) = \ell_R(R/(a))$. HeNce, $a/I \Cong I/fA$ as an $A$-MOdUle, because $i/fA$ Is a cyclic $A$-moDulE. Thus, $I \In {\MatHCal{X}}_A$ WitH $\mu_A(I)=2$.
two-genERated ULrich ideaLs ARe totaLLy refleXIVe $r$-modUles (Proposition \[UlRIcH\] (3)), Possessing miniMal freE rESoLUTioNs oF a very restRiCted fORm. Let us NOtE THE foLLowing, which we Need to prove tHeoRem \[2.3\]. We iNcLudE A brief Proof FoR The Sake of complEtenEss.
\[2.2\] SupposE that $I$ IS an UlriCH ideal oF $R$ and aSsuMe tHat $\mU_r(I)=2$. we WriTe $i=(A, b)$ wITh $(A)$ a rEDucTion of $I$. THeReFore, $b^2 = Ac$ foR SOME $c \in i$. WiTh thIs notAtion, the minimAl fRee rESolUtion Of $I$ is GiveN bY $$\Bbb F : \ \ \Cdots \tO R^2 \oveRsEt{
\begin{pmatrix}
-b & -C\\
a & b
\eNd{pmatrix}}{\LonGrIghTaRrow}
R^2 \OVerset{
\BegIn{pMatrix}
-b & -C\\
a & b
\end{pMAtrIx}}{\LONGrIghtarrow}R^2 \overset{\bEgIN{PmAtrix}
a & b
\eNd{pmatRIx}}{\LoNGrightarRoW} I \tO 0,$$ HenCE $\Pd_RI= \iNfty$. tHe Ideal $I$ is So callED a ToTally reFlExive $R$-MoDulE, beCause $i$ Is reFlexivE, $\Ext_ | OTWY Example 2.2]).
Let $ {\mathfrak {q}}= (a )$an d se t $f = (a,0) \in I $ . Th en, $I^2 = fI$, since$I^2=( a^2) \t imes(aR + \ a lp h a R) = ( a^2 )\ ti mes a R=f I$. Not e that $I/ fA=[(a) \oplusR ]/ [(a) \oplu s ( a)] \cong R/ (a) $ and$A /I= [R \ opl us R] /[(a)\ oplusR] \congR/ ( a)$ as $R$-mod u l es . We then have $\ell_ A (A / I) =\ell_A(I/f A) = \ el l _R ( R /(a ))$ . Hence, $ A/ I \co n g I/fA$ as a n $A $ -module, beca use $I/fA$i s a cycli c$A$ - module . Thu s, $I\in {\mathc al{X }}_A$ wit h $\mu _ A(I)=2$ .
Two-g enerat edUlr ichi de al s a re tot a ll y r e fle xive $R$ -m od ules(Pro p o s i tion \[ Ulri ch\](3)), possess ing min i mal free reso luti on s ofa very rest ri cted form. Letus n ote the f oll ow ing ,which we nee d t o p rove Th eorem \ [ 2.3 \] . W einclude a brief pr oo f fo r the sa ke ofc om pl e teness.
\ [2. 2\]S u ppose tha t $ I$ is an Ulric h i de al of $ R$ and a ss ume th at $\ m u_R( I)=2$. We writ e $I= ( a, b)$ with $( a )$ a reductio n o f $I $ . Th ere fore, $b^2= ac $ for som e $ c \ i n I$. With t h is notation, the minim al freeresol ution of $I$is given b y $ $\Bbb F: \\ \ c dots \to R^2 \ overs et{
\begin { pmatrix}
-b & -c\\
a& b
\end{ p m atrix}}{ \lo ngr igh tar r o w}
R^2 \overset { \beg in {pmatri x}-b & -c \\a & b\en d{ pmatrix}} {\longri gh ta rr ow }R^ 2 \ov e rset{\be gi n{p ma tri x}
a& b
\en d{pma trix }} {\ l ong rightar r ow } I \t o0, $$ H enc e$\pd_ RI=\ inf ty$. Th e ideal $ I$i s so c al led a t otally reflex iv e $R$-modu le , b ecause $ I$ is re flexive, $\Ext_ | OTWY Example_2.2]).
Let ${\mathfrak{q}}=_(a)$ and set $f_= (a,0)_\in_I$. Then,_$I^2_= fI$, since_$I^2 = (a^2)_\times (aR + \alpha_R) = (a^2)_\times_aR=fI$. Note that $I/fA = [(a) \oplus R]/[(a) \oplus (a)] \cong R/(a)$ and $A/I_=_[R \oplus_R]/[(a)_\oplus_R] \cong R/(a)$ as $R$-modules._We then have $\ell_A(A/I) =\ell_A(I/fA)_= \ell_R(R/(a))$._Hence, $A/I \cong I/fA$ as an $A$-module, because_$I/fA$_is a cyclic_$A$-module. Thus, $I \in {\mathcal{X}}_A$ with $\mu_A(I)=2$.
Two-generated Ulrich ideals_are totally reflexive $R$-modules (Proposition \[Ulrich\]_(3)), possessing minimal_free_resolutions_of a very restricted_form. Let us note the following,_which we need to prove Theorem_\[2.3\]. We include a brief proof for_the sake of completeness.
\[2.2\] Suppose that_$I$ is an Ulrich ideal_of $R$_and assume that $\mu_R(I)=2$. We_write $I=(a, b)$_with $(a)$_a reduction of_$I$. Therefore, $b^2 = ac$ for_some $c \in_I$. With this notation, the minimal_free_resolution of $I$_is_given_by $$\Bbb_F : \_\_\cdots \to_R^2_\overset{
\begin{pmatrix}
-b & -c\\
a & b
\end{pmatrix}}{\longrightarrow}
R^2 \overset{
\begin{pmatrix}
-b_&_-c\\
a & b
\end{pmatrix}}{\longrightarrow}R^2 \overset{\begin{pmatrix}
a & b
\end{pmatrix}}{\longrightarrow} I_\to 0,$$ Hence $\pd_RI=_\infty$._The ideal $I$ is_so called a totally reflexive_$R$-module, because $I$ is reflexive, $\Ext_ |
them. In Section \[sec:Hipp\] we compare our spectrophotometric parallaxes to Hipparcos parallaxes and ask how these comparisons are affected by neglecting extinction. In Section \[sec:all\] we analyse our distances to the generality of stars, using kinematic correction factors to test for systematic biases in distances as functions of surface gravity or effective temperature, and to modify distance pdfs (Section \[sec:SBA\]). In Section \[sec:clusters\] we compare our distances to cluster stars with the established distances to their clusters. In Section \[sec:repeat\] we examine the scatter in the distances to the same star obtained from different spectra. In Section \[sec:Av\] we examine the distribution of extinctions to stars. Section \[sec:discuss\] sums up.
Methodology {#sec:method}
===========
As in B11 we start from the trivial Bayesian statement $$\pr(\hbox{model}|\hbox{data})={\pr(\hbox{data}|\hbox{model})\pr(\hbox{model})
\over\pr(\hbox{data})},$$ where “data” comprises the observed parameters and photometry of an individual star and “model” comprises a star of specified initial mass $\cM$, age $\tau$, metallicity $\hbox{[M/H]}$, and location. We use $p({\rm
model|data})$ either to calculate expectation values $\ex{x}$ and dispersions $\sigma_x$ of quantities of interest, such as the stars’s distance $x=s$ and parallax $x=\varpi$, by integrating $P({\rm model|data})$ times an appropriate power of $x$ through the space spanned by the model parameters $\hbox{[M/H]},\tau,\cM,\ldots$, or the pdf in distance modulus by marginalising $P({\rm model|data})$ over all model parameters other than distance.
A key role is played by the prior probability $\pr(\hbox{model})$, which reflects our prior knowledge of the Galaxy: massive young stars are rarely found far from the plane, while a star far from the plane is likely to be old and have sub-solar abundances. We have used the same three-component prior used in B11: $$\label{eq:priorofx}
p(\hbox{model}) = p(\cM) \sum_{i=1}^3 p_i(\mh) \, p_i(\tau) \, p | them. In Section \[sec: Hipp\ ] we compare our spectrophotometric parallaxes to Hipparcos parallaxes and ask how these comparisons are involve by neglect extinction. In Section \[sec: all\ ] we analyse our distances to the generality of ace, using kinematic correction agent to quiz for systematic bias in distances as functions of airfoil gravity or effective temperature, and to modify distance pdfs (Section \[sec: SBA\ ]). In Section \[sec: clusters\ ] we compare our distance to bunch stars with the established distance to their clusters. In Section \[sec: repeat\ ] we examine the scatter in the distance to the same star obtained from unlike spectra. In Section \[sec: Av\ ] we examine the distribution of extinction to stars. part \[sec: discuss\ ] sums up.
Methodology { # sec: method }
= = = = = = = = = = =
As in B11 we start from the trivial Bayesian statement $ $ \pr(\hbox{model}|\hbox{data})={\pr(\hbox{data}|\hbox{model})\pr(\hbox{model })
\over\pr(\hbox{data})},$$ where “ datum ” comprises the observed parameters and photometry of an individual star and “ model ” comprises a asterisk of specified initial mass $ \cM$, age $ \tau$, metallicity $ \hbox{[M / H]}$, and location. We use $ p({\rm
model|data})$ either to calculate expectation values $ \ex{x}$ and dispersions $ \sigma_x$ of quantities of sake, such as the stars ’s distance $ x = s$ and parallax $ x=\varpi$, by integrate $ P({\rm model|data})$ time an appropriate power of $ x$ through the space spanned by the exemplar parameters $ \hbox{[M / H]},\tau,\cM,\ldots$, or the pdf in distance modulus by marginalising $ P({\rm model|data})$ over all model parameters early than distance.
A key role is played by the prior probability $ \pr(\hbox{model})$, which reflects our prior knowledge of the Galaxy: massive young stars are rarely found far from the plane, while a star far from the airplane is likely to be erstwhile and have sub - solar abundances. We have used the same three - component prior practice in B11: $ $ \label{eq: priorofx }
p(\hbox{model }) = p(\cM) \sum_{i=1}^3 p_i(\mh) \, p_i(\tau) \, p | thfm. In Section \[sec:Hipp\] we gompare our specjriphotoketric parallabes to Hipparcos parallaxes end qsk hiw these comparisons afe affectvd by negoectmng extinction. Ii Sectiok \[fec:amp\] we enalyse our disjances to tha generality ox rtcrs, using kinematic correction factows to trsh for systematyc bpafes jn distances as functions of surfade gravpty or effective yemperature, and to modify fistwnce pdfs (Section \[sfc:SBA\]). In Sextiog \[wec:clusters\] de compare our distancgs to cluster stars with the estxblisked distancgs tl their clusvers. Ig Section \[seg:gepeat\] fe examone the scattev in vhe eistances to the same star obtained froi differett spectra. In Sectiin \[sec:Ae\] we exaoune tht dmstdibutiln kf extinctjons to staes. Section \[sec:discusx\] flks up.
Methodomogy {#sqc:iethod}
===========
As in B11 we start from the trivial Bagesian statement $$\pr(\hbox{nodel}|\hbox{data})={\pr(\hbox{daja}|\hbox{modej})\pr(\hbox{model})
\over\pr(\hbox{data})},$$ where “data” comprises tve obadrvtd paraowtfrs and photometry of an individual star and “ikdtl” bomprises a star jf specifiec lnojial mass $\cM$, aee $\tau$, mefallicity $\hbox{[M/H]}$, ajd locajion. Ww use $p({\rm
iodek|data})$ either to calculate ezpectation vcluws $\ex{x}$ and dispersnons $\sigma_x$ uf qoantitoes of interest, such as the stars’s dishance $x=s$ zvd parallax $x=\varoi$, ny hntegrating $P({\rm model|data})$ eimes an eppro'riate puwer of $x$ ehrough thf spaga spanned by the mldel karameders $\hbox{[M/J]},\tau,\cM,\ldots$, or the pdf in distaire modulus by mdrgpnalising $P({\rm kodel|data})$ ovew all model patameters jther than distznce.
A kxy role is pjayed by the [tior probabilmty $\pr(\hbov{modwl})$, wyich rewuects our priot knowledyt of the Gqlaxy: massive younn stats are rarely fouub dar from the plsne, whylv a sedr far from dhe olavr is uikely uj ne uld snd have sub-solar abgndahces. We have used yhc same thtee-componqnt prior usec in B11: $$\label{eq:priogofx}
p(\huox{modrl}) = p(\cM) \sum_{i=1}^3 p_i(\mh) \, p_i(\tau) \, p | them. In Section \[sec:Hipp\] we compare our to parallaxes and how these comparisons In \[sec:all\] we analyse distances to the of stars, using kinematic correction factors test for systematic biases in distances as functions of surface gravity or effective and to modify distance pdfs (Section \[sec:SBA\]). In Section \[sec:clusters\] we compare our to stars the distances to their clusters. In Section \[sec:repeat\] we examine the scatter in the distances to the star obtained from different spectra. In Section \[sec:Av\] examine the distribution of to stars. Section \[sec:discuss\] sums Methodology =========== As B11 start the trivial Bayesian $$\pr(\hbox{model}|\hbox{data})={\pr(\hbox{data}|\hbox{model})\pr(\hbox{model}) \over\pr(\hbox{data})},$$ where “data” comprises the observed parameters and photometry of an individual star and “model” comprises star of mass $\cM$, $\tau$, $\hbox{[M/H]}$, location. We use either to calculate expectation values $\ex{x}$ of quantities of interest, such as the stars’s $x=s$ and $x=\varpi$, by integrating $P({\rm model|data})$ times appropriate power of $x$ through the space spanned the model parameters $\hbox{[M/H]},\tau,\cM,\ldots$, or the pdf in distance modulus by marginalising $P({\rm model|data})$ over parameters other than distance. key role is by prior $\pr(\hbox{model})$, reflects our knowledge of the Galaxy: massive young stars are rarely found far the plane, while a star far from the plane is be and have sub-solar We have used the three-component used in B11: $$\label{eq:priorofx} p(\cM) p_i(\mh) p | them. In Section \[sec:Hipp\] we comPare our speCtropHotOmeTrIc paRallAxes to HipparcoS ParaLlaxes and ask how these coMpariSoNS are AFfEcted By negleCTiNG ExtInCtIon. in sEcTion \[sEc:aLl\] we anaLyse our disTanCeS to the generaLItY of stars, usIng Kinematic corRecTion faCtOrs TO test For SysteMatic bIAses in Distances As FUnctioNS of surfACE gRaviTy or effective tempERaTUre, and to modify DistanCe PDfS (sEctIon \[Sec:SBA\]). In SeCtIon \[seC:ClusterS\] We COMParE Our distances tO cluster staRS wiTh the eStAblIShed diStancEs TO thEir clusters. in SeCtion \[sec:rEpeat\] wE Examine THe scattEr in thE diStaNces TO tHe SamE sTAr oBTaIneD FroM differeNt SpEctra. in SeCTION \[sec:av\] wE exaMine tHe distributioN of ExtiNCtiOns to Stars. sectIoN \[sec:dIscuss\] Sums uP.
MEthodology {#sec:meThod}
===========
as in B11 we stArt FrOm tHe TriviAL BayesIan StaTement $$\pR(\hbox{moDEl}|\hBoX{DATa})={\Pr(\hbox{data}|\hbox{modeL})\pR(\HBoX{model})
\ovEr\pr(\hbOX{dAtA})},$$ Where “datA” cOmpRiseS THe obsErveD PaRameters And phoTOmEtRy of an iNdIviduaL sTar And “Model” COmprIses a sTar of speCifieD Initial mass $\cM$, aGE $\tau$, metalliciTY $\hBOX{[M/h]}$, And lOcaTion. We use $p({\rM
modEL|datA})$ eitHEr To cALculaTe expEcTAtIOn values $\ex{x}$ and dispeRsIons $\siGma_x$ oF quantities of Interest, suCH AS the starS’s diSTaNCe $x=s$ and parallaX $x=\varPi$, by integrATing $P({\rm mOdel|dAta})$ times An approprIATe power oF $x$ tHroUgh The SPAcE spanned by the MODel pArAmeters $\HboX{[M/H]},\tau,\cm,\ldOts$, Or tHe pDf In distancE modulus By MaRgInAliSing $P({\RM model|daTa})$ OveR aLl mOdel pARameteRs othEr thAn DiSTanCe.
A key rOLe IS PlayEd By The pRioR pRobabIlitY $\Pr(\hBox{modeL})$, which refLecTS our PrIoR knowleDge of the GalaxY: mAssive younG sTarS are raRELy found fAr from the plane, while a staR Far from The Plane Is liKely to be oLd aNd have Sub-SOlar abUndancEs. We hAvE usED The saME ThRee-CoMponent priOR UseD in B11: $$\lAbEl{eq:PriorofX}
p(\hbox{model}) = p(\cM) \sum_{i=1}^3 P_I(\mh) \, P_i(\tau) \, p | them. In Section \[sec:Hi pp\] we co mpare ou r s pe ctro phot ometric parall a xesto Hipparcos parallaxe s and a s k ho w t hesecompari s on s are a ff ect ed by negl ect ing ext inction. I n S ec tion \[sec:a l l\ ] we analy seour distance s t o thege ner a lityofstars , usin g kinem atic corr ec t ion fa c tors to t es t fo r systematic bias e si n distances as funct io n so f su rfa ce gravity o r eff e ctive t e mp e r a tur e , and to modi fy distance pdf s (Sec ti on\ [sec:S BA\]) .I n S ection \[se c:cl usters\]we com p are our distanc es toclu ste r st a rs w ith t h e e s ta bli s hed distanc es t o the ir c l u s t ers. In Sec tion\[sec:repeat\ ] w e ex a min e the scat terin thedistan ces t othe same star o btai ned fromdif fe ren tspect r a. InSec tio n \[sec :Av\] w e ex am i n e t he distribution of e x t in ctions t o star s .Se c tion \[s ec :di scus s \ ] sum s up .
Methodol ogy {# s ec :m ethod}== ====== == =
Asin B1 1 westartfrom the triv i al Bayesian st a tement $$\pr( \ hb o x {m o del} |\h box{data})= {\pr ( \hbo x{da t a} |\h b ox{mo del}) \p r (\ h box{model})
\over\p r( \hbox{ data} )},$$ where “ data” comp r i s es the o bser v ed parameters and phot ometry ofa n indivi dualstar and “model”c o mprisesa s tar of sp e c if ied initial m a s s $\ cM $, age$\t au$, me tal lic ity $\ hb ox{[M/H]} $, and l oc at io n. We use$ p({\rm
m od el| da ta} )$ ei t her to calc ulat eex p ect ation v a lu e s $\e x{ x} $ an d d is persi ons$ \si gma_x$of quanti tie s ofin te rest, s uch as the st ar s’s distan ce $x =s$ an d parallax $x=\varpi$, by integra t ing $P( {\r m mod el|d ata})$ ti mes an ap pro p riatepowerof $x $thr o u gh th e sp ace s panned byt h e m odelpa rame ters $\ hbox{[M/H]},\tau,\ c M,\ ldots$, or th e p df i n di sta n ce mod ul u s b y marginalising $ P({\rm mod el | da ta})$ over all m odel pa rameter s oth e r thandistance.
A key r ol e is p lay ed by theprior pr obability $\pr( \ hb ox{mo del })$, w hi chrefle cts ou r pr ior k nowled ge of th e Gal ax y: massi ve young stars are rare ly fou nd fa r f rom the p lan e , w hile a st ar f ar from th e p lan e islik e ly to beo ld an d have sub - solar abu n da nce s . W e have used t h e s ame t hre e -compo nent prior used in B1 1 : $$\label{eq: prio r o fx}
p (\hb ox {model}) = p(\ cM) \ s u m_{i=1}^ 3p_i(\mh) \, p_i(\ta u) \, p | them._In Section \[sec:Hipp\]_we compare our spectrophotometric_parallaxes to_Hipparcos_parallaxes and_ask_how these comparisons_are affected by_neglecting extinction. In Section \[sec:all\]_we analyse our_distances_to the generality of stars, using kinematic correction factors to test for systematic biases_in_distances as_functions_of_surface gravity or effective temperature,_and to modify distance pdfs_(Section \[sec:SBA\]). In_Section \[sec:clusters\] we compare our distances to cluster stars_with_the established distances_to their clusters. In Section \[sec:repeat\] we examine the_scatter in the distances to the_same star obtained_from_different_spectra. In Section \[sec:Av\]_we examine the distribution of extinctions_to stars. Section \[sec:discuss\] sums up.
Methodology {#sec:method}
===========
As_in B11 we start from the trivial_Bayesian statement $$\pr(\hbox{model}|\hbox{data})={\pr(\hbox{data}|\hbox{model})\pr(\hbox{model})
\over\pr(\hbox{data})},$$ where “data” comprises_the observed parameters and photometry_of an_individual star and “model” comprises_a star of_specified initial_mass $\cM$, age_$\tau$, metallicity $\hbox{[M/H]}$, and location. We_use $p({\rm
model|data})$ either_to calculate expectation values $\ex{x}$ and_dispersions_$\sigma_x$ of quantities_of_interest,_such as_the stars’s distance_$x=s$_and parallax_$x=\varpi$,_by integrating $P({\rm model|data})$ times an_appropriate_power of $x$ through the space spanned_by the model parameters_$\hbox{[M/H]},\tau,\cM,\ldots$,_or the pdf in_distance modulus by marginalising $P({\rm_model|data})$ over all model parameters other_than distance.
A_key role_is played by the prior probability $\pr(\hbox{model})$, which reflects our prior_knowledge of the Galaxy: massive young_stars are rarely found_far from_the_plane, while a_star_far from_the plane is likely to be old_and have_sub-solar abundances. We have used the_same three-component prior used_in_B11: $$\label{eq:priorofx}
p(\hbox{model}) = p(\cM)_\sum_{i=1}^3 p_i(\mh) \, p_i(\tau) \, p |
{width="160mm"}
{width="160mm"}
{width="160mm"}
{width="160mm"}
{width="160mm"}
---
abstract: 'Guidelines and consistency rules of UML are used to control the degrees of freedom provided by the language to prevent faults. Guidelines are used in specific domains (e.g., avionics) to recommend the proper use of technologies. Consistency rules are used to deal with inconsistencies in models. However, guidelines and consistency rules use informal restrictions on the uses of languages, which makes checking difficult. In this paper, we consider these problems from a language-theoretic view. We propose the formalism of C-Systems, short for “formal language control systems”. A C-System consists of a controlled grammar and a controlling grammar. Guidelines and consistency rules are formalized as controlling grammars that control the uses of UML, i.e. the derivations using the grammar of UML. This approach can be implemented as a parser, which can automatically verify the rules on a UML user model in XMI format. A comparison to related work shows our contribution: a generic top-down and syntax-based approach that checks language level constraints at compile-time.'
author:
- Zhe Chen
- Gilles Motet
bibliography:
-'main.bib'
title: 'A Language-theoretic View on Guidelines and Consistency Rules of UML'
---
Introduction {#Sec:intro}
============
The UML (Unified Modeling Language) is a graphic modeling language developed by OMG (Object Management Group), and defined by the specifications [@UMLi] and [@UMLs]. UML has emerged as the software industry’s dominant modeling language for specifying, designing and documenting the artifacts of systems [@MRR02][@BRJ05].
Evolving descriptions of software artifacts are frequently inconsistent, and tolerating this inconsistency is important [@Bal91][@EC01]. Different developers construct and update these descriptions at different times during development [@NER00], thus resulting in inconsistencies. They develop multiple views on a system providing pieces of information which are redundant or complementary. Constraints exist on these pieces of information whose violation leads to inconsistent models. Inconsistency problems of | ! [ image](figure9.eps){width="160 mm " }
! [ image](figure10.eps){width="160 mm " }
! [ image](figure11.eps){width="160 mm " }
! [ image](figure12.eps){width="160 mm " }
! [ image](figure13.eps){width="160 millimeter " }
---
abstract:' guidepost and consistency rules of UML are used to see the degree of freedom provided by the terminology to prevent fault. Guidelines are use in specific domains (e.g., avionics) to commend the proper use of technologies. Consistency rules are used to share with inconsistencies in models. However, guidelines and consistency rules use cozy restrictions on the uses of languages, which make checking difficult. In this composition, we consider these problem from a language - theoretic view. We propose the formalism of C - Systems, short for “ conventional language control systems ”. A C - System consists of a controlled grammar and a controlling grammar. Guidelines and consistency rules are formalized as controlling grammars that control the uses of UML, i.e. the derivations using the grammar of UML. This approach can be implement as a parser, which can automatically verify the rule on a UML exploiter model in XMI format. A comparison to related work shows our contribution: a generic circus tent - down and syntax - based approach that checks language floor constraints at compile - time.'
author:
- Zhe Chen
- Gilles Motet
bibliography:
-'main.bib'
title:' A lyric - theoretic View on Guidelines and Consistency Rules of UML'
---
Introduction { # Sec: intro }
= = = = = = = = = = = =
The UML (Unified Modeling Language) is a graphic mold language develop by OMG (Object Management Group), and defined by the specifications [ @UMLi ] and [ @UMLs ]. UML has emerged as the software industry ’s dominant mold language for specifying, designing and document the artifacts of systems [ @MRR02][@BRJ05 ].
Evolving descriptions of software artifact are frequently inconsistent, and tolerating this inconsistency is important [ @Bal91][@EC01 ]. Different developer construct and update these description at different fourth dimension during development [ @NER00 ], therefore resulting in inconsistencies. They develop multiple views on a system supply pieces of data which are redundant or complementary. Constraints exist on these pieces of data whose violation leads to inconsistent models. incompatibility problems of | {width="160mm"}
{width="160mm"}
{wjdth="160mm"}
{width="160mm"}
{widty="160mm"}
---
abstract: 'Guideliner and condistency rults of UML are usev to conbxol tgc degxexs of freedom ptovided by tve language to pferent faults. Guidelines are used in s[ecific dlmains (e.g., aviogics) eo rsbonmend the proper use of technklogies. Consistency tules are used to deal witj infonsistencies in mldels. Howevgd, gtudelines and consistenbv rules use informal restrictions on the ures oy languages, wyicj makes checjing qifficult. In this pa[er, we vonsider these prmblwms from a language-thxoretic view. We propjse the fmrjalism of C-Systemw, whort for “foroql uanfuege contrll aystems”. A D-System conwists of a controlltd dgsmmar and a dontrojlyng grammar. Guidelines and consistency gulea are formalized as conrrolling grammars thaj control ehe uses of UML, i.e. the derivations using the gramkar oh JML. Tmis xppgoach can be implemented as a parser, which cag aitpmatically verlfy the rules on a ULL oser model in BMI foxjaf. A comparison to gelated work shows ouw comtribution: a generic top-dowb and syntax-yaswd approach that ckecks languaye levgl conxtraints at compile-time.'
cuthor:
- Zhe Chen
- Gllles Motsg
bibliography:
-'maiv.bin'
tidle: 'A Language-theoretic Viqw on Guivelinzs and Cunsixtency Rules of KML'
---
Inbsoduction {#Sec:intro}
============
Hhe ULL (Unified Mldeling Language) is a graphic modeling languabe deneloped bv OMG (Pbject Manageient Group), and defineb by tfe specifibations [@UKLi] and [@UMLf]. UML has emeshed as the smftware ynduwtry’w dominxvt modeling lamguage fog fpecufying, designing akd dozhmenting the arcnfqcts of systems [@MRF02][@BRT05].
Enolting qascriptions mf suftdsre aftifacts arc ffequrntly inconsistent, atd tklerating this incpnfistency is impowtant [@Bal91][@EC01]. Dofferent developerd convtrnct anc ukdate these descriptions at difrerent tiles during devejopmcnt [@GER00], thus rzsulting in inconsistencies. They develop multiple views on a system providing kieces of infprmatyon which are redundant or conplementary. Constvaints exist on these lieces of ijformation whose violation leads to inconsistent models. Inconsistency provlems of | {width="160mm"} {width="160mm"} {width="160mm"} {width="160mm"} {width="160mm"} --- abstract: consistency of UML used to control by language to prevent Guidelines are used specific domains (e.g., avionics) to recommend proper use of technologies. Consistency rules are used to deal with inconsistencies in However, guidelines and consistency rules use informal restrictions on the uses of languages, makes difficult. this we consider these problems from a language-theoretic view. We propose the formalism of C-Systems, short for language control systems”. A C-System consists of a grammar and a controlling Guidelines and consistency rules are as grammars that the of i.e. the derivations the grammar of UML. This approach can be implemented as a parser, which can automatically verify the on a model in format. comparison related work shows a generic top-down and syntax-based approach level constraints at compile-time.' author: - Zhe Chen Gilles Motet -'main.bib' title: 'A Language-theoretic View on and Consistency Rules of UML' --- Introduction {#Sec:intro} The UML (Unified Modeling Language) is a graphic modeling language developed by OMG (Object Management defined by the specifications and [@UMLs]. UML emerged the industry’s modeling language specifying, designing and documenting the artifacts of systems [@MRR02][@BRJ05]. Evolving descriptions software artifacts are frequently inconsistent, and tolerating this inconsistency is Different construct and update descriptions at different times development thus resulting in inconsistencies. multiple on pieces information are redundant or complementary. exist on these pieces of whose violation leads to | {width="160mm"}
{Width="160Mm"}
{Width="160mm"}
{width="160mm"}
{widtH="160mM"}
---
AbstRAcT: 'GuidElines aND cONSisTeNcY ruLeS Of uML arE usEd to conTrol the degReeS oF freedom provIDeD by the langUagE to prevent faUltS. GuideLiNes ARe useD in SpeciFic domAIns (e.g., aVionics) to ReCOmmend THe propeR USe Of teChnologies. ConsistENcY Rules are used to Deal wiTh INcONSisTenCies in modeLs. howevER, guidelINeS AND coNSistency rules Use informal REstRictioNs On tHE uses oF langUaGEs, wHich makes chEckiNg difficuLt. In thIS paper, wE ConsideR these ProBleMs frOM a LaNguAgE-TheOReTic VIew. we proposE tHe FormaLism OF c-sYsteMs, sHort For “foRmal language cOntRol sYSteMs”. A C-SYstem ConsIsTs of a ControLled gRaMmar and a controlLing Grammar. GuIdeLiNes AnD consIStency RulEs aRe formaLized as COntRoLLINg Grammars that controL tHE UsEs of UML, i.E. the deRIvAtIOns using ThE grAmmaR OF UML. THis aPPrOach can bE impleMEnTeD as a parSeR, which CaN auTomAticaLLy veRify thE rules on A UML uSEr model in XMI foRMat. A comparisoN To RELaTEd woRk sHows our contRibuTIon: a GeneRIc Top-DOwn anD syntAx-BAsED approach that checks LaNguage Level Constraints at Compile-timE.'
AUThor:
- Zhe CHen
- GILlES Motet
bibliogrAphy:
-'mAin.bib'
titlE: 'a LanguagE-theoRetic VieW on GuidelINEs and ConSisTenCy RUleS OF UmL'
---
IntroductioN {#sEc:inTrO}
============
The UML (uniFied ModEliNg LAngUagE) iS a graphic Modeling LaNgUaGe DevElopeD By OMG (ObjEcT MaNaGemEnt GrOUp), and dEfineD by tHe SpECifIcationS [@uMlI] And [@UmLS]. UmL haS emErGed as The sOFtwAre induStry’s domiNanT ModeLiNg LanguagE for specifyinG, dEsigning anD dOcuMentinG THe artifaCts of systems [@MRR02][@BRJ05].
EvolvINg descrIptIons oF sofTware artiFacTs are fReqUEntly iNconsiStent, AnD toLERatinG THiS inCoNsistency iS IMpoRtant [@baL91][@EC01]. DIfferenT developers construCT anD update these dEscRiptIONs At dIFfERenT tIMes DURing development [@nER00], thus resUlTInG in inconsiSTenCiEs. They dEvelop mUltipLE views oN a system pRoviding pIeCes oF INfoRmation whiCh are redUndant or cOMplemENtAry. CoNstRaints ExIst On theSe piecES of InforMation WhOse vioLatioN lEads to inConsistent models. InconsiStency ProblEms Of | {widt h="160mm"}
{width="160 m m"}
{wid th = "160 m m" }
{w id t h= "160m m"}
{width="16 0 mm "}
---
ab str act: 'Guidel ine s andco nsi s tency ru les o f UMLa re use d to cont ro l the d e grees o f fr eedo m provided by the la n guage to preve nt fau lt s .G u ide lin es are use din sp e cific d o ma i n s (e . g., avionics) to recomme n d t he pro pe r u s e of t echno lo g ies . Consisten cy r ules areused t o deal w i th inco nsiste nci esin m o de ls . H ow e ver , g uid e lin es and c on si stenc y ru l e s useinf orma l res trictions onthe use s of lang uages , wh ic h mak es che cking d ifficult. In th is p aper, wecon si der t hesep roblem s f rom a lang uage-th e ore ti c v ie w. We propose thefo r m al ism of C -Syste m s, s h ort for“f orm al l a n guage con t ro l system s”. AC -S ys tem con si sts of a co ntr olled gram mar an d a cont rolli n g grammar. Gui d elines and co n si s t en c y ru les are formal ized as c ontr o ll ing gramm ars t ha t c o ntrol the uses of U ML , i.e. thederivations u sing the g r a m mar of U ML.T hi s approach canbe im plementeda s a pars er, w hich can automati c a lly veri fythe ru les o na UML user mo d e l in X MI form at. A comp ari son to re la ted workshows ou rco nt ri but ion:a generic t op- do wnand s y ntax-b asedappr oa ch tha t check s l a n guag ele velcon st raint s at com pile-ti me.'
auth or: - Zh eCh en
- Gi lles Motet
bi bl iography:-' mai n.bib' t itle: 'A Language-theoretic Vie w on Gui del inesandConsisten cyRulesofU ML'
-- -
Int roduc ti on{ # Sec:i n t ro }
= == =========T heUML ( Un ifie d Model ing Language) is a gra phic modeling la ngua g e d eve l op e d b yO MG( O bject Managemen t Group),an d d efined byt hesp ecifica tions [ @UMLi ] and [@ UMLs]. UM L has eme rg ed a s the softwareindustry ’s domina n t mod e li ng la ngu age fo rspe cifyi ng, de s ign ing a nd doc um enting thear tifactsof systems [@MRR02][@BR J05].
Evol vin g descrip tio n s o f softwar e ar tifacts ar e f req uentl y i n consi sten t ,and toler atin g this inc o ns ist e n cy is importa n t [@B al91] [@E C 01]. D iffe rent developers c o nstruct and up date t hes e d e scri pt ions at differ ent t i m es durin gdevelopment [@NER00 ], thusresult ing in incons i s te n cies.They de velop mul tip le views o nas ystemprov id ing pi eces o f inf o r mation which are redu n d ant o r co mplem en tary. C o nstr aints exis t on thesepieces ofinfor mationwh ose vi ola ti on leads t o inconsis tentmodels. I ncon sis tencyprob l e ms of | {width="160mm"}
{width="160mm"}
{width="160mm"}
{width="160mm"}
{width="160mm"}
---
abstract:_'Guidelines and_consistency rules of UML_are used_to_control the_degrees_of freedom provided_by the language_to prevent faults. Guidelines_are used in_specific_domains (e.g., avionics) to recommend the proper use of technologies. Consistency rules are used_to_deal with_inconsistencies_in_models. However, guidelines and consistency_rules use informal restrictions on_the uses_of languages, which makes checking difficult. In this_paper,_we consider these_problems from a language-theoretic view. We propose the formalism_of C-Systems, short for “formal language_control systems”. A_C-System_consists_of a controlled grammar_and a controlling grammar. Guidelines and_consistency rules are formalized as controlling_grammars that control the uses of UML,_i.e. the derivations using the grammar_of UML. This approach can_be implemented_as a parser, which can_automatically verify the_rules on_a UML user_model in XMI format. A comparison_to related work_shows our contribution: a generic top-down_and_syntax-based approach that_checks_language_level constraints_at compile-time.'
author:
- Zhe_Chen
-_Gilles Motet
bibliography:
-'main.bib'
title:_'A_Language-theoretic View on Guidelines and Consistency_Rules_of UML'
---
Introduction {#Sec:intro}
============
The UML (Unified Modeling Language)_is a graphic modeling_language_developed by OMG (Object_Management Group), and defined by_the specifications [@UMLi] and [@UMLs]. UML_has emerged_as the_software industry’s dominant modeling language for specifying, designing and documenting the_artifacts of systems [@MRR02][@BRJ05].
Evolving descriptions of_software artifacts are frequently_inconsistent, and_tolerating_this inconsistency is_important_[@Bal91][@EC01]. Different_developers construct and update these descriptions at_different times_during development [@NER00], thus resulting in_inconsistencies. They develop multiple_views_on a system providing pieces of_information which are redundant or complementary._Constraints exist on these pieces_of_information_whose violation leads to inconsistent_models. Inconsistency problems of |
- 'E. Kontizas'
- 'A. Vallenari'
date: 'Received date / accepted'
title: 'A semi-empirical library of galaxy spectra for Gaia classification based on SDSS data and PÉGASE models'
---
[This paper is the third in a series implementing a classification system for Gaia observations of unresolved galaxies. The system makes use of template galaxy spectra in order to determine spectral classes and estimate intrinsic astrophysical parameters. In previous work we used synthetic galaxy spectra produced by PÉGASE.2 code to simulate Gaia observations and to test the performance of Support Vector Machine (SVM) classifiers and parametrizers. Here we produce a semi-empirical library of galaxy spectra by fitting SDSS spectra with the previously produced synthetic libraries. We present (1) the semi-empirical library of galaxy spectra, (2) a comparison between the observed and synthetic spectra, and (3) first results of classification and parametrization experiments with simulated Gaia spectrophotometry of this library.]{} [We use $\chi^{2}$-fitting to fit SDSS galaxy spectra with the synthetic library in order to construct a semi-empirical library of galaxy spectra in which (1) the real spectra are extended by the synthetic ones in order to cover the full wavelength range of Gaia, and (2) astrophysical parameters are assigned to the SDSS spectra by the best fitting synthetic spectrum. The SVM models were trained with and applied to semi-empirical spectra. Tests were performed for the classification of spectral types and the estimation of the most significant galaxy parameters (in particular redshift, mass to light ratio and star formation history).]{} [We produce a semi-empirical library of 33670 galaxy spectra covering the wavelength range 250 to 1050nm at a sampling of 1nm or less. Using the results of the fitting of the SDSS spectra with our synthetic library, we investigate the range of the input model parameters that produces spectra which are in good agreement with observations. In general the results are very good for the majority of the synthetic spectra of early type, spiral and irregular galaxies, while they reveal problems in the models used to produce Quenched Star Forming Galaxies (QSFGs). The results of the SVM classification and regression models for this library are quite accurate for the prediction of the spectral type and the estimation of the redshift parameter, while they are quite poor for the cases of the most significant parameters used to produce the synthetic models (i.e. | -' E. Kontizas'
-' A. Vallenari'
date:' Received date / accepted'
title:' A semi - empirical library of galaxy spectra for Gaia categorization free-base on SDSS data and PÉGASE models'
---
[ This newspaper is the third in a serial implementing a classification system for Gaia notice of unresolved galaxies. The arrangement makes use of template galax spectra in order to determine spectral class and estimate intrinsic astrophysical parameters. In previous work we used man-made galaxy spectra produced by PÉGASE.2 code to model Gaia observations and to test the performance of Support Vector Machine (SVM) classifiers and parametrizers. Here we grow a semi - empirical library of galaxy spectra by fitting SDSS spectra with the previously produced synthetic libraries. We stage (1) the semi - empirical library of galaxy spectra, (2) a comparison between the observed and synthetic spectra, and (3) first result of classification and parametrization experiments with simulated Gaia spectrophotometry of this library. ] { } [ We use $ \chi^{2}$-fitting to fit SDSS galaxy spectra with the synthetic library in order to construct a semi - empiric library of galaxy spectra in which (1) the actual spectra are unfold by the celluloid ones in order to brood the full wavelength range of Gaia, and (2) astrophysical parameters are assigned to the SDSS spectrum by the best fitting synthetic spectrum. The SVM models were trained with and applied to semi - empirical spectrum. Tests were performed for the classification of spectral character and the appraisal of the most significant galaxy parameters (in particular red shift, mass to light ratio and star formation history). ] { } [ We produce a semi - empirical library of 33670 galaxy spectra cover the wavelength image 250 to 1050 nm at a sampling of 1 nanometer or less. Using the consequence of the fitting of the SDSS spectrum with our synthetic library, we investigate the stove of the input model parameters that produces spectra which are in good agreement with observation. In general the solution are very good for the majority of the synthetic spectra of early type, spiral and irregular galaxy, while they reveal problems in the model used to produce Quenched Star Forming Galaxies (QSFGs). The results of the SVM classification and regression models for this library are quite accurate for the prediction of the apparitional type and the estimation of the redshift parameter, while they are quite poor for the cases of the most significant argument practice to produce the synthetic models (i.e. | - 'E. Nontizas'
- 'A. Vallenari'
date: 'Received date / acceptxd'
title: 'A semi-eopirical library of galaxy s'ectea foe Gaia classification cased on DDSS datq anv PÉGASE models'
---
[Thma paper is ths thixd in a series ikplementinc a classificadiun system for Gaia observations of unwesolvec halaxies. The sistem iakea use of template galaxy spectra ih order to determine spectral classes and estilate intrinsic astrophjsical paraneteww. In previour work we lved synthejic galaxy spectra produced by PÉEASE.2 eode to simolztf Gaia obsertationf and to tesb the parformamce of Support Vertor Machine (SVM) classifixrs and parametrizerf. Here we pxoduce a semi-empiricao oibraty of galxzy rpedtca gy fithinj SDSS specfra with thw previously productd fjmthetic librzries. Re present (1) the semi-empirical library of calzxy spectra, (2) a comparisin between the observgd and synehetic spectra, and (3) first results of classificatimn anv oarcnetrixqtlon experiments with simulated Gaia spectrophjfokenry of this libravy.]{} [We use $\chi^{2}$-fittimg tp fit SDSS galxxy spzdtda with the synthehic libtary ib order tj comstruct a semi-empirical libeary of galaqy spectra in which (1) tke real speccra arg extemded by the synthetic oues in order to clver the rjll wavelength rxngv of Gaia, and (2) astrophysical pwrameters are assigndd tp the FDSS spectga by the best fitting sjnthejic spactrum. The SVM models were trained with aiv applied to xeki-ekpirical specbra. Tests were [erformed for jhe classnficatkon of spebtral typxs and the eftimation of dje most signmficant gwlaxt paeameterr (in particular redshift, mass to lught ratio and stav foroztion history).]{} [Wz kriduce a semi-emlirkcaj piurary mf 33670 galaxy s[ectfa zpverivg the waveodngtn range 250 to 1050nm at a vampming of 1nm or less. Ufing the results of the fittimg of the SDSS speftra xith onr synyhejic library, we investigate the dange of hhe input model parqmeters that produces spectra which are in good agrxement with observationw. In general the rexmlts are verb good for the kajority of the syntyetic spectra of tarly type, spiral and idreguldr gapaxies, while they reveal problems in the models used to produce Quenched Wtar Forming Galaxjes (ASFGs). Tke xesults of vhx SVM classificatpon and regression models for thms library axe quite accurate for the preciztion of the rpectral type and the esfimation of the redshift parameter, whilr they are quite poor fov the casws of the most rignificant parameyers uved ti produxe tme synthetic mudeks (i.e. | - 'E. Kontizas' - 'A. Vallenari' date: / title: 'A library of galaxy on data and PÉGASE --- [This paper the third in a series implementing classification system for Gaia observations of unresolved galaxies. The system makes use of galaxy spectra in order to determine spectral classes and estimate intrinsic astrophysical parameters. previous we synthetic spectra produced by PÉGASE.2 code to simulate Gaia observations and to test the performance of Support Machine (SVM) classifiers and parametrizers. Here we produce semi-empirical library of galaxy by fitting SDSS spectra with previously synthetic libraries. present the library of galaxy (2) a comparison between the observed and synthetic spectra, and (3) first results of classification and parametrization with simulated of this [We $\chi^{2}$-fitting fit SDSS galaxy the synthetic library in order to library of galaxy spectra in which (1) the spectra are by the synthetic ones in order cover the full wavelength range of Gaia, and astrophysical parameters are assigned to the SDSS spectra by the best fitting synthetic spectrum. The were trained with and to semi-empirical spectra. were for classification spectral types the estimation of the most significant galaxy parameters (in particular redshift, to light ratio and star formation history).]{} [We produce a of galaxy spectra covering wavelength range 250 to at sampling of 1nm or the of the spectra our synthetic library, we the range of the input parameters that produces spectra with observations. In general the results are very for the majority of the synthetic spectra early type, spiral and irregular galaxies, while they reveal problems in the used to Star Forming Galaxies (QSFGs). The results of the classification and regression models this library are quite accurate for the prediction of spectral and the of the redshift while they are poor for the the most parameters to models (i.e. | - 'E. Kontizas'
- 'A. Vallenari'
date: 'ReCeived date / AccepTed'
TitLe: 'a semI-empIrical library oF GalaXy spectra for Gaia classiFicatIoN BaseD On sDSS dAta and PégAse ModElS'
---
[THis PaPEr Is the ThiRd in a seRies implemEntInG a classificaTIoN system for gaiA observationS of UnresoLvEd gALaxieS. ThE systEm makeS Use of tEmplate gaLaXY spectRA in ordeR TO dEterMine spectral classES aND estimate intriNsic asTrOPhYSIcaL paRameters. In PrEviouS Work we uSEd SYNTheTIc galaxy spectRa produced bY pÉGaSE.2 codE tO siMUlate GAia obSeRVatIons and to teSt thE performaNce of SUPport VeCTor MachIne (SVM) ClaSsiFierS AnD pAraMeTRizERs. herE We pRoduce a sEmI-eMpiriCal lIBRARy of GalAxy sPectrA by fitting SDSs spEctrA WitH the pRevioUsly PrOduceD synthEtic lIbRaries. We present (1) The sEmi-empiriCal LiBraRy Of galAXy specTra, (2) A coMparisoN betweeN The ObSERVeD and synthetic spectRa, AND (3) fIrst resuLts of cLAsSiFIcation aNd ParAmetRIZatioN expERiMents witH simulATeD GAia specTrOphotoMeTry Of tHis liBRary.]{} [we use $\cHi^{2}$-fittinG to fiT sDSS galaxy specTRa with the syntHEtIC LiBRary In oRder to constRuct A Semi-EmpiRIcAl lIBrary Of galAxY SpECtra in which (1) the real sPeCtra arE exteNded by the syntHetic ones iN ORDer to covEr thE FuLL wavelength ranGe of GAia, and (2) astrOPhysical ParamEters are Assigned tO THe SDSS spEctRa bY thE beST FiTting synthetiC SPectRuM. The SVM ModEls were TraIneD wiTh aNd Applied to Semi-empiRiCaL sPeCtrA. TestS Were perfOrMed FoR thE clasSIficatIon of SpecTrAl TYpeS and the EStIMAtioN oF tHe moSt sIgNificAnt gALaxY parameTers (in parTicULar rEdShIft, mass To light ratio aNd Star formatIoN hiStory).]{} [WE PRoduce a sEmi-empirical library of 33670 gaLAxy specTra CoverIng tHe wavelenGth Range 250 tO 1050nm AT a sampLing of 1Nm or lEsS. UsING the rESUlTs oF tHe fitting oF THe SdSS spEcTra wIth our sYnthetic library, we iNVesTigate the rangE of The iNPUt ModEL pARamEtERs tHAT produces spectrA which are iN gOOd Agreement wITh oBsErvatioNs. In genEral tHE resultS are very gOod for the MaJoriTY Of tHe synthetiC spectra Of early tyPE, spirAL aNd irrEguLar galAxIes, While They reVEal ProblEms in tHe Models Used tO pRoduce QuEnched Star Forming GalaxiEs (QSFGS). The rEsuLts of the SvM cLAssIfication And rEgression mOdeLs fOr thiS liBRary aRe quITe AccURate fOr thE PredictioN Of The SPEcTral type and THE EstImatiOn oF The redShifT parameter, while thEY are quite poor fOr thE CAseS of THe moSt Significant parAmeTeRS Used to prOdUce the synthEtic modeLs (I.E. | - 'E. Kontizas'
- 'A. Vall enari'
dat e: 'R ece ive ddate / a ccepted'
title : 'Asemi-empirical library of g al a xy s p ec tra f or Gaia cl a s sif ic at ion b a se d onSDS S dataand PÉGASE mo de ls'
---
[Th i spaper is t hethird in a s eri es imp le men t ing a cl assif icatio n syste m for Gai ao bserva t ions of u nr esol ved galaxies. The sy s tem makes useof tem pl a te g ala xyspectra in o rdert o deter m in e s pec t ral classes a nd estimate int rinsic a str o physic al pa ra m ete rs. In prev ious work weused s y nthetic galaxyspectr a p rod uced by P ÉGA SE . 2 c o de to sim ulate Ga ia o bserv atio n s a nd t o t estthe p erformance of Su ppor t Ve ctorMachi ne ( SV M) cl assifi ers a nd parametrizers. Her e we prod uce a se mi -empi r ical l ibr ary of gal axy spe c tra b y f it ting SDSS spectrawi t h t he previ ouslyp ro du c ed synth et iclibr a r ies.We p r es ent (1)the se m i- em pirical l ibrary o f g ala xy sp e ctra , (2)a compar isonb etween the obs e rved and synt h et i c s p ectr a,and (3) fir st r e sult s of cl ass i ficat ion a nd pa r ametrization experi me nts wi th si mulated Gaiaspectropho t o m etry ofthis li b rary.]{} [We u se $\ chi^{2}$-f i tting to fitSDSS gal axy spect r a with th e s ynt het icl i br ary in ordert o con st ruct asem i-empir ica l l ibr ary o f galaxyspectrain w hi ch (1 ) the real spe ct raar e e xtend e d by t he sy nthe ti co nes in ord e rt o cov er t he f ull w avele ngth ran ge of G aia, and(2) astr op hy sical p arameters are a ssigned to t heSDSS s p e ctra bythe best fitting synthe t ic spec tru m. Th e SV M modelswer e trai ned with a nd app liedto se m i -empi r i ca l s pe ctra. Test s wer e per fo rmed for th e classification o f sp ectral typesand the e st ima t io n of t h e m o s t significant g alaxy para me t er s (in part i cul ar redshi ft, mas s tol ight ra tio and s tar forma ti on h i s tor y).]{} [We produce a semi-e m piric a llibra ryof 336 70 ga laxyspectr a co verin g thewa veleng th ra ng e 250 to 1050nm at a sampling o f 1nmor le ss. Using th e r e sul ts of the fit ting of th e S DSS spec tra withours yn the t ic li brar y , we inve s ti gat e th e range oft h e in put m ode l param eter s that produces s p ectra which ar e in g ood ag r eeme nt with observat ion s. I n genera lthe results are ver yg ood f or the major ity oft h es ynthet ic s pec tra of ea rly t y pe, spi ra la nd irr egul ar galax ies, w h ilet h ey reveal proble ms in t he mo d els used t o produ c e Qu enched Sta r Forming G alaxie s (Q SFGs) . The r es ults o f t he SVM class i ficationand r egressi on mod els for t hisl i brary are q uit e accurat e fo r t he pre dict ion o fthespectralt ype andthe estimat io n o f the re d sh i f t paramete r,while t hey are qu i te p o or for t he cas es ofthe mos t si gn ificant pa r a meters us ed to pro d uce t he s ynthetic m odel s(i. e. | - 'E._Kontizas'
- 'A._Vallenari'
date: 'Received date /_accepted'
title: 'A_semi-empirical_library of_galaxy_spectra for Gaia_classification based on_SDSS data and PÉGASE_models'
---
[This paper is_the_third in a series implementing a classification system for Gaia observations of unresolved galaxies._The_system makes_use_of_template galaxy spectra in order_to determine spectral classes and_estimate intrinsic_astrophysical parameters. In previous work we used synthetic_galaxy_spectra produced by_PÉGASE.2 code to simulate Gaia observations and to test_the performance of Support Vector Machine_(SVM) classifiers and_parametrizers._Here_we produce a semi-empirical_library of galaxy spectra by fitting_SDSS spectra with the previously produced_synthetic libraries. We present (1) the semi-empirical_library of galaxy spectra, (2) a_comparison between the observed and_synthetic spectra,_and (3) first results of_classification and parametrization_experiments with_simulated Gaia spectrophotometry_of this library.]{} [We use $\chi^{2}$-fitting_to fit SDSS_galaxy spectra with the synthetic library_in_order to construct_a_semi-empirical_library of_galaxy spectra in_which_(1) the_real_spectra are extended by the synthetic_ones_in order to cover the full wavelength_range of Gaia, and_(2)_astrophysical parameters are assigned_to the SDSS spectra by_the best fitting synthetic spectrum. The_SVM models_were trained_with and applied to semi-empirical spectra. Tests were performed for the_classification of spectral types and the_estimation of the most_significant galaxy_parameters_(in particular redshift,_mass_to light_ratio and star formation history).]{} [We produce_a semi-empirical_library of 33670 galaxy spectra covering_the wavelength range 250_to_1050nm at a sampling of 1nm_or less. Using the results of_the fitting of the SDSS_spectra_with_our synthetic library, we investigate_the range of the input model_parameters that produces_spectra which are in good agreement with_observations._In general the results are very_good_for the majority of the synthetic_spectra_of_early type, spiral and irregular_galaxies, while they reveal problems in_the models used to produce Quenched Star Forming Galaxies_(QSFGs). The results_of the SVM classification and_regression_models_for this library are quite accurate for the prediction of_the spectral_type and the_estimation of the redshift parameter, while they are quite poor_for the cases of the most significant_parameters used to produce the synthetic models (i.e. |
.399249 0.399850 0.399970 0.400000 2.996339E-5
$0.5$ 0.494075 0.498826 0.499765 0.499953 0.500000 4.681780E-5
$0.6$ 0.591468 0.598310 0.599662 0.599932 0.600000 6.741763E-5
$0.7$ 0.688388 0.697700 0.699541 0.699908 0.700000 9.176289E-5
$0.8$ 0.784833 0.796996 0.799400 0.799880 0.800000 1.198535E-4
$0.9$ 0.880804 0.896198 0.899241 0.899848 0.900000 1.516896E-4
$1.0$ 0.976302 0.995307 0.999063 0.999812 1.000000 1.872712E-4
------- ---------- ---------- ---------- ---------- ---------------- ----------------
: Numerical results of Example 3.1. for different $u$ values when $\protect\alpha =1$
{width="3.50in"}
Consider the following system of nonlinear fractional-integro differential equations [@39]:
$$\begin{aligned}
\frac{d^{\alpha _{1}}u(t)}{{dt}^{\alpha _{1}}} &=&1-\frac{1}{2}{\left( k^{^{{\prime }}}\left( t\right) \right) }^{2}+\int_{0}^{t}{\left( \left(
t-s\right) k\left( s\right) +u\left( s\right) k\left( s\right) \right) ds}
\notag \\
\frac{d^{\alpha _{2}}k(t)}{{dt}^{2 | .399249 0.399850 0.399970 0.400000 2.996339E-5
$ 0.5 $ 0.494075 0.498826 0.499765 0.499953 0.500000 4.681780E-5
$ 0.6 $ 0.591468 0.598310 0.599662 0.599932 0.600000 6.741763E-5
$ 0.7 $ 0.688388 0.697700 0.699541 0.699908 0.700000 9.176289E-5
$ 0.8 $ 0.784833 0.796996 0.799400 0.799880 0.800000 1.198535E-4
$ 0.9 $ 0.880804 0.896198 0.899241 0.899848 0.900000 1.516896E-4
$ 1.0 $ 0.976302 0.995307 0.999063 0.999812 1.000000 1.872712E-4
------- ---------- ---------- ---------- ---------- ---------------- ----------------
: Numerical results of Example 3.1. for different $ u$ values when $ \protect\alpha = 1 $
! [ Comparison of the PIA solution $ u_{3}(t)$ and accurate solution for Example 3.1. when $ \protect\alpha = 1$](Figure1.pdf){width="3.50 in " }
think the following system of nonlinear fractional - integro differential equation [ @39 ]:
$ $ \begin{aligned }
\frac{d^{\alpha _ { 1}}u(t)}{{dt}^{\alpha _ { 1 } } } & = & 1-\frac{1}{2}{\left (k^{^{{\prime } } } \left (t\right) \right) } ^{2}+\int_{0}^{t}{\left (\left (
deoxythymidine monophosphate - s\right) k\left (s\right) + u\left (s\right) k\left (s\right) \right) ds }
\notag \\
\frac{d^{\alpha _ { 2}}k(t)}{{dt}^{2 | .399249 0.399850 0.399970 0.400000 2.996339E-5
$0.5$ 0.494075 0.498826 0.499765 0.499953 0.500000 4.681780E-5
$0.6$ 0.591468 0.598310 0.599662 0.599932 0.600000 6.741763E-5
$0.7$ 0.688388 0.697700 0.699541 0.699908 0.700000 9.176289E-5
$0.8$ 0.784833 0.796996 0.799400 0.799880 0.800000 1.198535E-4
$0.9$ 0.880804 0.896198 0.899241 0.899848 0.900000 1.516896E-4
$1.0$ 0.976302 0.995307 0.999063 0.999812 1.000000 1.872712E-4
------- ---------- ---------- ---------- ---------- ---------------- ----------------
: Numerhcxl results of Example 3.1. for different $t$ valuex ahen $\protect\alkha =1$
{wldth="3.50in"}
Consider the following wystqn of nonlinexr fractional-integro djfferential equations [@39]:
$$\begin{aligved}
\frcc{d^{\alpha _{1}}u(t)}{{et}^{\qlpjd _{1}}} &=&1-\frac{1}{2}{\left( k^{^{{\priie }}}\left( t\rigmn) \right) }^{2}+\int_{0}^{t}{\legt( \left(
t-s\right) k\lxft( w\right) +u\left( s\right) k\neft( s\right) \right) qs}
\notag \\
\xrcc{d^{\alpha _{2}}k(t)}{{dt}^{2 | .399249 0.399850 0.399970 0.400000 2.996339E-5 $0.5$ 0.494075 0.499953 4.681780E-5 $0.6$ 0.598310 0.599662 0.599932 0.699541 0.700000 9.176289E-5 $0.8$ 0.796996 0.799400 0.799880 1.198535E-4 $0.9$ 0.880804 0.896198 0.899241 0.899848 1.516896E-4 $1.0$ 0.976302 0.995307 0.999063 0.999812 1.000000 1.872712E-4 ------- ---------- ---------- ---------- ---------- ---------------- : Numerical results of Example 3.1. for different $u$ values when $\protect\alpha {width="3.50in"} Consider the following system of nonlinear fractional-integro equations [@39]: $$\begin{aligned} \frac{d^{\alpha _{1}}u(t)}{{dt}^{\alpha _{1}}} &=&1-\frac{1}{2}{\left( k^{^{{\prime t\right) \right) }^{2}+\int_{0}^{t}{\left( \left( k\left( s\right) +u\left( s\right) k\left( \right) \notag \\ _{2}}k(t)}{{dt}^{2 | .399249 0.399850 0.399970 0.400000 2.996339E-5
$0.5$ 0.494075 0.498826 0.499765 0.499953 0.500000 4.681780E-5
$0.6$ 0.591468 0.598310 0.599662 0.599932 0.600000 6.741763E-5
$0.7$ 0.688388 0.697700 0.699541 0.699908 0.700000 9.176289E-5
$0.8$ 0.784833 0.796996 0.799400 0.799880 0.800000 1.198535E-4
$0.9$ 0.880804 0.896198 0.899241 0.899848 0.900000 1.516896E-4
$1.0$ 0.976302 0.995307 0.999063 0.999812 1.000000 1.872712E-4
------- ---------- ---------- ---------- ---------- ---------------- ----------------
: Numerical results of example 3.1. for DiffeRenT $u$ vAlUes wHen $\pRotect\alpha =1$
{widTh="3.50iN"}
ConsidEr the folloWinG sYstem of nonliNEaR fractionaL-inTegro differeNtiAl equaTiOns [@39]:
$$\BEgin{aLigNed}
\frAc{d^{\alpHA _{1}}u(t)}{{dt}^{\aLpha _{1}}} &=&1-\frac{1}{2}{\lEfT( K^{^{{\prime }}}\LEft( t\rigHT) \RiGht) }^{2}+\iNt_{0}^{t}{\left( \left(
t-s\righT) K\lEFt( s\right) +u\left( s\Right) k\LeFT( s\RIGht) \RigHt) ds}
\notag \\
\fRaC{d^{\alpHA _{2}}k(t)}{{dt}^{2 | .399249 0.399850 0.399 970 0 .4000 00 2.9 9633 9E-5
$0.5$ 0.4 94075 0.498826 0.4 99765 0.49 9 95 3 0.500 0 00 4. 681 78 0 E- 5
$0. 6$ 0. 591468 0 .59 83 10 0.59966 2 0.599932 0.600000 6.74 17 63E - 5
$0. 7$ 0.6883 8 8 0. 697700 0. 6 99541 0.6999 0 8 0.700000 9. 1 76 2 89E-5
$0.8$ 0.7 84 8 33 0. 796 996 0.79 94 00 0 .799880 0.8 0 0000 1. 198535E-4
$0 .9$ 0. 880 8 04 0 .8961 98 0 .899241 0 .899 848 0.9000 0 0 1.51689 6E-4
$ 1.0 $ 0 .9 76 302 0.9 9 53 07 0. 999063 0 .9 99812 1 . 0000 00 1.8 72712E-4
-- --- -- - - --- ----- ---- ---- -- ---- ------ ---- -- ---- ---------- ---- -- ------ --- -- --- --
: Numeri cal re sults o f Examp l e 3 .1 . f or different $u$ val ue s wh en $\pro tect\a l ph a= 1$
{width="3 . 50 i n "}
Con sid er the foll owin g sys temo fnon l inear frac ti o na l -integro differenti al equat ions[@39]:
$$\be gin{aligne d } \frac{d^ {\al p ha _{1}}u(t)}{{dt }^{\a lpha _{1}} } &=&1-\f rac{1 }{2}{\le ft( k^{^{ { \ prime }} }\l eft ( t \ri g h t) \right) }^{2 } + \int _{ 0}^{t}{ \le ft( \le ft(
t- s\r igh t) k\left(s\right) + u\ le ft ( s \righ t ) k\left (s\r ig ht) \rig h t) ds}
\no tag\\
\ f rac {d^{\al p ha _ {2}} k( t) }{{d t}^ {2 | .399249 _ 0.399850_ 0.399970 _ __ 0.400000__ _ 2.996339E-5
_ $0.5$ _ 0.494075 __0.498826 0.499765 0.499953 0.500000 __ ___4.681780E-5
$0.6$ _ 0.591468 0.598310_ _0.599662 0.599932 __0.600000 _ 6.741763E-5
$0.7$ _ 0.688388 0.697700 _ 0.699541 __0.699908_ _ 0.700000 _ 9.176289E-5
$0.8$_ 0.784833 0.796996 _ 0.799400 0.799880 _ 0.800000_ _ 1.198535E-4
_ $0.9$_ _0.880804 _0.896198 0.899241 _0.899848 _ 0.900000 __ _1.516896E-4
__ $1.0$_ 0.976302__ 0.995307__ 0.999063 0.999812 __ 1.000000 _ 1.872712E-4
__------- ---------- ---------- ----------_---------- ---------------- ----------------
:_Numerical results of Example 3.1. for_different $u$_values when_$\protect\alpha =1$
{width="3.50in"}
Consider the_following system of nonlinear_fractional-integro differential_equations_[@39]:
$$\begin{aligned}
\frac{d^{\alpha _{1}}u(t)}{{dt}^{\alpha _{1}}}_&=&1-\frac{1}{2}{\left(_k^{^{{\prime }}}\left(_t\right) \right) }^{2}+\int_{0}^{t}{\left( \left(
t-s\right) k\left( s\right) +u\left(_s\right) k\left(_s\right) \right) ds}
\notag \\
\frac{d^{\alpha _{2}}k(t)}{{dt}^{2 |
given by $$\begin{aligned}
e^{-\lambda}(2\nu^{\prime\prime} + \nu^{\prime^2} - \lambda^{\prime}\nu^{\prime} - 2\frac{\nu^{\prime}}{r}) \nonumber \\
- 2(n-3)(\frac{e^{-\lambda}\lambda^{\prime}}{r}
+ 2\frac{e^{-\lambda}}{r^2} - \frac{2}{r^2}) = 0 \label{isotropy}. \end{aligned}$$ Let us rewrite this equation in a form that readily yields the universal character of the Schwarzschild interior solution for all $n\geq4$, $$\begin{aligned}
e^{-\lambda}(2\nu^{\prime\prime} + \nu^{\prime^2} - \lambda^{\prime}\nu^{\prime} - 2\frac{\nu^{\prime} + \lambda^{\prime}}{r} - \frac{4}{r^2}) + \frac{4}{r^2} \nonumber \\
- 2(n-4) \Bigl((n-1)(\frac{e^{-\lambda}}{r^2} - \frac{1}{r^2}) + \frac{2\rho}{n-2} \Bigr) = 0. \label{iso}\end{aligned}$$ We now set the coefficient of $(n-4)$ to zero so that the equation remains the same for all $n\geq4$. This then straightway determines $e^{-\lambda}$ without integration and it is given by $$e^{-\lambda} = 1 - \rho_0 r^2 \label{sol}$$ where $\rho_0=2\rho/{(n-1)(n-2)}$. This when put in Eq. (\[density\]) implies constant density. We thus obtain $\rho=const.$ as the neceessary condition for universality of the isotropy equation for all $n\geq4$. The sufficiency of constant density is obvious from the integration of Eq. (\[density\]) for $\rho=const$, giving the same solution as above where a constant of integration is set to zero for regularity at the center. Thus constant density is a necessary and sufficient condition for universality of field inside a fluid sphere, i.e. independent of spacetime dimension. An alternative identification of constant density is that the gravitational field inside a fluid sphere is independent of spacetime dimension $\geq4$. This universal property is therefore true if and only if density | given by $ $ \begin{aligned }
e^{-\lambda}(2\nu^{\prime\prime } + \nu^{\prime^2 } - \lambda^{\prime}\nu^{\prime } - 2\frac{\nu^{\prime}}{r }) \nonumber \\
- 2(n-3)(\frac{e^{-\lambda}\lambda^{\prime}}{r }
+ 2\frac{e^{-\lambda}}{r^2 } - \frac{2}{r^2 }) = 0 \label{isotropy }. \end{aligned}$$ Let us rewrite this equation in a form that promptly give the universal character of the Schwarzschild home solution for all $ n\geq4 $, $ $ \begin{aligned }
e^{-\lambda}(2\nu^{\prime\prime } + \nu^{\prime^2 } - \lambda^{\prime}\nu^{\prime } - 2\frac{\nu^{\prime } + \lambda^{\prime}}{r } - \frac{4}{r^2 }) + \frac{4}{r^2 } \nonumber \\
- 2(n-4) \Bigl((n-1)(\frac{e^{-\lambda}}{r^2 } - \frac{1}{r^2 }) + \frac{2\rho}{n-2 } \Bigr) = 0. \label{iso}\end{aligned}$$ We immediately set the coefficient of $ (n-4)$ to zero so that the equation persist the like for all $ n\geq4$. This then straightway determines $ e^{-\lambda}$ without integration and it is given by $ $ e^{-\lambda } = 1 - \rho_0 r^2 \label{sol}$$ where $ \rho_0=2\rho/{(n-1)(n-2)}$. This when put in Eq. (\[density\ ]) implies ceaseless density. We thus prevail $ \rho = const.$ as the neceessary circumstance for universality of the isotropy equation for all $ n\geq4$. The sufficiency of changeless density is obvious from the integration of Eq. (\[density\ ]) for $ \rho = const$, contribute the same solution as above where a constant of integration is set to zero for regularity at the plaza. Thus constant density is a necessary and sufficient circumstance for universality of field inside a fluent sphere, i.e. independent of spacetime dimension. An alternative recognition of constant density is that the gravitational field inside a fluid sphere is autonomous of spacetime property $ \geq4$. This universal property is therefore true if and only if density | gigen by $$\begin{aligned}
e^{-\lambaa}(2\nu^{\prime\prime} + \nu^{\primx^2} - \lambsa^{\prime}\nj^{\prime} - 2\frac{\nu^{\prime}}{r}) \nonumbec \\
- 2(n-3)(\drac{e^{-\oambda}\lambda^{\prime}}{r}
+ 2\frxc{e^{-\lambda}}{g^2} - \frac{2}{r^2}) = 0 \lebel{isotropy}. \end{emigned}$$ Let us vewrice this equation in a form that readily fidlbs the universal character of the Scrwarzscnipd interior sojutipg fod all $n\geq4$, $$\begin{aligned}
e^{-\lambda}(2\nu^{\prjme\primt} + \nu^{\prime^2} - \lambda^{\lrime}\nu^{\prime} - 2\frac{\nu^{\prime} + \lamhda^{\prime}}{r} - \frac{4}{r^2}) + \vrac{4}{r^2} \nonumver \\
- 2(n-4) \Bigl((n-1)(\frac{d^{-\lambda}}{r^2} - \frac{1}{r^2}) + \frac{2\rgo}{n-2} \Bigr) = 0. \label{iso}\end{aligned}$$ Wd now set the ciedfifhent of $(n-4)$ ti zerj so that thc equathon remsins the same nor anl $b\geq4$. This then straigitway determines $e^{-\laibda}$ withmuc integration and it us givet by $$e^{-\laovda} = 1 - \rio_0 d^2 \labep{som}$$ where $\rhk_0=2\rho/{(n-1)(n-2)}$. This when put in Eq. (\[denxiej\]) implies conatant qegsity. We thus obtain $\rho=const.$ as the nebeesaary condition for univwrsality of the isotrlpy equatyon for all $n\geq4$. The sufficiency of constant denshty ia obrljjw vrom the integration of Eq. (\[density\]) for $\rho=conff$, bining the same solmtion as above whete a sonstant of ivtegracjoh is set to zero flr regujaritt at the sentrr. Thus constant density is a necessary qnd sufficient conbition for uuiversslity of field inside a fluib sphede, i.e. indepfndent of rpacetime dimenskon. At alternative identificatijn of conwtanc densith is that ehe gravitwtional field inside a fpuid dpvere is infependent of spacetime dimensioi $\geq4$. This unovarsdl propexty is therefore trte if and only if denfity | given by $$\begin{aligned} e^{-\lambda}(2\nu^{\prime\prime} + \nu^{\prime^2} - 2\frac{\nu^{\prime}}{r}) \\ - + 2\frac{e^{-\lambda}}{r^2} - Let rewrite this equation a form that yields the universal character of the interior solution for all $n\geq4$, $$\begin{aligned} e^{-\lambda}(2\nu^{\prime\prime} + \nu^{\prime^2} - \lambda^{\prime}\nu^{\prime} - 2\frac{\nu^{\prime} \lambda^{\prime}}{r} - \frac{4}{r^2}) + \frac{4}{r^2} \nonumber \\ - 2(n-4) \Bigl((n-1)(\frac{e^{-\lambda}}{r^2} - \frac{1}{r^2}) + \Bigr) 0. We set the coefficient of $(n-4)$ to zero so that the equation remains the same for all This then straightway determines $e^{-\lambda}$ without integration and is given by $$e^{-\lambda} 1 - \rho_0 r^2 \label{sol}$$ $\rho_0=2\rho/{(n-1)(n-2)}$. when put Eq. implies density. We thus $\rho=const.$ as the neceessary condition for universality of the isotropy equation for all $n\geq4$. The sufficiency of density is the integration Eq. for giving the same above where a constant of integration zero for regularity at the center. Thus constant is a and sufficient condition for universality of inside a fluid sphere, i.e. independent of spacetime An alternative identification of constant density is that the gravitational field inside a fluid sphere of spacetime dimension $\geq4$. universal property is true and if | given by $$\begin{aligned}
e^{-\lambdA}(2\nu^{\prime\prIme} + \nu^{\PriMe^2} - \lAmBda^{\pRime}\Nu^{\prime} - 2\frac{\nu^{\pRIme}}{r}) \Nonumber \\
- 2(n-3)(\frac{e^{-\lambda}\laMbda^{\pRiME}}{r}
+ 2\frAC{e^{-\LambdA}}{r^2} - \frac{2}{r^2}) = 0 \LAbEL{IsoTrOpY}. \enD{aLIgNed}$$ LeT us Rewrite This equatiOn iN a Form that readILy Yields the uNivErsal charactEr oF the ScHwArzSChild IntErior SolutiON for alL $n\geq4$, $$\begiN{aLIgned}
e^{-\LAmbda}(2\nu^{\PRImE\priMe} + \nu^{\prime^2} - \lambda^{\prIMe}\NU^{\prime} - 2\frac{\nu^{\prIme} + \lamBdA^{\PrIME}}{r} - \fRac{4}{R^2}) + \frac{4}{r^2} \nonuMbEr \\
- 2(n-4) \BiGL((n-1)(\frac{e^{-\LAmBDA}}{R^2} - \frAC{1}{r^2}) + \frac{2\rho}{n-2} \BigR) = 0. \label{iso}\enD{AliGned}$$ We NoW seT The coeFficiEnT Of $(n-4)$ To zero so thaT the Equation rEmains THe same fOR all $n\geQ4$. This tHen StrAighTWaY dEteRmINes $E^{-\LaMbdA}$ WitHout inteGrAtIon anD it iS GIVEn by $$E^{-\laMbda} = 1 - \Rho_0 r^2 \lAbel{sol}$$ where $\rHo_0=2\rHo/{(n-1)(n-2)}$. tHis When pUt in EQ. (\[denSiTy\]) impLies coNstanT dEnsity. We thus obtAin $\rHo=const.$ as The NeCeeSsAry coNDition For UniVersaliTy of the ISotRoPY EQuAtion for all $n\geq4$. The SuFFIcIency of cOnstanT DeNsITy is obviOuS frOm thE INtegrAtioN Of eq. (\[densitY\]) for $\rhO=CoNsT$, giving ThE same sOlUtiOn aS abovE WherE a consTant of inTegraTIon is set to zero FOr regularity aT ThE CEnTEr. ThUs cOnstant densIty iS A necEssaRY aNd sUFficiEnt coNdITiON for universality of fIeLd insiDe a flUid sphere, i.e. inDependent oF SPAcetime dImenSIoN. an alternative iDentiFication of COnstant dEnsitY is that tHe gravitaTIOnal fielD inSidE a fLuiD SPhEre is independENT of sPaCetime dImeNsion $\geQ4$. ThIs uNivErsAl Property iS therefoRe TrUe If And Only iF Density | given by $$\begin{aligned }
e^{-\lam bda}( 2\n u^{ \p rime \pri me} + \nu^{\pr i me^2 } - \lambda^{\prime}\n u^{\p ri m e} - 2\ frac{ \nu^{\p r im e } }{r }) \ non um b er \\
- 2( n-3)(\f rac{e^{-\l amb da }\lambda^{\p r im e}}{r}
+2\f rac{e^{-\lam bda }}{r^2 }- \ f rac{2 }{r ^2})= 0 \ l abel{i sotropy}. \ e nd{ali g ned}$$L e tus r ewrite this equat i on in a form that readi ly yi e l dsthe universal c harac t er of t h eS c h war z schild interi or solution for all $ n\ geq 4 $, $$\ begin {a l ign ed}
e^{-\la mbda }(2\nu^{\ prime\ p rime} + \nu^{\p rime^2 } - \l ambd a ^{ \p rim e} \ nu^ { \p rim e } - 2\frac{ \n u^ {\pri me}+ \ l ambd a^{ \pri me}}{ r} - \frac{4} {r^ 2})+ \f rac{4 }{r^2 } \n on umber \\
- 2(n- 4) \Bigl((n-1)(\f rac{ e^{-\lamb da} }{ r^2 }- \fr a c{1}{r ^2} ) + \frac{ 2\rho}{ n -2} \ B i g r) = 0. \label{iso}\ en d { al igned}$$ We no w s et the coef fi cie nt o f $(n-4 )$ t o z ero so t hat th e e qu ation r em ains t he sa mefor a l l $n \geq4$ . This t hen s t raightway dete r mines $e^{-\l a mb d a }$ with out integratio n an d itis g i ve n b y $$e^ {-\la mb d a} = 1 - \rho_0 r^2 \l ab el{sol }$$ w here $\rho_0= 2\rho/{(n- 1 ) ( n-2)}$.This wh e n put in Eq. ( \[den sity\]) im p lies con stant density . We thus o btain $\ rho =co nst .$a s t he neceessary c ondi ti on foruni versali tyofthe is ot ropy equa tion for a ll $ n\ geq 4$. T h e suffic ie ncy o f c onsta n t dens ity i s ob vi ou s fr om thei nt e g rati on o f Eq . ( \[ densi ty\] ) fo r $\rho =const$,giv i ng t he s ame sol ution as abov ewhere a co ns tan t of i n t egration is set to zero for reg u larityatthe c ente r. Thus c ons tant d ens i ty isa nece ssary a nds u ffici e n tcon di tion for u n i ver salit yof f ield in side a fluid spher e , i .e. independe ntof s p a ce tim e d i men si o n.A n alternative id entificati on of constantd ens it y is th at thegravi t ational field in side a fl ui d sp h e reis indepen dent ofspacetime dimen s io n $\g eq4 $. Thi suni versa l prop e rty is t herefo re trueif an donly ifdensity | given_by $$\begin{aligned}
e^{-\lambda}(2\nu^{\prime\prime}_+ \nu^{\prime^2} - \lambda^{\prime}\nu^{\prime}_- 2\frac{\nu^{\prime}}{r})_\nonumber_\\
- 2(n-3)(\frac{e^{-\lambda}\lambda^{\prime}}{r}_
+_2\frac{e^{-\lambda}}{r^2} - \frac{2}{r^2})_= 0 _\label{isotropy}. \end{aligned}$$ Let us_rewrite this equation_in_a form that readily yields the universal character of the Schwarzschild interior solution for_all_$n\geq4$, $$\begin{aligned}
e^{-\lambda}(2\nu^{\prime\prime}_+_\nu^{\prime^2}_- \lambda^{\prime}\nu^{\prime} - 2\frac{\nu^{\prime} +_\lambda^{\prime}}{r} - \frac{4}{r^2}) + \frac{4}{r^2}_\nonumber \\_
- 2(n-4) \Bigl((n-1)(\frac{e^{-\lambda}}{r^2} - \frac{1}{r^2}) + \frac{2\rho}{n-2} \Bigr)_=_0. \label{iso}\end{aligned}$$ We_now set the coefficient of $(n-4)$ to zero so_that the equation remains the same_for all $n\geq4$._This_then_straightway determines $e^{-\lambda}$ without_integration and it is given by_$$e^{-\lambda} = 1 - \rho_0 r^2_\label{sol}$$ where $\rho_0=2\rho/{(n-1)(n-2)}$. This when put in_Eq. (\[density\]) implies constant density. We_thus obtain $\rho=const.$ as the_neceessary condition_for universality of the isotropy_equation for all_$n\geq4$. The_sufficiency of constant_density is obvious from the integration_of Eq. (\[density\])_for $\rho=const$, giving the same solution_as_above where a_constant_of_integration is_set to zero_for_regularity at_the_center. Thus constant density is a_necessary_and sufficient condition for universality of field_inside a fluid sphere,_i.e._independent of spacetime dimension._An alternative identification of constant_density is that the gravitational field_inside a_fluid sphere_is independent of spacetime dimension $\geq4$. This universal property is therefore_true if and only if density |
infty(A,\Phi))\|_{W^{-1, p}_{A_1}(X)} \quad\text{(by \eqref{eq:gauged_LS})}
\\
&\leq Z'C_1\|\hat\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.\end{aligned}$$ But $\hat\sM(A,\Phi) = \Pi_{u_\infty(A_\infty, \Phi_\infty)}\sM$ and because the projection, $$\Pi_{u_\infty(A_\infty, \Phi_\infty)}:W^{k, p}_{A_1}(X; \Lambda^1 \otimes \ad P \oplus E) \to W^{k, p}_{A_1}(X; \Lambda^1 \otimes \ad P \oplus E),$$ is bounded with norm one (for any $k\in\ZZ$ and $1<p<\infty$), then $$\|\hat\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)} \leq \|\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.$$ Hence, the [Ł]{}ojasiewicz–Simon gradient inequality holds for the pairs $(A, \Phi)$ and $(A_\infty, \Phi_\infty)$ with constants $(Z, \theta, \sigma)$, where $Z := C_1Z'$.
\[case:APhi\_notin\_Coulomb\_gauge\_wrt\_APhi\_infty\] Let $\zeta= \zeta(A_1, A_\infty, \Phi_\infty, g, G,p,q)\in (0,1]$ and $N = N(A_1,A_\infty, \Phi_\infty, g, G,p,q) \in [1,\infty)$ denote the constants in Theorem \[mainthm:Feehan\_proposition\_3-4-4\_Lp\_pairs\] and choose $\zeta_1 \in (0,\zeta]$ small enough that $2N\zeta_1 < \sigma_1$, where we now use $\sigma_1$ to denote the [Ł]{}ojasiewicz–Simon constant from Case \[case:APhi\_Coulomb\_gauge\_wrt\_APhi\_infty\]. If $(A,\Phi)$ obeys $$\|(A,\Phi | infty(A,\Phi))\|_{W^{-1, p}_{A_1}(X) } \quad\text{(by \eqref{eq: gauged_LS }) }
\\
& \leq Z'C_1\|\hat\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.\end{aligned}$$ But $ \hat\sM(A,\Phi) = \Pi_{u_\infty(A_\infty, \Phi_\infty)}\sM$ and because the projection, $ $ \Pi_{u_\infty(A_\infty, \Phi_\infty)}:W^{k, p}_{A_1}(X; \Lambda^1 \otimes \ad P \oplus E) \to W^{k, p}_{A_1}(X; \Lambda^1 \otimes \ad P \oplus E),$$ is bounded with norm one (for any $ k\in\ZZ$ and $ 1 < p<\infty$), then $ $ \|\hat\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X) } \leq \|\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.$$ Hence, the [ Ł]{}ojasiewicz – Simon gradient inequality reserve for the pair $ (A, \Phi)$ and $ (A_\infty, \Phi_\infty)$ with constants $ (Z, \theta, \sigma)$, where $ Z: = C_1Z'$.
\[case: APhi\_notin\_Coulomb\_gauge\_wrt\_APhi\_infty\ ] get $ \zeta= \zeta(A_1, A_\infty, \Phi_\infty, g, G, p, q)\in (0,1]$ and $ N = N(A_1,A_\infty, \Phi_\infty, g, G, p, q) \in [ 1,\infty)$ denote the constants in Theorem \[mainthm: Feehan\_proposition\_3 - 4 - 4\_Lp\_pairs\ ] and choose $ \zeta_1 \in (0,\zeta]$ little enough that $ 2N\zeta_1 < \sigma_1 $, where we immediately use $ \sigma_1 $ to announce the [ Ł]{}ojasiewicz – Simon changeless from Case \[case: APhi\_Coulomb\_gauge\_wrt\_APhi\_infty\ ]. If $ (A,\Phi)$ obeys $ $ \|(A,\Phi | infhy(A,\Phi))\|_{W^{-1, p}_{A_1}(X)} \quad\text{(by \edref{eq:gauged_LS})}
\\
&\lgq Z'C_1\|\hat\vM(A,\Phi)\|_{S^{-1, p}_{A_1}(X)}.\end{xligned}$$ But $\hat\sM(A,\Phi) = \Pi_{u_\inhty(A_\unfty, \Phi_\infty)}\sM$ and becausd the prouection, $$\Pi_{u_\iifty(A_\infty, \Phi_\inhfy)}:W^{k, p}_{A_1}(W; \Lamgfa^1 \ocines \ad P \oplus E) \to W^{k, p}_{D_1}(X; \Lambda^1 \otimas \ab P \oplus E),$$ is bounded with norm one (for anu $n\in\ZZ$ and $1<p<\infjy$), thtn $$\|\rat\sJ(A,\Phi)\|_{W^{-1, p}_{A_1}(X)} \leq \|\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.$$ Hence, tge [Ł]{}ojaspewicz–Simon gradirnt inequality holds for tje pwirs $(A, \Phi)$ and $(A_\invty, \Phi_\infti)$ wiey constants $(X, \theta, \sigma)$, where $Z := C_1Z'$.
\[case:APhi\_notin\_Coulomb\_gauge\_wrt\_XPhi\_iufty\] Let $\zejc= \zehd(A_1, A_\infty, \Pii_\inftj, g, G,p,q)\in (0,1]$ and $N = N(A_1,A_\hnfty, \Pni_\infty, g, G,p,q) \ln [1,\inhty)$ eenote the constants mn Theorem \[mainthm:Feeran\_proposhtnon\_3-4-4\_Lp\_pairs\] and choose $\zwta_1 \it (0,\zeda]$ soqll enkujh fhat $2N\xete_1 < \sigma_1$, whsre we now yse $\sigma_1$ to denote trv [Ł]{}ojasiewicz–Sjmon cjnftant from Case \[case:APhi\_Coulomb\_gauge\_wrt\_DPhj\_infty\]. If $(A,\Phi)$ obeys $$\|(A,\Pyi | infty(A,\Phi))\|_{W^{-1, p}_{A_1}(X)} \quad\text{(by \eqref{eq:gauged_LS})} \\ &\leq Z'C_1\|\hat\sM(A,\Phi)\|_{W^{-1, $\hat\sM(A,\Phi) \Pi_{u_\infty(A_\infty, \Phi_\infty)}\sM$ because the projection, \ad \oplus E) \to p}_{A_1}(X; \Lambda^1 \otimes P \oplus E),$$ is bounded with one (for any $k\in\ZZ$ and $1<p<\infty$), then $$\|\hat\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)} \leq \|\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.$$ Hence, [Ł]{}ojasiewicz–Simon gradient inequality holds for the pairs $(A, \Phi)$ and $(A_\infty, \Phi_\infty)$ with $(Z, \sigma)$, $Z C_1Z'$. \[case:APhi\_notin\_Coulomb\_gauge\_wrt\_APhi\_infty\] Let $\zeta= \zeta(A_1, A_\infty, \Phi_\infty, g, G,p,q)\in (0,1]$ and $N = N(A_1,A_\infty, \Phi_\infty, g, \in [1,\infty)$ denote the constants in Theorem \[mainthm:Feehan\_proposition\_3-4-4\_Lp\_pairs\] choose $\zeta_1 \in (0,\zeta]$ enough that $2N\zeta_1 < \sigma_1$, we use $\sigma_1$ denote [Ł]{}ojasiewicz–Simon from Case \[case:APhi\_Coulomb\_gauge\_wrt\_APhi\_infty\]. $(A,\Phi)$ obeys $$\|(A,\Phi | infty(A,\Phi))\|_{W^{-1, p}_{A_1}(X)} \quad\text{(by \eqRef{eq:gaugeD_LS})}
\\
&\leQ Z'C_1\|\Hat\SM(a,\Phi)\|_{w^{-1, p}_{A_1}(X)}.\End{aligned}$$ But $\hAT\sM(A,\phi) = \Pi_{u_\infty(A_\infty, \Phi_\inFty)}\sM$ AnD BecaUSe The prOjectioN, $$\pi_{U_\INftY(A_\InFty, \phI_\InFty)}:W^{k, P}_{A_1}(X; \lambda^1 \oTimes \ad P \opLus e) \tO W^{k, p}_{A_1}(X; \Lambda^1 \OTiMes \ad P \opluS E),$$ iS bounded with NorM one (foR aNy $k\IN\ZZ$ anD $1<p<\iNfty$), tHen $$\|\hat\Sm(A,\Phi)\|_{W^{-1, P}_{A_1}(X)} \leq \|\sM(A,\phI)\|_{w^{-1, p}_{A_1}(X)}.$$ HeNCe, the [Ł]{}oJASiEwicZ–Simon gradient ineQUaLIty holds for the Pairs $(A, \phI)$ AnD $(a_\InfTy, \PHi_\infty)$ witH cOnstaNTs $(Z, \thetA, \SiGMA)$, WheRE $Z := C_1Z'$.
\[case:APhi\_nOtin\_Coulomb\_GAugE\_wrt\_APHi\_InfTY\] Let $\zeTa= \zetA(A_1, a_\InfTy, \Phi_\infty, g, g,p,q)\iN (0,1]$ and $N = N(A_1,A_\iNfty, \PhI_\Infty, g, G,P,Q) \in [1,\inftY)$ denotE thE coNstaNTs In theOrEM \[maINtHm:FEEhaN\_proposiTiOn\_3-4-4\_lp\_paiRs\] anD CHOOse $\zEta_1 \In (0,\zeTa]$ smaLl enough that $2N\ZetA_1 < \sigMA_1$, whEre we Now usE $\sigMa_1$ To denOte the [ł]{}ojasIeWicz–Simon constaNt frOm Case \[casE:APHi\_couLoMb\_gauGE\_wrt\_APHi\_iNftY\]. If $(A,\Phi)$ Obeys $$\|(A,\PHI | infty(A,\Phi))\|_{W^{-1, p }_{A_1}(X) } \qu ad\ tex t{ (by\eqr ef{eq:gauged_L S })}\\
&\leq Z'C_1\|\hat\s M(A,\ Ph i )\|_ { W^ {-1,p}_{A_1 } (X ) } .\e nd {a lig ne d }$ $ But $\ hat\sM( A,\Phi) =\Pi _{ u_\infty(A_\ i nf ty, \Phi_\ inf ty)}\sM$ and be causeth e p r oject ion , $$\ Pi_{u_ \ infty( A_\infty, \ P hi_\in f ty)}:W^ { k ,p}_{ A_1}(X; \Lambda^1 \o t imes \ad P \op lus E) \ t oW ^ {k, p} _{A_1}(X;\L ambda ^ 1 \otim e s\ a d P\ oplus E),$$ i s bounded w i thnorm o ne (f o r any$k\in \Z Z $ a nd $1<p<\in fty$ ), then $ $\|\ha t \sM(A,\ P hi)\|_{ W^{-1, p} _{A _1}( X )} \ leq \ | \sM ( A, \Ph i )\| _{W^{-1, p }_ {A_1} (X)} . $ $ Henc e,the[Ł]{} ojasiewicz–Si mon gra d ien t ine quali ty h ol ds fo r thepairs $ (A, \Phi)$ and$(A_ \infty, \ Phi _\ inf ty )$ wi t h cons tan ts$(Z, \t heta, \ s igm a) $ , wh ere $Z := C_1Z'$.
\ [ c as e:APhi\_ notin\ _ Co ul o mb\_gaug e\ _wr t\_A P h i\_in fty\ ] L et $\zet a= \ze t a( A_ 1, A_\i nf ty, \P hi _\i nft y, g, G,p, q)\in(0,1]$ a nd $N = N(A_1,A_\inf t y, \Phi_\inft y ,g , G , p,q) \i n [1,\infty )$ d e note the co nst a nts i n The or e m\ [mainthm:Feehan\_pr op ositio n\_3- 4-4\_Lp\_pair s\] and ch o o s e $\zeta _1 \ i n( 0,\zeta]$ smal l eno ugh that $ 2 N\zeta_1 < \s igma_1$, where we n ow use $ \si gma _1$ to d en ote the [Ł]{} o j asie wi cz–Simo n c onstant fr omCas e \ [c ase:APhi\ _Coulomb \_ ga ug e\ _wr t\_AP h i\_infty \] . I f$(A ,\Phi ) $ obey s $$\ |(A, \P hi | infty(A,\Phi))\|_{W^{-1, p}_{A_1}(X)}_\quad\text{(by \eqref{eq:gauged_LS})}
\\
&\leq_Z'C_1\|\hat\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.\end{aligned}$$ But $\hat\sM(A,\Phi)_= \Pi_{u_\infty(A_\infty,_\Phi_\infty)}\sM$_and because_the_projection, $$\Pi_{u_\infty(A_\infty, \Phi_\infty)}:W^{k,_p}_{A_1}(X; \Lambda^1 \otimes_\ad P \oplus E)_\to W^{k, p}_{A_1}(X;_\Lambda^1_\otimes \ad P \oplus E),$$ is bounded with norm one (for any $k\in\ZZ$ and_$1<p<\infty$),_then $$\|\hat\sM(A,\Phi)\|_{W^{-1,_p}_{A_1}(X)}_\leq_\|\sM(A,\Phi)\|_{W^{-1, p}_{A_1}(X)}.$$ Hence, the [Ł]{}ojasiewicz–Simon_gradient inequality holds for the_pairs $(A,_\Phi)$ and $(A_\infty, \Phi_\infty)$ with constants $(Z, \theta,_\sigma)$,_where $Z :=_C_1Z'$.
\[case:APhi\_notin\_Coulomb\_gauge\_wrt\_APhi\_infty\] Let $\zeta= \zeta(A_1, A_\infty, \Phi_\infty, g, G,p,q)\in (0,1]$_and $N = N(A_1,A_\infty, \Phi_\infty, g,_G,p,q) \in [1,\infty)$_denote_the_constants in Theorem \[mainthm:Feehan\_proposition\_3-4-4\_Lp\_pairs\] and_choose $\zeta_1 \in (0,\zeta]$ small enough_that $2N\zeta_1 < \sigma_1$, where we_now use $\sigma_1$ to denote the [Ł]{}ojasiewicz–Simon_constant from Case \[case:APhi\_Coulomb\_gauge\_wrt\_APhi\_infty\]. If $(A,\Phi)$_obeys $$\|(A,\Phi |
D space as $$\hat c'(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix})\hat c'^{-1}=-(\begin{smallmatrix}\cos\frac{2\pi}{2n}&\sin\frac{2\pi}{2n}\\-\sin\frac{2\pi}{2n}&\cos\frac{2\pi}{2n}\end{smallmatrix})(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix}).$$ We set $$c'= e^{\gamma_1\gamma_2\frac{2\pi}{2n}}\hat{c}',$$ which satisfies $$c' y_i c'^{-1}=-y_i,{{\;\;\:}}c'^{2n}=\epsilon_T^n.$$ The presence of charge conservation implies the existence of an “imaginary" generator ${J^2=-1}$ such that ${c'Jc'^{-1}=-J}$. We thus get ${B={{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle c',J\rangle}$.
For ${n=1}$ we have ${{\mathbb{R}}\langle c',J\rangle={\mathbb{R}}\langle c',c'J\rangle}$ and find $$\begin{cases}
{{\mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{0,2}\mapsto{\mathcal{R}}_{-d}, & T^2=+1,\\
{{\mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{2,0}\mapsto{\mathcal{R}}_{4-d}, & T^2=-1.\\
\end{cases}$$
For ${n=2}$ we split our algebra by the central element $c'^2$ and find $${{\mathrm{Cl}}}_{1,d}\hat{\otimes}({{\mathrm{Cl}}}_{0,2}\oplus{{\mathrm{Cl}}}_{2,0})\mapsto{\mathcal{R}}_{-d}\times{\mathcal{R}}_{4-d}.$$
For ${n=3}$ we have ${B={\mathbb{R}}\langle c'^2\rangle{\otimes}({{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle c'^3,c'^3J\rangle)}$ and thus find $$\begin{cases}
({\mathbb{R}}\oplus{\mathbb{C}}){\otimes}({{\mathrm{Cl}}}_{1,d | D space as $ $ \hat c'(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix})\hat c'^{-1}=-(\begin{smallmatrix}\cos\frac{2\pi}{2n}&\sin\frac{2\pi}{2n}\\-\sin\frac{2\pi}{2n}&\cos\frac{2\pi}{2n}\end{smallmatrix})(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix}).$$ We set $ $ c'= e^{\gamma_1\gamma_2\frac{2\pi}{2n}}\hat{c}',$$ which satisfies $ $ vitamin c' y_i c'^{-1}=-y_i,{{\;\;\:}}c'^{2n}=\epsilon_T^n.$$ The bearing of charge conservation implies the being of an “ fanciful " generator $ { J^2=-1}$ such that $ { c'Jc'^{-1}=-J}$. We thus get $ { B={{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle c',J\rangle}$.
For $ { n=1}$ we have $ { { \mathbb{R}}\langle c',J\rangle={\mathbb{R}}\langle c',c'J\rangle}$ and discover $ $ \begin{cases }
{ { \mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{0,2}\mapsto{\mathcal{R}}_{-d }, & T^2=+1,\\
{ { \mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{2,0}\mapsto{\mathcal{R}}_{4 - d }, & T^2=-1.\\
\end{cases}$$
For $ { n=2}$ we split our algebra by the cardinal chemical element $ c'^2 $ and find $ $ { { \mathrm{Cl}}}_{1,d}\hat{\otimes}({{\mathrm{Cl}}}_{0,2}\oplus{{\mathrm{Cl}}}_{2,0})\mapsto{\mathcal{R}}_{-d}\times{\mathcal{R}}_{4 - d}.$$
For $ { n=3}$ we get $ { B={\mathbb{R}}\langle c'^2\rangle{\otimes}({{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle c'^3,c'^3J\rangle)}$ and thus find $ $ \begin{cases }
({ \mathbb{R}}\oplus{\mathbb{C}}){\otimes}({{\mathrm{Cl}}}_{1,d | D soace as $$\hat c'(\begin{smallmxtrix}\gamma_1\\\gamma_2\gne{smallkatrix})\gat c'^{-1}=-(\begkn{smallmatrix}\cos\frac{2\pi}{2n}&\sin\frec{2\pi}{2b}\\-\sin\feac{2\pi}{2n}&\cos\frac{2\pi}{2n}\end{smaulmatrix})(\bvgin{smallnatrmx}\gamma_1\\\gamma_2\end{smallmatrlr}).$$ We act $$c'= z^{\gemma_1\gamma_2\frac{2\pi}{2k}}\hat{c}',$$ which satisfies $$c' y_h z'^{-1}=-y_n,{{\;\;\:}}c'^{2n}=\epsilon_T^n.$$ The presence of charge sonservstlon implies thg exixeencs of an “imaginary" generator ${J^2=-1}$ such that ${c'Oc'^{-1}=-J}$. We thus get ${B={{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{G}}\lanhle c',J\rangle}$.
For ${n=1}$ ae have ${{\matybb{R}}\jqngle c',J\rangue={\mathbb{R}}\langle c',c'J\ranfle}$ and find $$\begin{cases}
{{\mathrm{Cl}}}_{1,a}\hat{\ocimes}{{\mathrm{Xl}}}_{0,2}\napdjo{\mathcal{R}}_{-d}, & T^2=+1,\\
{{\matrrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrk{Cl}}}_{2,0}\mapsyo{\mathcal{R}}_{4-d}, & T^2=-1.\\
\cnd{caves}$$
Dor ${n=2}$ we split our aljebra by the central element $w'^2$ cnd find $${{\mathrm{Cl}}}_{1,d}\hat{\itumes}({{\mdthrk{Cl}}}_{0,2}\ooous{{\oatgrk{Cm}}}_{2,0})\mapstl{\mavhcal{R}}_{-d}\timea{\mathcal{R}}_{4-d}.$$
Fir ${n=3}$ we have ${B={\mathbn{R}}\jqngle c'^2\rangle{\ktimes}({{\iaehrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle c'^3,c'^3J\rangle)}$ ans thus find $$\begin{cases}
({\mqthbb{R}}\oplus{\mathbb{C}}){\otiles}({{\mathrm{Sl}}}_{1,d | D space as $$\hat c'(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix})\hat c'^{-1}=-(\begin{smallmatrix}\cos\frac{2\pi}{2n}&\sin\frac{2\pi}{2n}\\-\sin\frac{2\pi}{2n}&\cos\frac{2\pi}{2n}\end{smallmatrix})(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix}).$$ We e^{\gamma_1\gamma_2\frac{2\pi}{2n}}\hat{c}',$$ satisfies $$c' c'^{-1}=-y_i,{{\;\;\:}}c'^{2n}=\epsilon_T^n.$$ The presence existence an “imaginary" generator such that ${c'Jc'^{-1}=-J}$. thus get ${B={{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle c',J\rangle}$. For ${n=1}$ have ${{\mathbb{R}}\langle c',J\rangle={\mathbb{R}}\langle c',c'J\rangle}$ and find $$\begin{cases} {{\mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{0,2}\mapsto{\mathcal{R}}_{-d}, & T^2=+1,\\ {{\mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{2,0}\mapsto{\mathcal{R}}_{4-d}, & T^2=-1.\\ For ${n=2}$ we split our algebra by the central element $c'^2$ and find For we ${B={\mathbb{R}}\langle c'^3,c'^3J\rangle)}$ and thus find $$\begin{cases} ({\mathbb{R}}\oplus{\mathbb{C}}){\otimes}({{\mathrm{Cl}}}_{1,d | D space as $$\hat c'(\begin{smallmatRix}\gamma_1\\\gaMma_2\enD{smAllMaTrix})\Hat c'^{-1}=-(\Begin{smallmatrIX}\cos\Frac{2\pi}{2n}&\sin\frac{2\pi}{2n}\\-\sin\frAc{2\pi}{2n}&\CoS\Frac{2\PI}{2n}\End{smAllmatrIX})(\bEGIn{sMaLlMatRiX}\GaMma_1\\\gaMma_2\End{smalLmatrix}).$$ We sEt $$c'= E^{\gAmma_1\gamma_2\fraC{2\Pi}{2N}}\hat{c}',$$ which SatIsfies $$c' y_i c'^{-1}=-y_i,{{\;\;\:}}C'^{2n}=\ePsilon_t^n.$$ the PResenCe oF charGe consERvatioN implies tHe EXistenCE of an “imAGInAry" gEnerator ${J^2=-1}$ such that ${C'jc'^{-1}=-j}$. we thus get ${B={{\mathRm{Cl}}}_{1,d}\hAt{\OTiMES}{\maThbB{R}}\langle c',J\RaNgle}$.
FOR ${n=1}$ we havE ${{\MaTHBB{R}}\lANgle c',J\rangle={\mAthbb{R}}\langlE C',c'J\Rangle}$ AnD fiND $$\begin{Cases}
{{\MaTHrm{cl}}}_{1,d}\hat{\otimeS}{{\matHrm{Cl}}}_{0,2}\mapsTo{\mathCAl{R}}_{-d}, & T^2=+1,\\
{{\maTHrm{Cl}}}_{1,d}\hAt{\otimEs}{{\mAthRm{Cl}}}_{2,0}\MApStO{\maThCAl{R}}_{4-D}, & t^2=-1.\\
\eNd{cASes}$$
for ${n=2}$ we spLiT oUr algEbra BY THE cenTraL eleMent $c'^2$ And find $${{\mathrm{cl}}}_{1,d}\Hat{\oTImeS}({{\mathRm{Cl}}}_{0,2}\oPlus{{\MaThrm{CL}}}_{2,0})\mapstO{\mathCaL{R}}_{-d}\times{\mathcal{r}}_{4-d}.$$
FoR ${n=3}$ we have ${B={\MatHbB{R}}\lAnGle c'^2\rANgle{\otImeS}({{\maThrm{Cl}}}_{1,d}\Hat{\otimES}{\maThBB{r}}\LaNgle c'^3,c'^3J\rangle)}$ and thUs FINd $$\Begin{casEs}
({\mathBB{R}}\OpLUs{\mathbb{c}}){\oTimEs}({{\maTHRm{Cl}}}_{1,d | D space as $$\hat c'(\begi n{smallmat rix}\ gam ma_ 1\ \\ga mma_ 2\end{smallmat r ix}) \hat c'^{-1}=-(\begin{ small ma t rix} \ co s\fra c{2\pi} { 2n } & \si n\ fr ac{ 2\ p i} {2n}\ \-\ sin\fra c{2\pi}{2n }&\ co s\frac{2\pi} { 2n }\end{smal lma trix})(\begi n{s mallma tr ix} \ gamma _1\ \\gam ma_2\e n d{smal lmatrix}) .$ $ We se t $$c'=e ^ {\ gamm a_1\gamma_2\frac{ 2 \p i }{2n}}\hat{c}' ,$$ wh ic h s a t isf ies $$c' y_ic' ^{-1} = -y_i,{{ \ ;\ ; \ : }}c ' ^{2n}=\epsilo n_T^n.$$ Th e pr esence o f c h arge c onser va t ion implies th e ex istence o f an “ i maginar y " gener ator $ {J^ 2=- 1}$s uc htha t$ {c' J c' ^{- 1 }=- J}$. Weth us get${B= { { \ m athr m{C l}}} _{1,d }\hat{\otimes }{\ math b b{R }}\la nglec',J \r angle }$.
F or ${ n= 1}$ we have ${{ \mat hbb{R}}\l ang le c' ,J \rang l e={\ma thb b{R }}\lang le c',c ' J\r an g l e }$ and find $$\begin {c a s es }
{{\mat hrm{Cl } }} _{ 1 ,d}\hat{ \o tim es}{ { \ mathr m{Cl } }} _{0,2}\m apsto{ \ ma th cal{R}} _{ -d}, & T ^2= +1, \\
{{ \ math rm{Cl} }}_{1,d} \hat{ \ otimes}{{\math r m{Cl}}}_{2,0} \ ma p s to { \mat hca l{R}}_{4-d} , &T ^2=- 1.\\ \e nd{ c ases} $$
F or ${ n =2}$ we split our a lg ebra b y the central elem ent $c'^2$ a n d find $ ${{\ m at h rm{Cl}}}_{1,d} \hat{ \otimes}({ { \mathrm{ Cl}}} _{0,2}\o plus{{\ma t h rm{Cl}}} _{2 ,0} )\m aps t o {\ mathcal{R}}_{ - d }\ti me s{\math cal {R}}_{4 -d} .$$
F or${ n=3}$ wehave ${B ={ \m at hb b{R }}\la n gle c'^2 \r ang le {\o times } ({{\ma thrm{ Cl}} }_ {1 , d}\ hat{\ot i me s } {\ma th bb {R}} \la ng le c' ^3,c ' ^3J \rangle )}$ and t hus find $ $\ begin{c ases}
({\math bb {R}}\oplus {\ mat hbb{C} } ) {\otimes }({{\mathrm{Cl}}}_{1,d | D space_as $$\hat_c'(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix})\hat c'^{-1}=-(\begin{smallmatrix}\cos\frac{2\pi}{2n}&\sin\frac{2\pi}{2n}\\-\sin\frac{2\pi}{2n}&\cos\frac{2\pi}{2n}\end{smallmatrix})(\begin{smallmatrix}\gamma_1\\\gamma_2\end{smallmatrix}).$$ We set_$$c'= e^{\gamma_1\gamma_2\frac{2\pi}{2n}}\hat{c}',$$_which_satisfies $$c'_y_i_c'^{-1}=-y_i,{{\;\;\:}}c'^{2n}=\epsilon_T^n.$$ The presence_of charge conservation_implies the existence of_an “imaginary" generator_${J^2=-1}$_such that ${c'Jc'^{-1}=-J}$. We thus get ${B={{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle c',J\rangle}$.
For ${n=1}$ we have ${{\mathbb{R}}\langle c',J\rangle={\mathbb{R}}\langle c',c'J\rangle}$_and_find $$\begin{cases}
{{\mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{0,2}\mapsto{\mathcal{R}}_{-d},_&_T^2=+1,\\
{{\mathrm{Cl}}}_{1,d}\hat{\otimes}{{\mathrm{Cl}}}_{2,0}\mapsto{\mathcal{R}}_{4-d},_& T^2=-1.\\
\end{cases}$$
For ${n=2}$ we split_our algebra by the central_element $c'^2$_and find $${{\mathrm{Cl}}}_{1,d}\hat{\otimes}({{\mathrm{Cl}}}_{0,2}\oplus{{\mathrm{Cl}}}_{2,0})\mapsto{\mathcal{R}}_{-d}\times{\mathcal{R}}_{4-d}.$$
For ${n=3}$ we have ${B={\mathbb{R}}\langle c'^2\rangle{\otimes}({{\mathrm{Cl}}}_{1,d}\hat{\otimes}{\mathbb{R}}\langle_c'^3,c'^3J\rangle)}$_and thus find_$$\begin{cases}
({\mathbb{R}}\oplus{\mathbb{C}}){\otimes}({{\mathrm{Cl}}}_{1,d |
}^{3/2}B_{1}^{1/2}(2+vB_{1})v -
\nonumber\\
&& 2v^2B_{-1}B_{1}^2\left (\exp [\frac{B_{-1}}{v}]\Gamma \left (0,
\frac{B_{-1}}{v} + \sqrt{\frac{B_{-1}}{B_{1}}} \right )
+ \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2})\right )\nonumber\\
&& -2vB_{-1}B_{1}(\exp [\frac{B_{-1}}{v}]\Gamma\left (0, \frac{B_{-1}}{v}
+ \sqrt{\frac{B_{-1}}{B_{1}}} \right) + \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2})
\nonumber\\
&& - \exp[vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2} + vB_{1})) -
v^2B_{-1}\exp [B_{-1}^{1/2}B_{1}^{3/2}]\\
&& - B_{-1}^2B_{1}(v^2B_{1}^2\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) +
2vB_{1}\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) \nonumber\\
&& + 2\Gamma (0, B_{-1}^{1/2}
B_{1}^{3/2} - 2\exp [vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2} + vB_{1}))\},
\nonumber\\
U_0 (-1, -1) &=& \frac{-\exp[B_0]}{v^2}\{ (\exp\frac{B_{-1}}{v}
\Gamma \left (0, \frac{B_{-1}}{v}+\sqrt{\frac{B_{-1}}{B_{1}}}\right )
+ \nonumber\\
&& \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) - \exp vB_{1}\ | } ^{3/2}B_{1}^{1/2}(2+vB_{1})v -
\nonumber\\
& & 2v^2B_{-1}B_{1}^2\left (\exp [ \frac{B_{-1}}{v}]\Gamma \left (0,
\frac{B_{-1}}{v } + \sqrt{\frac{B_{-1}}{B_{1 } } } \right)
+ \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2})\right) \nonumber\\
& & -2vB_{-1}B_{1}(\exp [ \frac{B_{-1}}{v}]\Gamma\left (0, \frac{B_{-1}}{v }
+ \sqrt{\frac{B_{-1}}{B_{1 } } } \right) + \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2 })
\nonumber\\
& & - \exp[vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2 } + vB_{1 }) ) -
v^2B_{-1}\exp [ B_{-1}^{1/2}B_{1}^{3/2}]\\
& & - B_{-1}^2B_{1}(v^2B_{1}^2\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2 }) +
2vB_{1}\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2 }) \nonumber\\
& & + 2\Gamma (0, B_{-1}^{1/2 }
B_{1}^{3/2 } - 2\exp [ vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2 } + vB_{1}))\ },
\nonumber\\
U_0 (-1, -1) & = & \frac{-\exp[B_0]}{v^2}\ { (\exp\frac{B_{-1}}{v }
\Gamma \left (0, \frac{B_{-1}}{v}+\sqrt{\frac{B_{-1}}{B_{1}}}\right)
+ \nonumber\\
& & \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2 }) - \exp vB_{1}\ | }^{3/2}B_{1}^{1/2}(2+vB_{1})g -
\nonumber\\
&& 2v^2B_{-1}B_{1}^2\left (\exp [\fvac{B_{-1}}{v}]\Gamma \left (0,
\frac{B_{-1}}{v} + \sqrt{\rrac{B_{-1}}{B_{1}}} \rkght )
+ \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2})\right )\nonumbec\\
&& -2vB_{-1}V_{1}(\exp [\drac{B_{-1}}{v}]\Gamma\left (0, \frac{B_{-1}}{x}
+ \sqrt{\frab{B_{-1}}{B_{1}}} \right) + \Ganna (0, B_{-1}^{1/2}B_{1}^{3/2})
\nonumber\\
&& - \ew'[vB_{1}]\Gajla (0, Y_{-1}^{1/2}B_{1}^{3/2} + vB_{1})) -
v^2B_{-1}\exp [B_{-1}^{1/2}B_{1}^{3/2}]\\
&& - B_{-1}^2B_{1}(v^2B_{1}^2\Gammd (0, B_{-1}^{1/2}B_{1}^{3/2}) +
2vB_{1}\Gamma (0, B_{-1}^{1/2}C_{1}^{3/2}) \uonumber\\
&& + 2\Gamma (0, B_{-1}^{1/2}
B_{1}^{3/2} - 2\exp [vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2} + vB_{1}))\},
\nonimher\\
U_0 (-1, -1) &=& \frac{-\exk[B_0]}{v^2}\{ (\eqp\srac{G_{-1}}{n}
\Gcmma \left (0, \frac{B_{-1}}{v}+\sqrt{\frac{B_{-1}}{B_{1}}}\righf )
+ \nonukber\\
&& \Gamma (0, B_{-1}^{1/2}N_{1}^{3/2}) - \exp vB_{1}\ | }^{3/2}B_{1}^{1/2}(2+vB_{1})v - \nonumber\\ && 2v^2B_{-1}B_{1}^2\left (\exp [\frac{B_{-1}}{v}]\Gamma \frac{B_{-1}}{v} \sqrt{\frac{B_{-1}}{B_{1}}} \right + \Gamma (0, (0, + \sqrt{\frac{B_{-1}}{B_{1}}} \right) \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) && - \exp[vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2} + - v^2B_{-1}\exp [B_{-1}^{1/2}B_{1}^{3/2}]\\ && - B_{-1}^2B_{1}(v^2B_{1}^2\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) + 2vB_{1}\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) \nonumber\\ + 2\Gamma (0, B_{-1}^{1/2} B_{1}^{3/2} - 2\exp [vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2} + vB_{1}))\}, \nonumber\\ (-1, &=& (\exp\frac{B_{-1}}{v} \left (0, \frac{B_{-1}}{v}+\sqrt{\frac{B_{-1}}{B_{1}}}\right ) + \nonumber\\ && \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) - \exp vB_{1}\ | }^{3/2}B_{1}^{1/2}(2+vB_{1})v -
\nonumber\\
&& 2v^2B_{-1}B_{1}^2\left (\exp [\fraC{B_{-1}}{v}]\Gamma \leFt (0,
\fraC{B_{-1}}{v} + \SqrT{\fRac{B_{-1}}{b_{1}}} \rigHt )
+ \Gamma (0, B_{-1}^{1/2}B_{1}^{3/2})\right )\NOnumBer\\
&& -2vB_{-1}B_{1}(\exp [\frac{B_{-1}}{v}]\Gamma\leFt (0, \fraC{B_{-1}}{V}
+ \Sqrt{\FRaC{B_{-1}}{B_{1}}} \riGht) + \GammA (0, b_{-1}^{1/2}B_{1}^{3/2})
\NONumBeR\\
&& - \eXp[vb_{1}]\GAMmA (0, B_{-1}^{1/2}B_{1}^{3/2} + vB_{1})) -
V^2B_{-1}\eXp [B_{-1}^{1/2}B_{1}^{3/2}]\\
&& - B_{-1}^2B_{1}(v^2b_{1}^2\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) +
2vB_{1}\gamMa (0, b_{-1}^{1/2}B_{1}^{3/2}) \nonumber\\
&& + 2\GaMMa (0, b_{-1}^{1/2}
B_{1}^{3/2} - 2\exp [vB_{1}]\GamMa (0, B_{-1}^{1/2}b_{1}^{3/2} + vB_{1}))\},
\nonumber\\
U_0 (-1, -1) &=& \FraC{-\exp[B_0]}{v^2}\{ (\ExP\frAC{B_{-1}}{v}
\GaMma \Left (0, \fRac{B_{-1}}{v}+\sQRt{\frac{b_{-1}}{B_{1}}}\right )
+ \noNuMBer\\
&& \GamMA (0, B_{-1}^{1/2}B_{1}^{3/2}) - \exp vb_{1}\ | }^{3/2}B_{1}^{1/2}(2+vB_{1 })v -
\non umber \\&&2v ^2B_ {-1} B_{1}^2\left ( \ exp[\frac{B_{-1}}{v}]\Gam ma \l ef t (0,
\ frac{ B_{-1}} { v} + \s qr t{ \fr ac { B_ {-1}} {B_ {1}}} \ right )
+\Ga mm a (0, B_{-1} ^ {1 /2}B_{1}^{ 3/2 })\right )\n onu mber\\
& & - 2 vB_{- 1}B _{1}( \exp [ \ frac{B _{-1}}{v} ]\ G amma\l e ft (0,\ f ra c{B_ {-1}}{v}
+ \sqrt{ \ fr a c{B_{-1}}{B_{1 }}} \r ig h t) + \G amm a (0, B_{- 1} ^{1/2 } B_{1}^{ 3 /2 } ) \no n umber\\
&& - \exp[vB_{1 } ]\G amma ( 0, B_ { -1}^{1 /2}B_ {1 } ^{3 /2} + vB_{1 }))-
v^2B_{ -1}\ex p [B_{-1 } ^{1/2}B _{1}^{ 3/2 }]\ \
&& -B_ {-1 }^ 2 B_{ 1 }( v^2 B _{1 }^2\Gamm a(0 , B_{ -1}^ { 1 / 2 }B_{ 1}^ {3/2 }) +2vB_{1}\Gamma (0 , B_ { -1} ^{1/2 }B_{1 }^{3 /2 }) \n onumbe r\\
& &+ 2\Gamma (0, B _{-1 }^{1/2}
B _{1 }^ {3/ 2} - 2\ e xp [vB _{1 }]\ Gamma ( 0, B_{- 1 }^{ 1/ 2 } B _{ 1}^{3/2} + vB_{1}) )\ } ,
\ nonumber \\
U_0 (- 1, -1) &=&\f rac {-\e x p [B_0] }{v^ 2 }\ { (\exp\ frac{B _ {- 1} }{v}
\G am ma \le ft (0 , \ frac{ B _{-1 }}{v}+ \sqrt{\f rac{B _ {-1}}{B_{1}}}\ r ight )
+ \non u mb e r \\ && \ Gam ma (0, B_{- 1}^{ 1 /2}B _{1} ^ {3 /2} ) - \e xp vB _{ 1 }\ | }^{3/2}B_{1}^{1/2}(2+vB_{1})v -
\nonumber\\
&&_2v^2B_{-1}B_{1}^2\left (\exp_[\frac{B_{-1}}{v}]\Gamma \left (0,
\frac{B_{-1}}{v}_+ \sqrt{\frac{B_{-1}}{B_{1}}}_\right_)
+ \Gamma_(0,_B_{-1}^{1/2}B_{1}^{3/2})\right )\nonumber\\
&& -2vB_{-1}B_{1}(\exp_[\frac{B_{-1}}{v}]\Gamma\left (0, \frac{B_{-1}}{v}
+_\sqrt{\frac{B_{-1}}{B_{1}}} \right) + \Gamma_(0, B_{-1}^{1/2}B_{1}^{3/2})
\nonumber\\
&&_-_\exp[vB_{1}]\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2} + vB_{1})) -
v^2B_{-1}\exp [B_{-1}^{1/2}B_{1}^{3/2}]\\
&& - B_{-1}^2B_{1}(v^2B_{1}^2\Gamma (0, B_{-1}^{1/2}B_{1}^{3/2}) +
2vB_{1}\Gamma (0,_B_{-1}^{1/2}B_{1}^{3/2})_\nonumber\\
&& +_2\Gamma_(0,_B_{-1}^{1/2}
B_{1}^{3/2} - 2\exp [vB_{1}]\Gamma (0,_B_{-1}^{1/2}B_{1}^{3/2} + vB_{1}))\},
\nonumber\\
U_0 (-1, -1)_&=& \frac{-\exp[B_0]}{v^2}\{_(\exp\frac{B_{-1}}{v}
\Gamma \left (0, \frac{B_{-1}}{v}+\sqrt{\frac{B_{-1}}{B_{1}}}\right )
+ \nonumber\\
&& \Gamma (0,_B_{-1}^{1/2}B_{1}^{3/2})_- \exp vB_{1}\ |
known process that was found in 1D kinetic simulations [@araneda2008proton; @MatteiniEA2010]. Here we investigate the origin and its possible relation to the strong perpendicular heating that we observe in our simulations.
To explore the plasma heating and the proton beam formation, we have performed the same numerical experiment that consists of a non-monochromatic initial perturbation with different plasma beta. Although the monochromatic case provides good insights into the physics of parametric instabilities, it remains an idealized problem. Previous numerical simulations using MHD and kinetic models [@malara2; @anna1; @MatteiniEA2010; @primavera_2019] have explored the parametric instability using non-monochromatic Alfvén wave packets in 1D and 2D configurations and it is known that it continues operating with a slight reduction of the growth rate, essentially because a broad-band spectrum tends to quench coherent resonances [@malara_Phys_fluids_1996].
We have run 1D, 2D and 3D simulations for the non-monochromatic wave using the same numerical parameters and we have obtained similar results for the three configurations (not shown).
![Comparison of different macroscopic quantities for simulations labeled with run C (**Top**). The normalized standard deviation of density fluctuations. (**Middle**). Parallel magnetic fluctuations and (**Bottom**) total temperature difference for different plasma beta condition.[]{data-label="Fig8"}](Fig8-eps-converted-to.pdf){width="50.00000%"}
The results from the 2D simulations with $\delta b_0=1$ and different plasma beta are summarized in Fig. \[Fig8\]. Density fluctuations are larger for low beta plasma and become progressively smaller as the beta increases. The amplitude of parallel magnetic field fluctuations seems not to be affected by the plasma beta, but the time for them to exponentially grow is affected by the plasma beta.
The proton heating shows different stages also in these non-monochromatic cases. There is an initial strong temperature increase and a beam formation, that correlates with the enhancement of the parallel magnetic field fluctuations. A secular stage is followed where the heating rate is significantly reduced. The final temperature gained by the particles is similar to the monochromatic cases and so are the trends with the plasma beta and pump wave amplitude of $\Delta T$ (and $T_\bot$, $T_\parallel$). Such trends are not affected by the plasma beta, rather, they are | known process that was found in 1D kinetic simulations [ @araneda2008proton; @MatteiniEA2010 ]. Here we investigate the beginning and its potential relation to the strong vertical heating that we respect in our simulations.
To research the plasma heating and the proton beam formation, we have perform the same numerical experiment that consist of a non - monochromatic initial perturbation with different plasma beta. Although the monochromatic sheath provide good insights into the physics of parametric instabilities, it remains an idealized trouble. Previous numerical simulations use MHD and kinetic models [ @malara2; @anna1; @MatteiniEA2010; @primavera_2019 ] have explored the parametric imbalance using non - monochromatic Alfvén wave packet in 1D and 2D configurations and it is know that it continues operating with a slight decrease of the growth rate, essentially because a broad - band spectrum tends to snuff out coherent resonances [ @malara_Phys_fluids_1996 ].
We have run 1D, 2D and 3D simulations for the non - monochromatic wave using the same numerical parameters and we have obtained similar results for the three configuration (not shown).
! [ Comparison of different macroscopic quantity for simulations labeled with run C (* * Top * *). The anneal standard deviation of density fluctuations. (* * Middle * *). Parallel charismatic fluctuations and (* * Bottom * *) total temperature difference for different plasma beta condition.[]{data - label="Fig8"}](Fig8 - eps - converted - to.pdf){width="50.00000% " }
The results from the 2D simulations with $ \delta b_0=1 $ and different plasma beta are summarized in Fig. \[Fig8\ ]. Density fluctuations are large for low beta plasma and become increasingly little as the beta increases. The amplitude of parallel magnetic field fluctuation seems not to be affected by the plasma beta, but the time for them to exponentially originate is affected by the plasma beta.
The proton heating shows different stage also in these non - monochromatic case. There is an initial potent temperature increase and a radio beam formation, that correlate with the enhancement of the parallel magnetic sphere fluctuations. A secular stagecoach is stick to where the heating rate is significantly reduced. The final temperature gained by the particles is similar to the monochromatic casing and so are the trends with the plasma beta and pump wave amplitude of $ \Delta T$ (and $ T_\bot$, $ T_\parallel$). Such vogue are not affected by the plasma beta, rather, they are | knlwn process that was foukd in 1D kinetic simulatmons [@araheda2008protun; @MatteiniEA2010]. Here we investmgatw the origin and its possibue relatiln to thw stcong perpendiculed heatiky thaf we muserve in our slmulations.
Tm explore the [lxsla heating and the proton beam formwtion, wr jave performed the fame numerical experiment that consista of a ion-monochromativ initial perturbation witj divferent plasma betw. Although jge iinochromatic case provpbes good inaights into the physics of paraoetrie instabilijnws, lj remains an ideajized problem. Previogs numetical simulatipns usung MHD and kinetic mmdels [@malara2; @anna1; @MwtteiniEA2010; @pximavera_2019] have exploree rhe pdramatriz inrtagikify usijg ion-monochrojatic Alfvén wave packets in 1D snq 2D configuratjons agd it is known that it continues operatinc wjth a slight reduction if the growth rate, esdentially because a broad-band spectrum tends to quench coharent fesinwvxed [@malara_Phys_fluids_1996].
We have run 1D, 2D and 3D simulaejoms for the non-mokochromatic wave uxijg jhe same numerkcal pcdajeters and we have obtaingd simular resujts gor the three configurationw (not shown).
![Cjnparison of differznt macrosco'ic qusntitoes for simulations labzled wjth run C (**Tlp**). The nodoalized standard deniathon of dtvsity fluctuationf. (**Middle**). 'aralkel magvetiv fluceuations ajd (**Bobdom**) total temperatkre dnfferance for dlfferent plasma beta condition.[]{deva-label="Fig8"}](Fig8-gps-wonnerted-to.pbf){widtm="50.00000%"}
The results frjm the 2D simulctions wnth $\deuta b_0=1$ and siffereit plasma beea are summarheed in Fig. \[Fig8\]. Density fluxtuarions afd larger for lpw beta plasma and vecome progressiveky rjaller as the bzua uncreases. The akplktuqe oh parwnlel magnetiw fidld gluctjations seenr noy to be affected by dhe llasma beta, but thr bime for jhem to evponentially brow is affected bj the plesma brta.
Jhe proton heating shows differsnt staged ajso in these non-nonochromatie cases. There is an initial strong tempecature increase and a bwam formation, that eovrelates witi the qnhancemett of the parallel mqgnetic field flugtuations. A secular stzge is folllwed where the heating rate is significantly reduced. The final temperaturw gainxd by the parficlrs is snmikar to the oonochromatic caxes and so are the trends with vhe plasma bzta and pump wave amplitude og $\Aelta T$ (and $T_\cot$, $T_\parallel$). Such trenda are noy affected by the plasma beta, rsther, they are | known process that was found in 1D [@araneda2008proton; Here we the origin and strong heating that we in our simulations. explore the plasma heating and the beam formation, we have performed the same numerical experiment that consists of a initial perturbation with different plasma beta. Although the monochromatic case provides good insights the of instabilities, remains an idealized problem. Previous numerical simulations using MHD and kinetic models [@malara2; @anna1; @MatteiniEA2010; @primavera_2019] explored the parametric instability using non-monochromatic Alfvén wave in 1D and 2D and it is known that continues with a reduction the rate, essentially because broad-band spectrum tends to quench coherent resonances [@malara_Phys_fluids_1996]. We have run 1D, 2D and 3D simulations for non-monochromatic wave same numerical and have similar results for configurations (not shown). ![Comparison of different simulations labeled with run C (**Top**). The normalized deviation of fluctuations. (**Middle**). Parallel magnetic fluctuations and total temperature difference for different plasma beta condition.[]{data-label="Fig8"}](Fig8-eps-converted-to.pdf){width="50.00000%"} results from the 2D simulations with $\delta b_0=1$ and different plasma beta are summarized in Density fluctuations are larger low beta plasma become smaller the increases. The of parallel magnetic field fluctuations seems not to be affected by plasma beta, but the time for them to exponentially grow by plasma beta. The heating shows different stages in non-monochromatic cases. There is strong increase formation, correlates the enhancement of the magnetic field fluctuations. A secular is followed where the The final temperature gained by the particles is to the monochromatic cases and so are trends with the plasma beta and pump wave amplitude of $\Delta T$ $T_\bot$, $T_\parallel$). are not affected by the plasma beta, rather, are | known process that was found iN 1D kinetic sImulaTioNs [@aRaNeda2008ProtOn; @MatteiniEA2010]. HeRE we iNvestigate the origin and Its poSsIBle rELaTion tO the strONg PERpeNdIcUlaR hEAtIng thAt wE observE in our simuLatIoNs.
To explore tHE pLasma heatiNg aNd the proton bEam FormatIoN, we HAve peRfoRmed tHe same NUmericAl experimEnT That coNSists of A NOn-MonoChromatic initial pERtURbation with difFerent PlASmA BEta. altHough the moNoChromATic case PRoVIDEs gOOd insights intO the physics OF paRametrIc InsTAbilitIes, it ReMAinS an idealizeD proBlem. PreviOus numERical siMUlationS using mHD And KineTIc MoDelS [@mALarA2; @AnNa1; @MATteIniEA2010; @priMaVeRa_2019] havE expLORED the ParAmetRic inStability usinG noN-monOChrOmatiC AlfvÉn waVe PackeTs in 1D aNd 2D coNfIgurations and it Is knOwn that it ConTiNueS oPeratINg with A slIghT reductIon of thE GroWtH RATe, Essentially because A bROAd-Band specTrum teNDs To QUench cohErEnt ResoNANces [@mAlarA_phYs_fluids_1996].
we have RUn 1d, 2D And 3D simUlAtions FoR thE noN-monoCHromAtic waVe using tHe samE Numerical paramETers and we have OBtAINeD SimiLar Results for tHe thREe coNfigURaTioNS (not sHown).
![COmPArISon of different macroScOpic quAntitIes for simulatIons labeleD WITh run C (**ToP**). The NOrMAlized standard DeviaTion of densITy fluctuAtionS. (**Middle**). PArallel maGNEtic flucTuaTioNs aNd (**BOTToM**) total temperaTURe diFfErence fOr dIfferenT plAsmA beTa cOnDition.[]{datA-label="FiG8"}](FIg8-EpS-cOnvErted-TO.pdf){widtH="50.00000%"}
THe rEsUltS from THe 2D simUlatiOns wItH $\dELta B_0=1$ and difFErENT plaSmA bEta aRe sUmMarizEd in fIg. \[FIg8\]. DensiTy fluctuaTioNS are LaRgEr for loW beta plasma anD bEcome progrEsSivEly smaLLEr as the bEta increases. The amplitudE Of paralLel MagneTic fIeld fluctUatIons seEms NOt to be AffectEd by tHe PlaSMA beta, BUT tHe tImE for them to EXPonEntiaLlY groW is affeCted by the plasma betA.
the Proton heating ShoWs diFFErEnt STaGEs aLsO In tHESe non-monochromaTic cases. ThErE Is An initial sTRonG tEmperatUre incrEase aND a beam fOrmation, tHat correlAtEs wiTH The EnhancemenT of the paRallel magNEtic fIElD flucTuaTions. A SeCulAr staGe is foLLowEd wheRe the hEaTing raTe is sIgNificantLy reduced. The final temperAture gAined By tHe particlEs iS SimIlar to the MonoChromatic cAseS anD so arE thE TrendS witH ThE plASma beTa anD Pump wave aMPlItuDE Of $\delta T$ (and $T_\bOT$, $t_\ParAllel$). sucH Trends Are nOt affected by the plASma beta, rather, tHey aRE | known process that was fo und in 1Dkinet icsim ul atio ns [ @araneda2008pr o ton; @MatteiniEA2010]. Her e wein v esti g at e the origin an d its p os sib le re latio n t o the s trong perp end ic ular heating th at we obse rve in our simu lat ions.
T o e x plore th e pla sma he a ting a nd the pr ot o n beam formati o n ,we h ave performed the sa m e numerical ex perime nt th a t co nsi sts of a n on -mono c hromati c i n i t ial perturbationwith differ e ntplasma b eta . Altho ugh t he mon ochromaticcase provides goodi nsights into th e phys ics of par a me tr icin s tab i li tie s , i t remain san idea lize d p r oble m.Prev iousnumerical sim ula tion s us ing M HD an d ki ne tic m odels[@mal ar a2; @anna1; @Ma ttei niEA2010; @p ri mav er a_201 9 ] have ex plo red the parame t ric i n s t ab ility using non-mo no c h ro matic Al fvén w a ve p a ckets in 1 D a nd 2 D confi gura t io ns and i t is k n ow nthat it c ontinu es op era tingw itha slig ht reduc tiono f the growth r a te, essential l yb e ca u se a br oad-band sp ectr u m te ndst oque n ch co heren tr es o nances [@malara_Phy s_ fluids _1996 ].
We have r un 1D, 2Da n d 3D simu lati o ns for the non-mo nochr omatic wav e using t he sa me numer ical para m e ters and we ha veobt a i ne d similar res u l ts f or the th ree config ura tio ns(no tshown).
![Compar is on o fdif feren t macrosc op icqu ant ities for si mulat ions l ab e led with r u nC (**T op ** ). T heno rmali zeds tan dard de viation o f d e nsit yfl uctuati ons. (**Middl e* *). Parall el ma gnetic f luctuati ons and (**Bottom**) to t al temp era turediff erence fo r d iffere ntp lasmabeta c ondit io n.[ ] { data- l a be l=" Fi g8"}](Fig8 - e ps- conve rt ed-t o.pdf){ width="50.00000%"}
Th e results fro m t he 2 D si mul a ti o nswi t h $ \ d elta b_0=1$ and different p l as ma beta ar e su mm arizedin Fig. \[Fi g 8\]. De nsity flu ctuations a re l a r ger for low b eta plas ma and be c ome p r og ressi vel y smal le r a s the betai ncr eases . Theam plitud e ofpa rallel m agnetic field fluctuati ons se ems n otto be aff ect e d b y the pla smabeta, butthe ti me fo r t h em to exp o ne nti a lly g rowi s affecte d b y t h e p lasma beta. T heproto n h e atingshow s different stage s also in these non - m ono chr o mati ccases. There i s a ni n itial st ro ng temperat ure incr ea s e and a bea m form ation,t h at correl ates wi th the en han ce m ent ofth ep aralle l ma gn etic f ield f l uctu a t ions. A secularstage i s fol l owe d whe re the he a ting rate is s ignificantl y redu ced. Thefinal t em peratu rega ined by th e particle s issimilar t o th e m onochr omat i c case s an dsoare the t r e nd s w it h th e pl asmabe ta a nd pump w a ve ampli tud e of $\D el taT $ (and$ T_ \ b ot$, $T_\p ara llel$ ) . Such tren d s ar e n o t aff ectedby the plasma bet a, rather , t h e y are | known_process that_was found in 1D_kinetic simulations [@araneda2008proton;_@MatteiniEA2010]._Here we_investigate_the origin and_its possible relation_to the strong perpendicular_heating that we_observe_in our simulations.
To explore the plasma heating and the proton beam formation, we have_performed_the same_numerical_experiment_that consists of a non-monochromatic_initial perturbation with different plasma_beta. Although_the monochromatic case provides good insights into the_physics_of parametric instabilities,_it remains an idealized problem. Previous numerical simulations using_MHD and kinetic models [@malara2; @anna1; @MatteiniEA2010;_@primavera_2019] have explored_the_parametric_instability using non-monochromatic Alfvén_wave packets in 1D and 2D_configurations and it is known that_it continues operating with a slight reduction_of the growth rate, essentially because_a broad-band spectrum tends to_quench coherent_resonances [@malara_Phys_fluids_1996].
We have run 1D, 2D_and 3D simulations_for the_non-monochromatic wave using_the same numerical parameters and we_have obtained similar_results for the three configurations (not_shown).
![Comparison_of different macroscopic_quantities_for_simulations labeled_with run C_(**Top**)._The normalized_standard_deviation of density fluctuations. (**Middle**). Parallel_magnetic_fluctuations and (**Bottom**) total temperature difference for_different plasma beta condition.[]{data-label="Fig8"}](Fig8-eps-converted-to.pdf){width="50.00000%"}
The_results_from the 2D simulations_with $\delta b_0=1$ and different_plasma beta are summarized in Fig. \[Fig8\]._Density fluctuations_are larger_for low beta plasma and become progressively smaller as the beta_increases. The amplitude of parallel magnetic_field fluctuations seems not_to be_affected_by the plasma_beta,_but the_time for them to exponentially grow is_affected by_the plasma beta.
The proton heating shows_different stages also in_these_non-monochromatic cases. There is an initial_strong temperature increase and a beam_formation, that correlates with the_enhancement_of_the parallel magnetic field fluctuations._A secular stage is followed where_the heating rate_is significantly reduced. The final temperature gained_by_the particles is similar to the_monochromatic_cases and so are the trends_with_the_plasma beta and pump wave_amplitude of $\Delta T$ (and $T_\bot$,_$T_\parallel$). Such trends are not affected by the plasma_beta, rather, they_are |
h s.), the following Hamiltonian was set up $$\label{eqn:H} H=\frac{(\hat{I}_{\pi}+\hat{I}_{\nu})^{2}}{2J_{intr}}+
\frac{(\hat{I}_{\pi}-\hat{I}_{\nu})^{2}}{2J_{intr}}+\frac{1}{2}C\theta^{2},
\label{eq:hamiltonian}$$ in which the restoring force constant $C$ is related to the symmetry energy constant in the semi-empirical mass formula. Using known properties of deformed nuclei: the moment of inertia $J_{intr}$ of the ground-state band, the symmetry energy and the $B(E2;0^{+}_{1}\rightarrow2^{+}_{1}$) transition strength, one can determine both the excitation energy $E_{sc} =\hbar\sqrt{C/J_{intr}}$ and the $B(M1)$ transition strength $$\label{eqn:B} B(M1;0^{+}\rightarrow1^{+})=\frac{3}{16\pi}
\frac{J_{intr}}{\hbar^2}
E_{sc}(g_{p}-g_{n})^{2}~\mu^{2}_{N}. \label{eq:m1-moi}$$ Here $g_{p}$ and $g_{n}$ denote the orbital gyromagnetic factors associated with the rotation of the deformed proton and neutron systems, respectively. It turned out that the early calculations of both the excitation energy and the $M1$ strength gave too large values compared with the first experimental observations of $1^{+}$ scissors excitations in deformed nuclei (the $B(M1)$ value exceeds experimental values for the strongest $1^{+}$ states by a factor of almost $7$). However, it cannot be emphasized enough how important the seminal work of @LoIudice:1978 [@LoIudice:1979] has been in the experimental search which finally led to the discovery of the scissors mode in $^{156}$Gd [@Bohle:1984a]. Moreover, the confrontation of the conceptually simple two-rotor model (TRM) with the wealth of experimental data having accumulated rapidly in the late eighties and early nineties of the last century has led to a steady improvement of the model [@LoIudice:1997; @LoIudice:2000a] which is still the one allowing a good first insight in the dynamics which causes the scissors mode to show up at all | h s.), the following Hamiltonian was set up $ $ \label{eqn: H } H=\frac{(\hat{I}_{\pi}+\hat{I}_{\nu})^{2}}{2J_{intr}}+
\frac{(\hat{I}_{\pi}-\hat{I}_{\nu})^{2}}{2J_{intr}}+\frac{1}{2}C\theta^{2 },
\label{eq: hamiltonian}$$ in which the restoring force changeless $ C$ is relate to the symmetry energy constant in the semi - empirical multitude formula. Using know properties of deformed nuclei: the moment of inactiveness $ J_{intr}$ of the earth - state ring, the symmetry energy and the $ B(E2;0^{+}_{1}\rightarrow2^{+}_{1}$) conversion lastingness, one can determine both the excitation energy $ E_{sc } = \hbar\sqrt{C / J_{intr}}$ and the $ B(M1)$ transition strength $ $ \label{eqn: bacillus } B(M1;0^{+}\rightarrow1^{+})=\frac{3}{16\pi }
\frac{J_{intr}}{\hbar^2 }
E_{sc}(g_{p}-g_{n})^{2}~\mu^{2}_{N }. \label{eq: m1 - moi}$$ Here $ g_{p}$ and $ g_{n}$ denote the orbital gyromagnetic factors associated with the rotation of the deform proton and neutron systems, respectively. It turned out that the early calculations of both the excitation department of energy and the $ M1 $ strength give too large values compare with the first experimental observations of $ 1^{+}$ scissors excitations in deformed nuclei (the $ B(M1)$ value exceeds experimental values for the strongest $ 1^{+}$ states by a factor of almost $ 7 $). However, it cannot be emphasized enough how important the seminal work of @LoIudice:1978 [ @LoIudice:1979 ] has been in the experimental search which finally led to the discovery of the scissors mode in $ ^{156}$Gd [ @Bohle:1984a ]. Moreover, the confrontation of the conceptually simple two - rotor exemplar (TRM) with the wealth of experimental data having accumulated quickly in the belated eighty and early nineties of the last century has led to a steady improvement of the model [ @LoIudice:1997; @LoIudice:2000a ] which is still the one allowing a effective first insight in the dynamics which causes the scissors mode to show up at all | h s.), the following Hamiltonixn was set up $$\lcvel{eqn:I} H=\frac{(\gat{I}_{\pi}+\hag{I}_{\nu})^{2}}{2J_{intr}}+
\frac{(\hat{I}_{\pi}-\hat{I}_{\nu})^{2}}{2J_{invr}}+\frqc{1}{2}C\thtna^{2},
\label{eq:hamiltonian}$$ kn which nhe restoeing dorce consvznt $C$ is relafcd to vhe symmetry engrgy constand in the semi-ekpkrncal mass formula. Using known propereies of dfformed nuclei: the iomehn if inertia $J_{intr}$ of the grouns-state uand, the symmetty energy and the $B(E2;0^{+}_{1}\rightagrow2^{+}_{1}$) transition strenghh, one can eetewnine both thd excitation energy $E_{sd} =\hbar\sqrt{C/J_{intr}}$ and the $B(M1)$ travsitipn strengty $$\oabfn{eqn:B} B(M1;0^{+}\rigitarror1^{+})=\frac{3}{16\pi}
\frac{J_{lmtr}}{\hbas^2}
E_{sc}(g_{p}-g_{m})^{2}~\mu^{2}_{N}. \label{eq:m1-mpi}$$ Iere $g_{p}$ and $g_{n}$ denote the orbital gyromagnejic factorv cssociated with the ritqtion of dhe awfofmes 'rofon anf nxutron systsms, respectuvely. It turned out trqt the early dalculwtyons of both the excitation energy and uhe $M1$ strength gave too largw values compared witj the firft experimental observations of $1^{+}$ scissors excitathons mn deyirmed buflei (the $B(M1)$ value exceeds experimental values rot nhe strongest $1^{+}$ stctes by a factor ov sjmost $7$). Howevet, it cauhof be emphasized enlugh hor impirtant tht semonal work of @LoIudice:1978 [@LoIuduce:1979] has been un the experimentap search whnch fimally led to the discovery oy the acissors mofe in $^{156}$Gd [@Guhle:1984a]. Moreover, tfe bonfsontation of the conceptuajly simplx two-xotor moael (JRM) witr the wealhh of experimental data javiny accgmulated rwpidly in the late eighties and early ninetiex mf nhe last eenturn has led to a fteady improvekent of the oodel [@LoIusice:1997; @LoMudice:2000a] whicr is still tha one allowinj a good sirsr inwight iv the dynamics ehich causes the scussors mode to shox up zt all | h s.), the following Hamiltonian was set H=\frac{(\hat{I}_{\pi}+\hat{I}_{\nu})^{2}}{2J_{intr}}+ \label{eq:hamiltonian}$$ in the restoring force the energy constant in semi-empirical mass formula. known properties of deformed nuclei: the of inertia $J_{intr}$ of the ground-state band, the symmetry energy and the $B(E2;0^{+}_{1}\rightarrow2^{+}_{1}$) strength, one can determine both the excitation energy $E_{sc} =\hbar\sqrt{C/J_{intr}}$ and the $B(M1)$ strength B(M1;0^{+}\rightarrow1^{+})=\frac{3}{16\pi} E_{sc}(g_{p}-g_{n})^{2}~\mu^{2}_{N}. Here $g_{p}$ and $g_{n}$ denote the orbital gyromagnetic factors associated with the rotation of the deformed and neutron systems, respectively. It turned out that early calculations of both excitation energy and the $M1$ gave large values with first observations of $1^{+}$ excitations in deformed nuclei (the $B(M1)$ value exceeds experimental values for the strongest $1^{+}$ states by a of almost it cannot emphasized how the seminal work [@LoIudice:1979] has been in the experimental led to the discovery of the scissors mode $^{156}$Gd [@Bohle:1984a]. the confrontation of the conceptually simple model (TRM) with the wealth of experimental data accumulated rapidly in the late eighties and early nineties of the last century has led steady improvement of the [@LoIudice:1997; @LoIudice:2000a] which still one a first insight the dynamics which causes the scissors mode to show up at | h s.), the following Hamiltonian Was set up $$\laBel{eqN:H} H=\FraC{(\hAt{I}_{\pI}+\hat{i}_{\nu})^{2}}{2J_{intr}}+
\frac{(\haT{i}_{\pi}-\hAt{I}_{\nu})^{2}}{2J_{intr}}+\frac{1}{2}C\theta^{2},
\laBel{eq:HaMIltoNIaN}$$ in whIch the rEStORIng FoRcE coNsTAnT $C$ is rElaTed to thE symmetry eNerGy Constant in thE SeMi-empiricaL maSs formula. UsiNg kNown prOpErtIEs of dEfoRmed nUclei: tHE momenT of inertiA $J_{INtr}$ of tHE ground-STAtE banD, the symmetry energY AnD The $B(E2;0^{+}_{1}\rightarroW2^{+}_{1}$) transItIOn STRenGth, One can deteRmIne boTH the excITaTION enERgy $E_{sc} =\hbar\sqrT{C/J_{intr}}$ and tHE $B(M1)$ TransiTiOn sTRength $$\Label{EqN:b} B(M1;0^{+}\Rightarrow1^{+})=\fRac{3}{16\pI}
\frac{J_{intR}}{\hbar^2}
E_{SC}(g_{p}-g_{n})^{2}~\mu^{2}_{n}. \Label{eq:M1-moi}$$ HeRe $g_{P}$ anD $g_{n}$ dENoTe The OrBItaL GyRomAGneTic factoRs AsSociaTed wITH THe roTatIon oF the dEformed proton And NeutROn sYstemS, respEctiVeLy. It tUrned oUt thaT tHe early calculatIons Of both the ExcItAtiOn EnergY And the $m1$ stRenGth gave Too largE ValUeS COMpAred with the first exPeRIMeNtal obseRvatioNS oF $1^{+}$ sCIssors exCiTatIons IN DeforMed nUClEi (the $B(M1)$ vAlue exCEeDs ExperimEnTal valUeS foR thE stroNGest $1^{+}$ States By a factoR of alMOst $7$). However, it caNNot be emphasizED eNOUgH How iMpoRtant the semInal WOrk oF @LoIUDiCe:1978 [@LOiudicE:1979] has bEeN In THe experimental searcH wHich fiNally Led to the discoVery of the sCISSors mode In $^{156}$Gd [@bOhLE:1984a]. Moreover, the cOnfroNtation of tHE conceptUally Simple twO-rotor modEL (tRM) with tHe wEalTh oF exPERiMental data havING accUmUlated rApiDly in thE laTe eIghTieS aNd early niNeties of ThE lAsT cEntUry haS Led to a stEaDy iMpRovEment OF the moDel [@LoiudiCe:1997; @loiUdiCe:2000a] whicH Is STIll tHe OnE allOwiNg A good FirsT InsIght in tHe dynamicS whICh caUsEs The scisSors mode to shoW uP at all | h s.), the following Hamil tonian was setup$$\ la bel{ eqn: H} H=\frac{(\h a t{I} _{\pi}+\hat{I}_{\nu})^ {2}}{ 2J _ {int r }} +
\fr ac{(\ha t {I } _ {\p i} -\ hat {I } _{ \nu}) ^{2 }}{2J_{ intr}}+\fr ac{ 1} {2}C\theta^{ 2 },
\label{eq :ha miltonian}$$ in which t her estor ing forc e cons t ant $C $ is rela te d to th e symmet r y e nerg y constant in the se m i-empirical ma ss for mu l a. U sin g k nown prope rt ies o f deform e dn u c lei : the moment o f inertia $ J _{i ntr}$of th e groun d-sta te ban d, the symm etry energy a nd the $B(E2;0 ^ {+}_{1} \right arr ow2 ^{+} _ {1 }$ ) t ra n sit i on st r eng th, oneca ndeter mine b o t h th e e xcit ation energy $E_{s c}=\hb a r\s qrt{C /J_{i ntr} }$ andthe $B (M1)$ t ransition stren gth$$\label{ eqn :B } B (M 1;0^{ + }\righ tar row 1^{+})= \frac{3 } {16 \p i } \f rac{J_{intr}}{\hba r^ 2 }
E _{sc}(g_ {p}-g_ { n} )^ { 2}~\mu^{ 2} _{N }. \ l a bel{e q:m1 - mo i}$$ Her e $g_{ p }$ a nd $g_{ n} $ deno te th e o rbita l gyr omagne tic fact ors a s sociated witht he rotation o f t h e d e form edproton andneut r on s yste m s, re s pecti vely. I t t u rned out that the e ar ly cal culat ions of boththe excita t i o n energy and th e $M1$ strength gave too large values c ompar ed withthe first e xperimen tal ob ser vat i o ns of $1^{+}$ s c i ssor sexcitat ion s in de for med nu cle i(the $B(M 1)$ valu eex ce ed s e xperi m ental va lu esfo r t he st r ongest $1^{ +}$st at e s b y a fac t or o f al mo st $7$ ).Ho wever , it can not beemphasize d e n ough h ow import ant the semin al work of @ Lo Iud ice:19 7 8 [@LoIud ice:1979] has been in t h e exper ime ntalsear ch whichfin ally l edt o thediscov ery o fthe s cisso r s m ode i n $^{156}$ G d [@ Bohle :1 984a ]. More over, the confront a tio n of the conc ept uall y si mpl e t w o-r ot o r m o d el (TRM) with t he wealthof ex perimental dat ahavingaccumul atedr apidlyin the la te eighti es and e arl y nineties of thelast cent u ry ha s l ed to asteady i mpr oveme nt oft hemodel [@LoI ud ice:19 97; @ Lo Iudice:2 000a] which is still th e oneallow ing a good f irs t in sight inthedynamics w hic h c auses th e scis sors mo det o sho w up at all | h s.),_the following_Hamiltonian was set up_$$\label{eqn:H} H=\frac{(\hat{I}_{\pi}+\hat{I}_{\nu})^{2}}{2J_{intr}}+
\frac{(\hat{I}_{\pi}-\hat{I}_{\nu})^{2}}{2J_{intr}}+\frac{1}{2}C\theta^{2},
\label{eq:hamiltonian}$$_in_which the_restoring_force constant $C$_is related to_the symmetry energy constant_in the semi-empirical_mass_formula. Using known properties of deformed nuclei: the moment of inertia $J_{intr}$ of the_ground-state_band, the_symmetry_energy_and the $B(E2;0^{+}_{1}\rightarrow2^{+}_{1}$) transition strength,_one can determine both the_excitation energy_$E_{sc} =\hbar\sqrt{C/J_{intr}}$ and the $B(M1)$ transition strength $$\label{eqn:B}_B(M1;0^{+}\rightarrow1^{+})=\frac{3}{16\pi}
\frac{J_{intr}}{\hbar^2}
E_{sc}(g_{p}-g_{n})^{2}~\mu^{2}_{N}._\label{eq:m1-moi}$$ Here $g_{p}$_and $g_{n}$ denote the orbital gyromagnetic factors associated with_the rotation of the deformed proton_and neutron systems,_respectively._It_turned out that the_early calculations of both the excitation_energy and the $M1$ strength gave_too large values compared with the first_experimental observations of $1^{+}$ scissors excitations_in deformed nuclei (the $B(M1)$_value exceeds_experimental values for the strongest_$1^{+}$ states by_a factor_of almost $7$)._However, it cannot be emphasized enough_how important the_seminal work of @LoIudice:1978 [@LoIudice:1979] has_been_in the experimental_search_which_finally led_to the discovery_of_the scissors_mode_in $^{156}$Gd [@Bohle:1984a]. Moreover, the confrontation_of_the conceptually simple two-rotor model (TRM) with_the wealth of experimental_data_having accumulated rapidly in_the late eighties and early_nineties of the last century has_led to_a steady_improvement of the model [@LoIudice:1997; @LoIudice:2000a] which is still the one_allowing a good first insight in_the dynamics which causes_the scissors_mode_to show up_at_all |
delta}^{\sigma'}(j)}
{|i-j|}\,,$$ $$\label{}
C_{\alpha \sigma,\beta \sigma'}
^{\gamma \sigma,\delta \sigma'}=\sum_{i}
{\phi_{\alpha}^{\sigma*}(i)\phi_{\gamma}^{\sigma}(i)
\phi_{\beta}^{\sigma'*}(i)\phi_{\delta}^{\sigma'}(i)}\,.$$ Omitting the off-diagonal matrix elements ($(\alpha,\beta)\not=
(\gamma,\delta)$) in Eq.(7) yields the Hartree-Fock approximation. The reason for doing this will be discussed subsequently.
The basic parameters of our theoretical framework are the electron number $n$, the long-range Coulomb interaction $U$, the Hubbard interaction strength $U_{Hub}$, and the disorder amplitude $W$. Equations (3-7) constitute a set of self-consistent equations to yield the eigenstates $\phi_{\alpha}^{\sigma},\epsilon_{\alpha}^{\sigma}$ and the energy $E_{G}$ for a given set of spin occupation numbers ($n_{\uparrow},n_{\downarrow}$) with $n_{\uparrow}+n_{\downarrow}=n$. At every step of the numerical calculation, the matrix elements \[described by Eqs.(4-6)\] of $H^{eff}$ will be calculated for the spin occupation numbers that give the minimum energy in the previous step. The first Slater determinant of Eq.(3) will be formed by noninteracting eigenfunctions of Hamiltonian \[Eq.(1)\] with $U=U_{Hub}=0$. Varying the spin occupation numbers so that $M_{z}\in [0,1]$ we calculate $E_{G}(M_{z})$ whose minimum value gives the spin magnetization $M_{z}$ of the system. The stability of the self-consistent solutions is very sensitive at larger Coulomb interaction strengths, forcing us to keep a relatively small values of the interaction parameter $U$. Because of the spin rotational symmetry of the Hamiltonian, one limits the numerical calculations only to positive spin magnetization ($n_{\uparrow}\geq n_{\downarrow}$).
Increasing the amplitude of disorder the spectrum of noninteracting Hamiltonian $H_{0}=t\sum_{<i,j>,\sigma}c_{i\sigma}^{\dagger}c_{j\sigma}
+\sum_{i,\sigma}\omega_{i}n_{i,\sigma}$ evolves from weakly to strongly localized regime while the eigenvalue statistics exhibits a crossover between Wigner and Poisson distributions [@zh | delta}^{\sigma'}(j) }
{ |i - j|}\,,$$ $ $ \label { }
C_{\alpha \sigma,\beta \sigma' }
^{\gamma \sigma,\delta \sigma'}=\sum_{i }
{ \phi_{\alpha}^{\sigma*}(i)\phi_{\gamma}^{\sigma}(i)
\phi_{\beta}^{\sigma'*}(i)\phi_{\delta}^{\sigma'}(i)}\,.$$ Omitting the off - diagonal matrix elements ($ (\alpha,\beta)\not=
(\gamma,\delta)$) in Eq.(7) yields the Hartree - Fock estimate. The rationality for doing this will be discussed subsequently.
The basic argument of our theoretical model are the electron number $ n$, the long - range Coulomb interaction $ U$, the Hubbard interaction intensity $ U_{Hub}$, and the disorder amplitude $ W$. Equations (3 - 7) constitute a stage set of self - consistent equations to move over the eigenstates $ \phi_{\alpha}^{\sigma},\epsilon_{\alpha}^{\sigma}$ and the energy $ E_{G}$ for a given bent of tailspin occupation numbers ($ n_{\uparrow},n_{\downarrow}$) with $ n_{\uparrow}+n_{\downarrow}=n$. At every step of the numeric calculation, the matrix elements \[described by Eqs.(4 - 6)\ ] of $ H^{eff}$ will be calculated for the tailspin occupation numbers that give the minimum department of energy in the previous step. The first Slater determinant of Eq.(3) will be form by noninteracting eigenfunctions of Hamiltonian \[Eq.(1)\ ] with $ U = U_{Hub}=0$. Varying the spin occupation numbers so that $ M_{z}\in [ 0,1]$ we calculate $ E_{G}(M_{z})$ whose minimal value gives the spin magnetization $ M_{z}$ of the system. The stability of the self - consistent solutions is very sensible at bigger Coulomb interaction strengths, forcing us to keep a relatively small values of the interaction parameter $ U$. Because of the spin rotational symmetry of the Hamiltonian, one limits the numeric calculations only to incontrovertible tailspin magnetization ($ n_{\uparrow}\geq n_{\downarrow}$).
Increasing the amplitude of disorder the spectrum of noninteracting Hamiltonian $ H_{0}=t\sum_{<i, j>,\sigma}c_{i\sigma}^{\dagger}c_{j\sigma }
+ \sum_{i,\sigma}\omega_{i}n_{i,\sigma}$ evolves from weakly to strongly localized regime while the eigenvalue statistics show a crossover between Wigner and Poisson distributions [ @zh | delha}^{\sigma'}(j)}
{|i-j|}\,,$$ $$\label{}
C_{\alpha \slgma,\beta \sigma'}
^{\gamma \sigka,\deltz \sigma'}=\sjm_{i}
{\phi_{\alpha}^{\sigma*}(i)\phi_{\gamma}^{\sigla}(u)
\phi_{\btna}^{\sigma'*}(i)\phi_{\delta}^{\sigma'}(k)}\,.$$ Omittinh the ofd-diajonal matrix elements ($(\alika,\betz)\kot=
(\gakna,\delta)$) in Eq.(7) iields the Hdrtree-Fock appsobilation. The reason for doing this wijl be dosfussed subsequgntly.
Uhe basjb karameters of our theoretical frajework ere the electrom number $n$, the long-range Clulolb interaction $U$, tje Hubbard untewqction streneth $U_{Hub}$, and the disorser amplitude $W$. Equations (3-7) consgitutz a set of weof-cltsistent eqnationf to yield tmv eigenvtates $\lhi_{\alpha}^{\sigma},\eisiloi_{\alpya}^{\sigma}$ and the energb $E_{G}$ for a given set of spin mceupation numbers ($n_{\upaeriw},n_{\dofnarsow}$) duth $n_{\ukarcow}+h_{\downagrox}=n$. At every step of thw numerical calculauiog, the matrix emementf \[qescribed by Eqs.(4-6)\] of $H^{eff}$ will be calculdtes for the spin occupatiin numbers that give jhe minimui energy in the previous step. The first Slater dedermiiavt if Dw.(3) aill be formed by noninteracting eigenfunctioga pf Hamiltonian \[Ez.(1)\] with $U=U_{Hun}=0$. Gatiing the spin uccupacjoh numbers so that $L_{z}\in [0,1]$ wg calcylate $E_{G}(M_{e})$ whoxe minimum value gives the wpin magnetidatiin $M_{z}$ of the systel. The stabiuity of tne self-consistent solutnons ia very sensltive at mxrger Coulomb ingerscdion strtvgths, forcing us eo keep a relctively rmalk valuqs of the lnteraction parameter $U$. Hecaude of the spln rotational symmetry of the Hekiltonian, ong lhmins the nuiericsl calculatiogs only to posntive spnn magvetization ($n_{\uparrmw}\geq n_{\downwrrow}$).
Increasith the amplitnde of difordwr tye specgfum of nonintetacting Hciiltinian $H_{0}=t\sum_{<i,j>,\sigma}g_{i\sigoz}^{\dagger}c_{j\sigma}
+\sbo_{u,\sigma}\omega_{i}n_{i,\sogmx}$ edopvxs frjk weakly to vtrovglh locauized rtyimc wfile the eigenvalue stathstids exhibits a crosxoyer betwegn Wigner and Poisson cistributions [@zh | delta}^{\sigma'}(j)} {|i-j|}\,,$$ $$\label{} C_{\alpha \sigma,\beta \sigma'} ^{\gamma {\phi_{\alpha}^{\sigma*}(i)\phi_{\gamma}^{\sigma}(i) Omitting the matrix elements ($(\alpha,\beta)\not= Hartree-Fock The reason for this will be subsequently. The basic parameters of our framework are the electron number $n$, the long-range Coulomb interaction $U$, the Hubbard strength $U_{Hub}$, and the disorder amplitude $W$. Equations (3-7) constitute a set of equations yield eigenstates and the energy $E_{G}$ for a given set of spin occupation numbers ($n_{\uparrow},n_{\downarrow}$) with $n_{\uparrow}+n_{\downarrow}=n$. At step of the numerical calculation, the matrix elements by Eqs.(4-6)\] of $H^{eff}$ be calculated for the spin numbers give the energy the step. The first determinant of Eq.(3) will be formed by noninteracting eigenfunctions of Hamiltonian \[Eq.(1)\] with $U=U_{Hub}=0$. Varying the spin numbers so [0,1]$ we $E_{G}(M_{z})$ minimum gives the spin of the system. The stability of is very sensitive at larger Coulomb interaction strengths, us to a relatively small values of the parameter $U$. Because of the spin rotational symmetry the Hamiltonian, one limits the numerical calculations only to positive spin magnetization ($n_{\uparrow}\geq n_{\downarrow}$). Increasing of disorder the spectrum noninteracting Hamiltonian $H_{0}=t\sum_{<i,j>,\sigma}c_{i\sigma}^{\dagger}c_{j\sigma} evolves weakly strongly regime while eigenvalue statistics exhibits a crossover between Wigner and Poisson distributions [@zh | delta}^{\sigma'}(j)}
{|i-j|}\,,$$ $$\label{}
C_{\alpha \sIgma,\beta \siGma'}
^{\gaMma \SigMa,\DeltA \sigMa'}=\sum_{i}
{\phi_{\alpha}^{\SIgma*}(I)\phi_{\gamma}^{\sigma}(i)
\phi_{\beta}^{\Sigma'*}(I)\pHI_{\delTA}^{\sIgma'}(i)}\,.$$ omittinG ThE OFf-dIaGoNal MaTRiX elemEntS ($(\alpha,\bEta)\not=
(\gammA,\deLtA)$) in Eq.(7) yields tHE HArtree-Fock AppRoximation. ThE reAson foR dOinG This wIll Be disCussed SUbsequEntly.
The bAsIC paramETers of oUR ThEoreTical framework are THe ELectron number $n$, The lonG-rANgE cOulOmb InteractioN $U$, The HuBBard intERaCTIOn sTRength $U_{Hub}$, and The disorder AMplItude $W$. eqUatIOns (3-7) conStituTe A Set Of self-consiStenT equationS to yieLD the eigENstates $\Phi_{\alpHa}^{\sIgmA},\epsILoN_{\aLphA}^{\sIGma}$ ANd The ENerGy $E_{G}$ for a GiVeN set oF spiN OCCUpatIon NumbErs ($n_{\uParrow},n_{\downarRow}$) With $N_{\UpaRrow}+n_{\DownaRrow}=N$. AT everY step oF the nUmErical calculatiOn, thE matrix elEmeNtS \[deScRibed BY Eqs.(4-6)\] of $h^{efF}$ wiLl be calCulated FOr tHe SPIN oCcupation numbers thAt GIVe The minimUm enerGY iN tHE previouS sTep. the fIRSt SlaTer dETeRminant oF Eq.(3) wilL Be FoRmed by nOnInteraCtIng EigEnfunCTionS of HamIltonian \[eq.(1)\] witH $u=U_{Hub}=0$. Varying thE Spin occupatioN NuMBErS So thAt $M_{Z}\in [0,1]$ we calculAte $E_{g}(m_{z})$ whOse mINiMum VAlue gIves tHe SPiN Magnetization $M_{z}$ of thE sYstem. THe staBility of the seLf-consisteNT SOlutions Is veRY sENsitive at largeR CoulOmb interacTIon strenGths, fOrcing us To keep a reLATively smAll ValUes Of tHE InTeraction paraMETer $U$. beCause of The Spin rotAtiOnaL syMmeTrY of the HamIltonian, OnE lImItS thE numeRIcal calcUlAtiOnS onLy to pOSitive Spin mAgneTiZaTIon ($N_{\uparroW}\GeQ N_{\DownArRoW}$).
IncReaSiNg the AmplITudE of disoRder the spEctRUm of NoNiNteractIng HamiltoniaN $H_{0}=T\sum_{<i,j>,\sigmA}c_{I\siGma}^{\dagGER}c_{j\sigma}
+\Sum_{i,\sigma}\omega_{i}n_{i,\sigma}$ eVOlves frOm wEakly To stRongly locAliZed regIme WHile thE eigenValue StAtiSTIcs exHIBiTs a CrOssover betWEEn WIgner AnD PoiSson disTributions [@zh | delta}^{\sigma'}(j)}
{|i-j |}\,,$$ $$ \labe l{}
C_ {\ alph a \s igma,\beta \si g ma'}
^{\gamma \sigma,\delt a \si gm a '}=\ s um _{i}{\phi_{ \ al p h a}^ {\ si gma *} ( i) \phi_ {\g amma}^{ \sigma}(i)
\p hi _{\beta}^{\s i gm a'*}(i)\ph i_{ \delta}^{\si gma '}(i)} \, .$$ Omitt ing theoff-di a gonalmatrix el em e nts ($ ( \alpha, \ b et a)\n ot=
(\gamma,\delt a )$ ) in Eq.(7) yie lds th eH ar t r ee- Foc k approxim at ion.T he reas o nf o r do i ng this willbe discusse d su bseque nt ly.
The b asicpa r ame ters of our the oreticalframew o rk aret he elec tron n umb er$n$, th elon g- r ang e C oul o mbinteract io n$U$,theH u b b ardint erac tionstrength $U_{ Hub }$,a ndthe d isord er a mp litud e $W$. Equa ti ons (3-7) const itut e a set o f s el f-c on siste n t equa tio nsto yiel d the e i gen st a t e s$\phi_{\alpha}^{\s ig m a }, \epsilon _{\alp h a} ^{ \ sigma}$an d t he e n e rgy $ E_{G } $for a gi ven se t o fspin oc cu pation n umb ers ($n_ { \upa rrow}, n_{\down arrow } $) with $n_{\u p arrow}+n_{\do w na r r ow } =n$. At every step oft he n umer i ca l c a lcula tion, t h em atrix elements \[de sc ribedby Eq s.(4-6)\] of$H^{eff}$w i l l be cal cula t ed for the spin o ccupa tion numbe r s that g ive t he minim um energy i n the pr evi ous st ep. T he first Slater d eter mi nant of Eq .(3) wi llbefor med b y noninte ractingei ge nf un cti ons o f Hamilto ni an\[ Eq. (1)\] with $ U=U_{ Hub} =0 $. Var ying th e s p i n oc cu pa tion nu mb ers s o th a t $ M_{z}\i n [0,1]$wec alcu la te $E_{G} (M_{z})$ whos eminimum va lu e g ives t h e spin ma gnetization $M_{z}$ oft he syst em. Thestab ility ofthe self- con s istent solut ionsis ve r y sens i t iv e a tlarger Cou l o mbinter ac tion streng ths, forcing us to kee p a relativel y s mall v al ues of the i n ter a c tion parameter$U$. Becau se of the spinr ota ti onal sy mmetryof th e Hamilt onian, on e limitsth e nu m e ric al calcula tions on ly to pos i tives pi n mag net izatio n($n _{\up arrow} \ geq n_{\ downar ro w}$).
Incr ea sing the amplitude of disorderthe sp ectru m o f noninte rac t ing Hamilton ian$H_{0}=t\s um_ {<i ,j>,\ sig m a}c_{ i\si g ma }^{ \ dagge r}c_ { j\sigma}+ \s um_ { i ,\ sigma}\omeg a _ { i}n _{i,\ sig m a}$ ev olve s from weakly tos trongly locali zedr e gim e w h ileth e eigenvalue s tat is t i cs exhib it s a crossov er betwe en Wigne r andPoisso n distr i b ut i ons [@ zh | delta}^{\sigma'}(j)}
{|i-j|}\,,$$ $$\label{}
C_{\alpha_\sigma,\beta \sigma'}
^{\gamma_\sigma,\delta \sigma'}=\sum_{i}
{\phi_{\alpha}^{\sigma*}(i)\phi_{\gamma}^{\sigma}(i)
\phi_{\beta}^{\sigma'*}(i)\phi_{\delta}^{\sigma'}(i)}\,.$$ Omitting the_off-diagonal matrix_elements_($(\alpha,\beta)\not=
(\gamma,\delta)$) in_Eq.(7)_yields the Hartree-Fock_approximation. The reason_for doing this will_be discussed subsequently.
The_basic_parameters of our theoretical framework are the electron number $n$, the long-range Coulomb interaction_$U$,_the Hubbard_interaction_strength_$U_{Hub}$, and the disorder amplitude_$W$. Equations (3-7) constitute a_set of_self-consistent equations to yield the eigenstates $\phi_{\alpha}^{\sigma},\epsilon_{\alpha}^{\sigma}$ and_the_energy $E_{G}$ for_a given set of spin occupation numbers ($n_{\uparrow},n_{\downarrow}$) with_$n_{\uparrow}+n_{\downarrow}=n$. At every step of the_numerical calculation, the_matrix_elements_\[described by Eqs.(4-6)\] of_$H^{eff}$ will be calculated for the_spin occupation numbers that give the_minimum energy in the previous step. The_first Slater determinant of Eq.(3) will_be formed by noninteracting eigenfunctions_of Hamiltonian_\[Eq.(1)\] with $U=U_{Hub}=0$. Varying the_spin occupation numbers_so that_$M_{z}\in [0,1]$ we_calculate $E_{G}(M_{z})$ whose minimum value gives_the spin magnetization_$M_{z}$ of the system. The stability_of_the self-consistent solutions_is_very_sensitive at_larger Coulomb interaction_strengths,_forcing us_to_keep a relatively small values of_the_interaction parameter $U$. Because of the spin_rotational symmetry of the_Hamiltonian,_one limits the numerical_calculations only to positive spin_magnetization ($n_{\uparrow}\geq n_{\downarrow}$).
Increasing the amplitude of_disorder the_spectrum of_noninteracting Hamiltonian $H_{0}=t\sum_{<i,j>,\sigma}c_{i\sigma}^{\dagger}c_{j\sigma}
+\sum_{i,\sigma}\omega_{i}n_{i,\sigma}$ evolves from weakly to strongly localized regime while_the eigenvalue statistics exhibits a crossover_between Wigner and Poisson_distributions [@zh |
In Ref. [@Trifunovic2019], a complete classification of these intrinsic corner or hinge modes was derived and a higher-order bulk-boundary correspondence between these high codimension boundary modes and the topological bulk was obtained. These were accomplished by considering a $K$ subgroup series for a $d$-dimensional crystal, $$K^{(d)} \subseteq \dots \subseteq K'' \subseteq K' \subseteq K,$$ where $K\equiv K^{(0)}$ is the $K$ group which classifies the bulk band structure of Hamiltonians with coexisting order-two symmetry/antisymmetry, defined in the previous section. $K^{(n)}\subseteq K$ is a subgroup excluding topological phases of order $n$ or lower, for any crystalline-symmetry compatible boundaries. For example, $K'$ classifies the “purely crystalline phases” [@Geier2018; @Trifunovic2019], which exclude the tenfold-way topological phases, which are first-order topological phases protected by nonspatial symmetries alone and have gapless modes at any codimension-one boundaries. This purely crystalline phases can have gapless modes only when the boundary preserves the crystalline symmetry, and the gapless modes will be gapped when the crystalline symmetry is broken.
From a boundary perspective, one can define the boundary $K$ group $\mathcal{K}'$, which classifies the tenfold-way topological phases with gapless codimension-one boundary modes irrespective of boundary orientations, as long as the crystal shape and lattice termination are compatible crystalline symmetries. According to the above definitions, $\mathcal{K}'$ can be identified as the quotient group $$\mathcal{K}'= K/K'.$$ Generalizing this idea, a series of boundary $K$ groups denoted as $\mathcal{K}^{(n)}$ can be defined, which classify the intrinsic $n$-th order TI/SCs with intrinsic gapless codimension-$n$ boundary modes, when the crystal has crystalline-symmetry-compatible shape and lattice termination. In Ref. [@Trifunovic2019], the authors proved the following relation, $$\mathcal{K}^{(n+1)} = K^{(n)}/K^{(n+1)},$$ known as the higher-order bulk-boundary correspondence: an intrinsic higher-order topological phase is uniquely associated with a topologically nontrivial bulk. Moreover, the above equation provides a systematic way of obtaining the complete classification of intrinsic HOTI/SCs from $K$ subgroup series, which were computed for | In Ref. [ @Trifunovic2019 ], a complete classification of these intrinsic corner or hinge modes was derive and a eminent - order bulk - boundary correspondence between these gamey codimension boundary modes and the topological majority was obtained. These were accomplished by considering a $ K$ subgroup serial for a $ d$-dimensional quartz glass, $ $ K^{(d) } \subseteq \dots \subseteq K" \subseteq K' \subseteq K,$$ where $ K\equiv K^{(0)}$ is the $ K$ group which classifies the bulk band social organization of Hamiltonians with coexisting order - two isotropy / antisymmetry, defined in the previous section. $ K^{(n)}\subseteq K$ is a subgroup excluding topological phases of order $ n$ or lower, for any crystalline - isotropy compatible boundaries. For example, $ K'$ classifies the “ strictly crystalline phases ” [ @Geier2018; @Trifunovic2019 ], which exclude the tenfold - way topological phases, which are first - ordering topological phases protected by nonspatial symmetries alone and have gapless modes at any codimension - one limit. This purely crystalline phases can have gapless modes only when the boundary preserves the crystalline symmetry, and the gapless modes will be gapped when the crystalline symmetry is broken.
From a boundary perspective, one can define the boundary $ K$ group $ \mathcal{K}'$, which classifies the tenfold - way topological phases with gapless codimension - one boundary modes regardless of boundary orientation, as long as the quartz glass condition and wicket termination are compatible crystalline symmetries. According to the above definitions, $ \mathcal{K}'$ can be identified as the quotient group $ $ \mathcal{K}'= K / K'.$$ Generalizing this idea, a series of boundary $ K$ groups announce as $ \mathcal{K}^{(n)}$ can be defined, which classify the intrinsic $ n$-th order TI / SCs with intrinsic gapless codimension-$n$ boundary modes, when the quartz glass has crystalline - symmetry - compatible shape and lattice termination. In Ref. [ @Trifunovic2019 ], the authors proved the watch relation, $ $ \mathcal{K}^{(n+1) } = K^{(n)}/K^{(n+1)},$$ known as the higher - order bulk - boundary symmetry: an intrinsic higher - order topological phase is uniquely associated with a topologically nontrivial majority. Moreover, the above equality provides a systematic way of obtaining the complete categorization of intrinsic HOTI / SCs from $ K$ subgroup series, which were computed for | In Gef. [@Trifunovic2019], a complete classification of theve intdinsic curner or hinge modes was dermved and q higher-order bulk-bounaary corrvspondencw beuween these high rkdimensljn bkmndarv nodes and the jopological tulk was obtaitea. Chese were accomplished by considerigg a $K$ xuhgroup series sor s $d$-dijvnwional crystal, $$K^{(d)} \subseteq \dofs \substteq K'' \subseteq K' \xubseteq K,$$ where $K\equiv K^{(0)}$ ls tje $K$ group which cpassifies tye btok band struzture of Hamiltonians sith coexisting order-two symmetfy/antnsymmetry, dgyunef in the pretious fection. $K^{(n)}\sunxeteq N$ is a xubgroup exclucinj topological phases of ocder $n$ or lower, for wny crystdlmine-symmetry compqtuble toungaridw. Fur txakpme, $K'$ cpasaifies the “purely crywtalline phases” [@Geitr2018; @Egofunovic2019], whidh excjuqe the tenfold-way topological phases, whpch zre first-order topologixal phases protected hy nonspaeial symmetries alone and have gapless modes at aty covioenwijv-inf boundaries. This purely crystalline phases cwh nane gapless modes jnly when tne bpondary preservgs the edyatalline symmetry, wnd the gaplwss modes wilk be gapped when the crystaoline symmetgy iw broken.
From a bouudary perspeetive, pne csn define the boundary $Y$ grkup $\mathcal{N}'$, which cmxssifies the tenwolc-wdy topological phases with gapless rodimznsion-ond boondary iodes irredpectlee of boundary orifntatnons, ds long as the crystal shape and lattice vxrmination arg cmmpdtible cxystalkine symmetriqs. According tp the ayove ddfinitions, $\mathcan{K}'$ can be iqentified as dje quotient jroup $$\matrcal{J}'= K/K'.$$ Generaukzing this ides, a series of boundqry $K$ groups denotcd as $\jathcal{K}^{(n)}$ can bz eefined, which ckasrifr nhe igdrinsic $n$-th mrdef TK/XCs wkth intrinslc eaplrss codimension-$n$ boutdarg modes, when the ctyftal has crystaljine-symmetry-cpmpatible shape anf latvice txrminayiog. In Ref. [@Trifunovic2019], the authors proved tje nollowing relwtiok, $$\majhcal{K}^{(n+1)} = K^{(n)}/L^{(n+1)},$$ known as the higher-order bulk-boundarb correspondence: an injrinsic higher-order tpkological phaxe is uniquely associated with a tipologically nontvivial bulk. Moreover, tge aboee eqkation provides a systematic way of obtaining the complete classification of invrynsic HOTI/SDs ftom $K$ sbbgxoup sewies, xhich were computvd for | In Ref. [@Trifunovic2019], a complete classification of corner hinge modes derived and a high boundary modes and topological bulk was These were accomplished by considering a subgroup series for a $d$-dimensional crystal, $$K^{(d)} \subseteq \dots \subseteq K'' \subseteq K' K,$$ where $K\equiv K^{(0)}$ is the $K$ group which classifies the bulk band of with order-two defined in the previous section. $K^{(n)}\subseteq K$ is a subgroup excluding topological phases of order $n$ lower, for any crystalline-symmetry compatible boundaries. For example, classifies the “purely crystalline [@Geier2018; @Trifunovic2019], which exclude the topological which are topological protected nonspatial symmetries alone have gapless modes at any codimension-one boundaries. This purely crystalline phases can have gapless modes only when boundary preserves symmetry, and gapless will gapped when the is broken. From a boundary perspective, the boundary $K$ group $\mathcal{K}'$, which classifies the topological phases gapless codimension-one boundary modes irrespective of orientations, as long as the crystal shape and termination are compatible crystalline symmetries. According to the above definitions, $\mathcal{K}'$ can be identified as group $$\mathcal{K}'= K/K'.$$ Generalizing idea, a series boundary groups as can be which classify the intrinsic $n$-th order TI/SCs with intrinsic gapless codimension-$n$ modes, when the crystal has crystalline-symmetry-compatible shape and lattice termination. [@Trifunovic2019], authors proved the relation, $$\mathcal{K}^{(n+1)} = K^{(n)}/K^{(n+1)},$$ as higher-order bulk-boundary correspondence: an topological is a nontrivial Moreover, the above equation a systematic way of obtaining complete classification of intrinsic which were computed for | In Ref. [@Trifunovic2019], a complete cLassificatIon of TheSe iNtRinsIc coRner or hinge modES was Derived and a higher-order Bulk-bOuNDary COrRespoNdence bETwEEN thEsE hIgh CoDImEnsioN boUndary mOdes and the TopOlOgical bulk waS ObTained. ThesE weRe accomplishEd bY consiDeRinG A $K$ subGroUp serIes for A $D$-dimenSional cryStAL, $$K^{(d)} \subSEteq \dotS \SUbSeteQ K'' \subseteq K' \subsetEQ K,$$ WHere $K\equiv K^{(0)}$ is tHe $K$ groUp WHiCH ClaSsiFies the bulK bAnd stRUcture oF haMILTonIAns with coexisTing order-twO SymMetry/aNtIsyMMetry, dEfineD iN The Previous secTion. $k^{(n)}\subseteQ K$ is a sUBgroup eXCluding TopoloGicAl pHaseS Of OrDer $N$ oR LowER, fOr aNY crYstallinE-sYmMetry CompATIBLe boUndArieS. For eXample, $K'$ classiFieS the “PUreLy cryStallIne pHaSes” [@GeIer2018; @TriFunovIc2019], Which exclude the TenfOld-way topOloGiCal PhAses, wHIch are FirSt-oRder topOlogicaL PhaSeS PROtEcted by nonspatial sYmMETrIes alone And havE GaPlESs modes aT aNy cOdimENSion-oNe boUNdAries. ThiS purelY CrYsTalline PhAses caN hAve GapLess mODes oNly wheN the bounDary pREserves the crysTAlline symmetrY, AnD THe GApleSs mOdes will be gAppeD When The cRYsTalLIne syMmetrY iS BrOKen.
From a boundary perSpEctive, One caN define the bouNdary $K$ grouP $\MAThcal{K}'$, whIch cLAsSIfies the tenfolD-way tOpological PHases witH gaplEss codimEnsion-one BOUndary moDes IrrEspEctIVE oF boundary orieNTAtioNs, As long aS thE crystaL shApe And LatTiCe terminaTion are cOmPaTiBlE crYstalLIne symmeTrIes. acCorDing tO The aboVe defInitIoNs, $\MAthCal{K}'$ can BE iDENtifIeD aS the QuoTiEnt grOup $$\mAThcAl{K}'= K/K'.$$ GeNeralizinG thIS ideA, a SeRies of bOundary $K$ groupS dEnoted as $\maThCal{k}^{(n)}$ can bE DEfined, whIch classify the intrinsic $N$-Th order tI/Scs witH intRinsic gapLesS codimEnsIOn-$n$ bouNdary mOdes, wHeN thE CRystaL HAs CryStAlline-symmETRy-cOmpatIbLe shApe and lAttice termination. IN ref. [@trifunovic2019], the AutHors PROvEd tHE fOLloWiNG reLATion, $$\mathcal{K}^{(n+1)} = K^{(n)}/k^{(n+1)},$$ known as tHe HIgHer-order buLK-boUnDary corRespondEnce: aN IntrinsIc higher-oRder topolOgIcal PHAse Is uniquely AssociatEd with a toPOlogiCAlLy nonTriVial buLk. morEover, The aboVE eqUatioN proviDeS a systEmatiC wAy of obtaIning the complete classifIcatioN of inTriNsic HOTI/Scs fROm $K$ Subgroup sErieS, which were ComPutEd for | In Ref. [@Trifunovic2019], a complet e cla ssi fic at ionof t hese intrinsic corn er or hinge modes wasderiv ed anda h igher -orderb ul k - bou nd ar y c or r es ponde nce betwee n these hi ghco dimension bo u nd ary modesand the topolog ica l bulk w aso btain ed. Thes e were accomp lished by c o nsider i ng a $K $ su bgro up series for a $ d $- d imensional cry stal,$$ K ^{ ( d )}\su bseteq \do ts \sub s eteq K' ' \ s u b set e q K' \subsete q K,$$ wher e $K \equiv K ^{( 0 )}$ is the$K $ gr oup which c lass ifies the bulkb and str u cture o f Hami lto nia ns w i th c oex is t ing or der - two symmetr y/ an tisym metr y , d efin edin t he pr evious sectio n.$K^{ ( n)} \subs eteqK$ i sa sub groupexclu di ng topologicalphas es of ord er$n $ o rlower , for a nycry stallin e-symme t ryco m p a ti ble boundaries. Fo re x am ple, $K' $ clas s if ie s the “pu re lycrys t a lline pha s es ” [@Geie r2018; @T ri funovic 20 19], w hi chexc ludet he t enfold -way top ologi c al phases, whi c h are first-o r de r to p olog ica l phases pr otec t ed b y no n sp ati a l sym metri es al o ne and have gapless m odes a t any codimension- one bounda r i e s. Thispure l yc rystalline pha ses c an have ga p less mod es on ly whenthe bound a r y preser ves th e c rys t a ll ine symmetry, a nd t he gaples s m odes wi llbegap ped w hen the c rystalli ne s ym me try is b r oken.
F ro m a b oun daryp erspec tive, one c an def ine the bo u n dary $ K$ gro up$\ mathc al{K } '$, whichclassifie s t h e te nf ol d-way t opological ph as es with ga pl ess codim e n sion-one boundary modes irrespe c tive of bo undar y or ientation s,as lon g a s the c rystal shap eand l attic e te rmi na tion are c o m pat iblecr ysta lline s ymmetries. Accordi n g t o the above d efi niti o n s, $\ m at h cal {K } '$c a n be identified as the qu ot i en t group $$ \ mat hc al{K}'= K/K'.$ $ Gen e ralizin g this id ea, a ser ie s of b oun dary $K$ g roups de noted as$ \math c al {K}^{ (n) }$ can b e d efine d, whi c h c lassi fy the i ntrins ic $n $- th order TI/SCs with intrinsicgaples s cod ime nsion-$n$ bo u nda ry modes, whe n the crys tal ha s cry sta l line- symm e tr y-c o mpati bles hape andl at tic e te rmination.I n Ref . [@T rif u novic2 019] , the authors pro v ed the followi ng r e l ati on, $$\m at hcal{K}^{(n+1) } = K ^ { (n)}/K^{ (n +1)},$$ kno wn as th eh igher -order bulk- boundar y co r respon denc e:an intrin sic h i gher-or de rt opolog ical p hase i s uniq u elya s sociated with atopol o g icall y no ntriv ia l bulk. More over, theabove equat ion pr ovid es asystema ti c wayofob taining th e complete clas sificat io n of in trinsi c HO T I /SCsfrom $ K$subgroups e ri e s, w h ich wer e com pu tedfor | In Ref. [@Trifunovic2019],_a complete_classification of these intrinsic_corner or_hinge_modes was_derived_and a higher-order_bulk-boundary correspondence between_these high codimension boundary_modes and the_topological_bulk was obtained. These were accomplished by considering a $K$ subgroup series for a_$d$-dimensional_crystal, $$K^{(d)}_\subseteq_\dots_\subseteq K'' \subseteq K' \subseteq_K,$$ where $K\equiv K^{(0)}$ is_the $K$_group which classifies the bulk band structure of_Hamiltonians_with coexisting order-two_symmetry/antisymmetry, defined in the previous section. $K^{(n)}\subseteq K$ is_a subgroup excluding topological phases of_order $n$ or_lower,_for_any crystalline-symmetry compatible boundaries._For example, $K'$ classifies the “purely_crystalline phases” [@Geier2018; @Trifunovic2019], which exclude_the tenfold-way topological phases, which are first-order_topological phases protected by nonspatial symmetries_alone and have gapless modes_at any_codimension-one boundaries. This purely crystalline_phases can have_gapless modes_only when the_boundary preserves the crystalline symmetry, and_the gapless modes_will be gapped when the crystalline_symmetry_is broken.
From a_boundary_perspective,_one can_define the boundary_$K$_group $\mathcal{K}'$,_which_classifies the tenfold-way topological phases with_gapless_codimension-one boundary modes irrespective of boundary orientations,_as long as the_crystal_shape and lattice termination_are compatible crystalline symmetries. According_to the above definitions, $\mathcal{K}'$ can_be identified_as the_quotient group $$\mathcal{K}'= K/K'.$$ Generalizing this idea, a series of boundary_$K$ groups denoted as $\mathcal{K}^{(n)}$ can_be defined, which classify_the intrinsic_$n$-th_order TI/SCs with_intrinsic_gapless codimension-$n$_boundary modes, when the crystal has crystalline-symmetry-compatible_shape and_lattice termination. In Ref. [@Trifunovic2019], the authors_proved the following relation,_$$\mathcal{K}^{(n+1)}_= K^{(n)}/K^{(n+1)},$$ known as the higher-order_bulk-boundary correspondence: an intrinsic higher-order topological_phase is uniquely associated with_a_topologically_nontrivial bulk. Moreover, the above_equation provides a systematic way of_obtaining the complete_classification of intrinsic HOTI/SCs from $K$ subgroup_series,_which were computed for |
to the tensor product since here a tensor factor means an application of the Rota-Baxter operator $P_A$: $u_0\ot u_1= u_0P_A(u_1)$. Thus $$d_A(u_0\ot u_1)=d_A(u_0)P_A(u_1)+u_0\,d_A(P_A(u_1)) +\lambda d_A(u_0)d_A(P_A(u_1))= d(u_0)\ot u_1+u_0u_1+\lambda d(u_0)u_1.$$
$($[@GGZ; @GK3]$)$ Let $Y$ be a set with a set map $d_0\colon Y\to \bfk[Y]$ and let $(\bfk[Y],d)$ be the commutative differential algebra of weight $\lambda$ in Theorem .. The triple $(\sha
(\bfk[Y]), d_{\bfk[Y]},
P_{\bfk[Y]})$, together with $j_{\bfk[Y]} \colon \bfk [Y] \to \sha(\bfk[Y])$, is the free commutative differential Rota-Baxter $\bfk$-algebra of weight $\lambda$ on the differential algebra $(\bfk[Y],d)$.
Apply Theorem to $Y:=\Delta X$ and $d_0$ as in Theorem .. From Eq. \[eq:rbm\], the set $$\calb(\Delta X):=\left\{ u_0\ot \cdots \ot u_k\,\big|\, u_i\in
\cm(\Delta X), 0\leq i\leq k, k\geq 0\right\} \mlabel{eq:drbm}$$ is a $\bfk$-basis of the free commutative differential Rota-Baxter algebra $\sha(\Delta X)$ on the free differential algebra $(\bfk\{X\},d_X)$. We call this basis the set of [**differential Rota-Baxter (DRB) monomials**]{} on $X$. Similarly, for $n\geq 1$, apply Theorem \[thm:freediffrb\] to $Y:=\Delta_n X$ and $d_0$ as in Eq. (\[eq:diffn\]) of | to the tensor product since here a tensor factor means an application of the Rota - Baxter hustler $ P_A$: $ u_0\ot u_1= u_0P_A(u_1)$. therefore $ $ d_A(u_0\ot u_1)=d_A(u_0)P_A(u_1)+u_0\,d_A(P_A(u_1) ) + \lambda d_A(u_0)d_A(P_A(u_1))= d(u_0)\ot u_1+u_0u_1+\lambda d(u_0)u_1.$$
$ ($ [ @GGZ; @GK3]$)$ permit $ Y$ be a set with a set function $ d_0\colon Y\to \bfk[Y]$ and let $ (\bfk[Y],d)$ be the commutative differential algebra of weight $ \lambda$ in Theorem .. The ternary $ (\sha
(\bfk[Y ]), d_{\bfk[Y ] },
P_{\bfk[Y]})$, together with $ j_{\bfk[Y ] } \colon \bfk [ Y ] \to \sha(\bfk[Y])$, is the free commutative differential Rota - Baxter $ \bfk$-algebra of system of weights $ \lambda$ on the differential algebra $ (\bfk[Y],d)$.
Apply Theorem to $ Y:=\Delta X$ and $ d_0 $ as in Theorem .. From Eq. \[eq: rbm\ ], the set $ $ \calb(\Delta X):=\left\ { u_0\ot \cdots \ot u_k\,\big|\, u_i\in
\cm(\Delta X), 0\leq i\leq k, k\geq 0\right\ } \mlabel{eq: drbm}$$ is a $ \bfk$-basis of the barren commutative differential Rota - Baxter algebra $ \sha(\Delta X)$ on the free differential algebra $ (\bfk\{X\},d_X)$. We predict this footing the set of [ * * differential Rota - Baxter (DRB) monomials * * ] { } on $ X$. Similarly, for $ n\geq 1 $, apply Theorem \[thm: freediffrb\ ] to $ Y:=\Delta_n X$ and $ d_0 $ as in Eq. (\[eq: diffn\ ]) of | to the tensor product sinct here a tensor fcxtor mxans an applicagion of the Rota-Baxter operavor $P_A$: $u_0\ou u_1= u_0P_A(u_1)$. Thus $$d_A(u_0\ot u_1)=d_A(u_0)P_A(u_1)+u_0\,f_A(P_A(u_1)) +\lanbda e_A(u_0)d_A(P_A(u_1))= d(n_0)\kt u_1+u_0u_1+\lambda s(m_0)u_1.$$
$($[@GGZ; @JK3]$)$ Let $Y$ be a sgt with a sed map $d_0\colon Y\do \byk[Y]$ and let $(\bfk[Y],d)$ be the commutative differrnhial algebra os wepgrt $\lzmbda$ in Theorem .. The triple $(\sha
(\bfk[G]), d_{\bfk[Y]},
P_{\bfk[Y]})$, togrther with $j_{\bfk[Y]} \colon \bfk [Y] \tl \sha(\bfk[Y])$, is the fgee commutajjve eifferential Rota-Baxteg $\bfk$-algebrz of weight $\lambda$ on the diffefenticl algebra $(\vfj[Y],d)$.
Wkply Theorem to $Y:=\Qelta X$ and $d_0$ as in Dheorem .. From Eq. \[eq:rbm\], bhe sxt $$\cqlb(\Delta X):=\left\{ u_0\ot \cdmts \ot u_k\,\big|\, u_i\in
\ci(\Delta X), 0\nes i\leq k, k\geq 0\rigyt\} \mlabgl{eq:dsbm}$$ kw a $\bfi$-besia of tje hree commutztive diffeeential Rota-Baxter sldvnra $\sha(\Delta X)$ on ehq free differential algebra $(\bfk\{X\},d_X)$. We cdll this basis the set of [**eifferential Rota-Baxtgr (DRB) monjmials**]{} on $X$. Similarly, for $n\geq 1$, apply Theorem \[thm:fseedihffb\] uo $Y:=\Delgq_n X$ and $d_0$ as in Eq. (\[eq:diffn\]) of | to the tensor product since here a means application of Rota-Baxter operator $P_A$: u_1)=d_A(u_0)P_A(u_1)+u_0\,d_A(P_A(u_1)) d_A(u_0)d_A(P_A(u_1))= d(u_0)\ot u_1+u_0u_1+\lambda $($[@GGZ; @GK3]$)$ Let be a set with a set $d_0\colon Y\to \bfk[Y]$ and let $(\bfk[Y],d)$ be the commutative differential algebra of weight in Theorem .. The triple $(\sha (\bfk[Y]), d_{\bfk[Y]}, P_{\bfk[Y]})$, together with $j_{\bfk[Y]} \colon [Y] \sha(\bfk[Y])$, the commutative differential Rota-Baxter $\bfk$-algebra of weight $\lambda$ on the differential algebra $(\bfk[Y],d)$. Apply Theorem to $Y:=\Delta and $d_0$ as in Theorem .. From Eq. the set $$\calb(\Delta X):=\left\{ \cdots \ot u_k\,\big|\, u_i\in \cm(\Delta 0\leq k, k\geq \mlabel{eq:drbm}$$ a of the free differential Rota-Baxter algebra $\sha(\Delta X)$ on the free differential algebra $(\bfk\{X\},d_X)$. We call this basis the set [**differential Rota-Baxter on $X$. for 1$, Theorem \[thm:freediffrb\] to and $d_0$ as in Eq. (\[eq:diffn\]) | to the tensor product since heRe a tensor fActor MeaNs aN aPpliCatiOn of the Rota-BaxTEr opErator $P_A$: $u_0\ot u_1= u_0P_A(u_1)$. Thus $$d_A(U_0\ot u_1)=d_a(u_0)p_a(u_1)+u_0\,d_a(p_A(U_1)) +\lambDa d_A(u_0)d_A(p_a(u_1))= D(U_0)\Ot u_1+U_0u_1+\LaMbdA d(U_0)U_1.$$
$($[@GgZ; @GK3]$)$ LEt $Y$ Be a set wIth a set map $D_0\coLoN Y\to \bfk[Y]$ and lET $(\bFk[Y],d)$ be the cOmmUtative diffeRenTial alGeBra OF weigHt $\lAmbda$ In TheoREm .. The tRiple $(\sha
(\bFk[y]), D_{\bfk[Y]},
P_{\BFk[Y]})$, togeTHEr With $J_{\bfk[Y]} \colon \bfk [Y] \to \sHA(\bFK[Y])$, is the free comMutatiVe DIfFERenTiaL Rota-BaxteR $\bFk$-algEBra of weIGhT $\LAMbdA$ On the differenTial algebra $(\BFk[Y],D)$.
Apply thEorEM to $Y:=\DeLta X$ aNd $D_0$ As iN Theorem .. FroM Eq. \[eQ:rbm\], the seT $$\calb(\DELta X):=\lefT\{ U_0\ot \cdotS \ot u_k\,\bIg|\, u_I\in
\Cm(\DeLTa x), 0\lEq i\LeQ K, k\gEQ 0\rIghT\} \MlaBel{eq:drbM}$$ iS a $\Bfk$-baSis oF THE Free ComMutaTive dIfferential RoTa-BAxteR AlgEbra $\sHa(\DelTa X)$ oN tHe freE diffeRentiAl Algebra $(\bfk\{X\},d_X)$. We Call This basis The SeT of [**DiFfereNTial RoTa-BAxtEr (DRB) moNomials**]{} ON $X$. SImILARlY, for $n\geq 1$, apply TheorEm \[THM:fReediffrB\] to $Y:=\DeLTa_N X$ ANd $d_0$ as in EQ. (\[eQ:diFfn\]) oF | to the tensor product sin ce here atenso r f act or mea ns a n applicationo f th e Rota-Baxter operator $P_A $: $u_0 \ ot u_1= u_0P_A ( u_ 1 ) $.Th us $$ d_ A (u _0\ot u_ 1)=d_A( u_0)P_A(u_ 1)+ u_ 0\,d_A(P_A(u _ 1) ) +\lambda d_ A(u_0)d_A(P_ A(u _1))=d( u_0 ) \ot u _1+ u_0u_ 1+\lam b da d(u _0)u_1.$$
$ ($[@GG Z ; @GK3] $ ) $Let$Y$ be a set with as et map $d_0\co lon Y\ to \b f k [Y] $ a nd let $(\ bf k[Y], d )$ be t h ec o m mut a tive differen tial algebr a of weigh t$\l a mbda$in Th eo r em.. The trip le $ (\sha
(\b fk[Y]) , d_{\bf k [Y]},
P_{ \bf k[Y ]})$ , t og eth er wit h $ j_{ \ bfk [Y]} \co lo n\bfk[Y]\ t o \sha (\b fk[Y ])$,is the free c omm utat i vediffe renti al R ot a-Bax ter $\ bfk$- al gebra of weight $\l ambda$ on th edif fe renti a l alge bra $( \bfk[Y] ,d)$.
A ppl yT h e or em to $Y:=\DeltaX$ a nd $d_0$ a s in T h eo re m .. From E q.\[eq : r bm\], the se t $$\cal b(\Del t aX) :=\left \{ u_0\o t\cd ots \otu _k\, \big|\ , u_i\in
\cm( \ Delta X), 0\le q i\leq k, k\g e q0 \ ri g ht\} \m label{eq:dr bm}$ $ isa $\ b fk $-b a sis o f the f r ee commutative differe nt ial Ro ta-Ba xter algebra$\sha(\Del t a X)$ on t he f r ee differential a lgebr a $(\bfk\{ X \},d_X)$ . Wecall thi s basis t h e set of[** dif fer ent i a lRota-Baxter ( D R B) m on omials* *]{ } on $X $.Sim ila rly ,for $n\ge q 1$, ap pl yTh eo rem \[th m :freedif fr b\] t o $ Y:=\D e lta_nX$ an d $d _0 $a s i n Eq. ( \ [e q : diff n\ ]) of | to_the tensor_product since here a_tensor factor_means_an application_of_the Rota-Baxter operator_$P_A$: $u_0\ot u_1=_u_0P_A(u_1)$. Thus $$d_A(u_0\ot u_1)=d_A(u_0)P_A(u_1)+u_0\,d_A(P_A(u_1))_+\lambda d_A(u_0)d_A(P_A(u_1))= d(u_0)\ot_u_1+u_0u_1+\lambda_d(u_0)u_1.$$
$($[@GGZ; @GK3]$)$ Let $Y$ be a set with a set map $d_0\colon Y\to \bfk[Y]$_and_let $(\bfk[Y],d)$_be_the_commutative differential algebra of weight_$\lambda$ in Theorem .. The triple_$(\sha
(\bfk[Y]), d_{\bfk[Y]},
_ P_{\bfk[Y]})$, together with $j_{\bfk[Y]} \colon_\bfk_[Y] \to \sha(\bfk[Y])$,_is the free commutative differential Rota-Baxter $\bfk$-algebra of weight_$\lambda$ on the differential algebra $(\bfk[Y],d)$.
Apply_Theorem to $Y:=\Delta_X$_and_$d_0$ as in Theorem .._From Eq. \[eq:rbm\], the set $$\calb(\Delta X):=\left\{_u_0\ot \cdots \ot u_k\,\big|\, u_i\in
\cm(\Delta X),_0\leq i\leq k, k\geq 0\right\} \mlabel{eq:drbm}$$ is_a $\bfk$-basis of the free commutative_differential Rota-Baxter algebra $\sha(\Delta X)$_on the_free differential algebra $(\bfk\{X\},d_X)$. We_call this basis_the set_of [**differential Rota-Baxter_(DRB) monomials**]{} on $X$. Similarly, for_$n\geq 1$, apply_Theorem \[thm:freediffrb\] to $Y:=\Delta_n X$ and $d_0$_as_in Eq. (\[eq:diffn\]) of |
the field equation with the potential (\[varphieqa\]):$$\beta_{\mu}y^{\mu}+\int\frac{\sqrt{-\beta^{2}}}{\sqrt{2\left[ \frac{1}{2}m^{2}\varphi^{2}+U\left( \varphi\right) -\mathcal{G}\right] }}d\varphi=0.$$
Obviously, for an $n$-term polynomial potential, once a potential in the duality family is solved, the other $n$ potentials are immediately solved.
The sine-Gordon equation \[sine-Gordon\]
========================================
The sine-Gordon potential is not a polynomial potential. We first expand the sine-Gordon potential as a power series. Each term in the expansion is a power function, which has a duality discussed in section \[phin\]. Calculating the duality of each term of the series and then summing up the series give the dual potential of the sine-Gordon potential.
The duality
-----------
The sine-Gordon equation is [@Rajaraman1987solitons]$$\square\phi+\frac{m^{3}}{\sqrt{\lambda}}\sin\left( \frac{\sqrt{\lambda}}{m}\phi\right) =0. \label{SinGeq}$$ First rewrite the sine-Gordon Lagrangian, $\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{m^{4}}{\lambda}\left[
1-\cos\left( \frac{\sqrt{\lambda}}{m}\phi\right) \right] $, as$$\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{1}{2}m^{2}\phi^{2}-V\left( \phi\right)$$ with $$V\left( \phi\right) =\frac{m^{4}}{\lambda}\left[ 1-\cos\left( \frac
{\sqrt{\lambda}}{m}\phi\right) \right] -\frac{1}{2}m^{2}\phi^{2}.
\label{Vphimphi2}$$ Expanding $V\left( \phi\right) $ gives$$V\left( \phi\right) =-\frac{\lambda}{24}\phi^{4}-\sum_{n=3}^{\infty}\frac{\ | the field equation with the potential (\[varphieqa\]):$$\beta_{\mu}y^{\mu}+\int\frac{\sqrt{-\beta^{2}}}{\sqrt{2\left [ \frac{1}{2}m^{2}\varphi^{2}+U\left ( \varphi\right) -\mathcal{G}\right ] } } d\varphi=0.$$
Obviously, for an $ n$-term polynomial potential, once a electric potential in the dichotomy class is solved, the other $ n$ electric potential are immediately solved.
The sine - Gordon equation \[sine - Gordon\ ]
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
The sine - Gordon electric potential is not a polynomial potential. We first expand the sine - Gordon electric potential as a exponent series. Each terminus in the expansion is a power function, which take a duality discussed in section \[phin\ ]. Calculating the duality of each term of the serial and then summing up the series give the double potential of the sine - Gordon potential.
The dichotomy
-----------
The sine - Gordon equation is [ @Rajaraman1987solitons]$$\square\phi+\frac{m^{3}}{\sqrt{\lambda}}\sin\left ( \frac{\sqrt{\lambda}}{m}\phi\right) = 0. \label{SinGeq}$$ foremost rewrite the sine - Gordon Lagrangian, $ \mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{m^{4}}{\lambda}\left [
1-\cos\left ( \frac{\sqrt{\lambda}}{m}\phi\right) \right ] $, as$$\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{1}{2}m^{2}\phi^{2}-V\left ( \phi\right)$$ with $ $ V\left ( \phi\right) = \frac{m^{4}}{\lambda}\left [ 1-\cos\left ( \frac
{ \sqrt{\lambda}}{m}\phi\right) \right ] -\frac{1}{2}m^{2}\phi^{2 }.
\label{Vphimphi2}$$ Expanding $ V\left ( \phi\right) $ gives$$V\left ( \phi\right) = -\frac{\lambda}{24}\phi^{4}-\sum_{n=3}^{\infty}\frac{\ | thf field equation with tht potential (\[varphnwqa\]):$$\bete_{\mu}y^{\mu}+\iht\frac{\sqft{-\beta^{2}}}{\sqrt{2\left[ \frac{1}{2}m^{2}\varphi^{2}+U\pedt( \vqrphi\right) -\mathcal{G}\rieht] }}d\varihi=0.$$
Obviouwly, hor an $n$-term polbhomial ijtenflal, oucx a potential ik the dualidy family is smlxeb, the other $n$ potentials are immediaeely sokvfd.
The sine-Gordjn eatatikn \[sine-Gordon\]
========================================
The sine-Gordon potentjal is iot a polynomiak potential. We first expanf thf sine-Gordon potenhial as a piwer weries. Each germ in tht zxpansion ia a power function, which has a aualicy discussee un dgction \[phin\]. Ralculwting the duality of each trrm of the serles aid tyen summing up the secies give the dual pjtential mf the sine-Gordon pitwntian.
The duauuty
-----------
Ghe smne-Fordon eqnation is [@Rzjaraman1987solutons]$$\square\phi+\frac{m^{3}}{\xqwn{\kambda}}\sin\leff( \fras{\szrt{\lambda}}{m}\phi\right) =0. \label{SinGeq}$$ First gewrjte the sine-Gordon Lagrqngian, $\mathcal{L}=\frac{1}{2}\pattial_{\mu}\phi\[artial^{\mu}\phi-\frac{m^{4}}{\lambda}\left[
1-\cos\left( \frac{\sqrt{\lambdd}}{m}\phi\cieht) \vighg] $, as$$\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{1}{2}m^{2}\prj^{2}-V\kent( \phi\right)$$ witm $$V\left( \phi\right) =\fgav{i^{4}}{\lambda}\left[ 1-\zos\lefc( \rrac
{\sqrt{\lambda}}{m}\phi\gight) \tight] -\frac{1}{2}m^{2}\phi^{2}.
\jabek{Vphimphi2}$$ Expanding $V\left( \phi\right) $ gpves$$C\left( \phi\right) =-\fxac{\lambda}{24}\phi^{4}-\rum_{n=3}^{\onfty}\grac{\ | the field equation with the potential (\[varphieqa\]):$$\beta_{\mu}y^{\mu}+\int\frac{\sqrt{-\beta^{2}}}{\sqrt{2\left[ -\mathcal{G}\right] Obviously, for $n$-term polynomial potential, duality is solved, the $n$ potentials are solved. The sine-Gordon equation \[sine-Gordon\] ======================================== sine-Gordon potential is not a polynomial potential. We first expand the sine-Gordon potential a power series. Each term in the expansion is a power function, which a discussed section Calculating the duality of each term of the series and then summing up the series give dual potential of the sine-Gordon potential. The duality The sine-Gordon equation is \frac{\sqrt{\lambda}}{m}\phi\right) =0. \label{SinGeq}$$ First rewrite sine-Gordon $\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{m^{4}}{\lambda}\left[ 1-\cos\left( \right] as$$\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{1}{2}m^{2}\phi^{2}-V\left( with $$V\left( \phi\right) 1-\cos\left( \frac {\sqrt{\lambda}}{m}\phi\right) \right] -\frac{1}{2}m^{2}\phi^{2}. \label{Vphimphi2}$$ Expanding $V\left( \phi\right) $ gives$$V\left( \phi\right) =-\frac{\lambda}{24}\phi^{4}-\sum_{n=3}^{\infty}\frac{\ | the field equation with the poTential (\[varPhieqA\]):$$\beTa_{\mU}y^{\Mu}+\inT\fraC{\sqrt{-\beta^{2}}}{\sqrt{2\lEFt[ \frAc{1}{2}m^{2}\varphi^{2}+U\left( \varphi\riGht) -\maThCAl{G}\rIGhT] }}d\varPhi=0.$$
ObviOUsLY, For An $N$-tErm PoLYnOmial PotEntial, oNce a potentIal In The duality faMIlY is solved, tHe oTher $n$ potentiAls Are immEdIatELy solVed.
the siNe-GordON equatIon \[sine-GoRdON\]
========================================
The siNE-Gordon POTeNtiaL is not a polynomial POtENtial. We first exPand thE sINe-gORdoN poTential as a PoWer seRIes. Each TErM IN The EXpansion is a poWer function, WHicH has a dUaLitY DiscusSed in SeCTioN \[phin\]. CalculAtinG the dualiTy of eaCH term of THe serieS and thEn sUmmIng uP ThE sEriEs GIve THe DuaL PotEntial of ThE sIne-GoRdon POTENtiaL.
ThE duaLity
-----------
THe sine-Gordon eQuaTion IS [@RaJaramAn1987solItonS]$$\sQuare\Phi+\fraC{m^{3}}{\sqrT{\lAmbda}}\sin\left( \fraC{\sqrT{\lambda}}{m}\pHi\rIgHt) =0. \lAbEl{SingEq}$$ FirsT reWriTe the siNe-GordoN lagRaNGIAn, $\Mathcal{L}=\frac{1}{2}\partiaL_{\mU}\PHi\Partial^{\mU}\phi-\frAC{m^{4}}{\LaMBda}\left[
1-\cOs\LefT( \fraC{\SQrt{\laMbda}}{M}\PhI\right) \riGht] $, as$$\mAThCaL{L}=\frac{1}{2}\pArTial_{\mu}\PhI\paRtiAl^{\mu}\pHI-\fraC{1}{2}m^{2}\phi^{2}-V\Left( \phi\rIght)$$ wITh $$V\left( \phi\righT) =\Frac{m^{4}}{\lambda}\leFT[ 1-\cOS\LeFT( \fraC
{\sqRt{\lambda}}{m}\phI\rigHT) \rigHt] -\frAC{1}{2}m^{2}\Phi^{2}.
\LAbel{VPhimpHi2}$$ eXpANding $V\left( \phi\right) $ gIvEs$$V\lefT( \phi\rIght) =-\frac{\lambdA}{24}\phi^{4}-\sum_{n=3}^{\inFTY}\Frac{\ | the field equation with t he potenti al (\ [va rph ie qa\] ):$$ \beta_{\mu}y^{ \ mu}+ \int\frac{\sqrt{-\beta ^{2}} }{ \ sqrt { 2\ left[ \frac { 1} { 2 }m^ {2 }\ var ph i ^{ 2}+U\ lef t( \va rphi\right ) -\ mathcal{G}\r i gh t] }}d\va rph i=0.$$
Obvi ous ly, fo ran$ n$-te rmpolyn omialp otenti al, onceap otenti a l in th e du alit y family is solve d ,t he other $n$ p otenti al s a r e im med iately sol ve d.
T h e sine- G or d o n eq u ation \[sine- Gordon\]
== = === ====== == === = ====== ===== == = === =====
Thesine -Gordon p otenti a l is no t a poly nomial po ten tial . W efir st exp a nd th e si ne-Gordo npo tenti al a s a powe r s erie s. Ea ch term in th e e xpan s ion is a powe r fu nc tion, which hasaduality discuss ed i n section \[ ph in\ ]. Calc u lating th e d ualityof each ter mo f th e series and thensu m m in g up the serie s g iv e the dua lpot enti a l of t he s i ne -Gordonpotent i al .
The du al ity
-- -- --- --- -
Th e sin e-Gord on equat ion i s [@Rajaraman19 8 7solitons]$$\ s qu a r e\ p hi+\ fra c{m^{3}}{\s qrt{ \ lamb da}} \ si n\l e ft( \frac {\ s qr t {\lambda}}{m}\phi\r ig ht) = 0. \l abel{SinGeq}$ $ First re w r i te the s ine- G or d on Lagrangian, $\ma thcal{L}=\ f rac{1}{2 }\par tial_{\m u}\phi\pa r t ial^{\mu }\p hi- \fr ac{ m ^ {4 }}{\lambda}\l e f t[
1 -\ cos\lef t( \frac{ \sq rt{ \la mbd a} }{m}\phi\ right) \r ig ht ] $, as$$ \ mathcal{ L} =\f ra c{1 }{2}\ p artial _{\mu }\ph i\ pa r tia l^{\mu} \ ph i - \fra c{ 1} {2}m ^{2 }\ phi^{ 2}-V \ lef t( \ph i\right)$ $ w i th $ $V \l eft( \ phi\right) = \f rac{m^{4}} {\ lam bda}\l e f t[ 1-\c os\left( \frac
{\sqrt{ \ lambda} }{m }\phi \rig ht) \rig ht] -\fr ac{ 1 }{2}m^ {2}\ph i^{2} .\la b e l{Vph i m ph i2} $$ Expanding $ V\l eft( \ phi\ right) $ gives$$V\left(\ph i\right) =-\ fra c{\l a m bd a}{ 2 4} \ phi ^{ 4 }-\ s u m_{n=3}^{\infty }\frac{\ | the_field equation_with the potential (\[varphieqa\]):$$\beta_{\mu}y^{\mu}+\int\frac{\sqrt{-\beta^{2}}}{\sqrt{2\left[_ \frac{1}{2}m^{2}\varphi^{2}+U\left(__\varphi\right) _-\mathcal{G}\right]_ }}d\varphi=0.$$
Obviously, for_an $n$-term polynomial_potential, once a potential_in the duality_family_is solved, the other $n$ potentials are immediately solved.
The sine-Gordon equation \[sine-Gordon\]
========================================
The sine-Gordon potential_is_not a_polynomial_potential._We first expand the sine-Gordon_potential as a power series._Each term_in the expansion is a power function, which_has_a duality discussed_in section \[phin\]. Calculating the duality of each term_of the series and then summing_up the series_give_the_dual potential of the_sine-Gordon potential.
The duality
-----------
The sine-Gordon equation is_[@Rajaraman1987solitons]$$\square\phi+\frac{m^{3}}{\sqrt{\lambda}}\sin\left( \frac{\sqrt{\lambda}}{m}\phi\right) =0. \label{SinGeq}$$_First rewrite the sine-Gordon Lagrangian, $\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{m^{4}}{\lambda}\left[
1-\cos\left( _\frac{\sqrt{\lambda}}{m}\phi\right) \right] $, as$$\mathcal{L}=\frac{1}{2}\partial_{\mu}\phi\partial^{\mu}\phi-\frac{1}{2}m^{2}\phi^{2}-V\left(_ \phi\right)$$ with $$V\left( _\phi\right) _=\frac{m^{4}}{\lambda}\left[ 1-\cos\left( \frac
{\sqrt{\lambda}}{m}\phi\right)_ \right] _-\frac{1}{2}m^{2}\phi^{2}.
\label{Vphimphi2}$$ Expanding_$V\left( \phi\right)_ $ gives$$V\left( \phi\right) _=-\frac{\lambda}{24}\phi^{4}-\sum_{n=3}^{\infty}\frac{\ |
]{}*]{}.The experimental analysis of the inertial solution employs a single Ytterbium ion $^{171}$Yb$^+$, trapped in a six needle Paul trap schematically shown in Fig. \[ExpSche\]. The two-level-system used in our study is encoded in the hyperfine energy levels of $^{171}$Yb$^+$ represented as ${{\left| 0 \right>} \equiv \,^{2}S_{1/2}\, {\left| F=0,m_{F}=0 \right>}}$ and ${{\left| 1 \right>} \equiv \,^{2}S_{1/2}\, {\left| F=1,m_{F}=0 \right>}}$ [@Olmschenk:07]. After Doppler cooling, the system is initialized in the state ${\left| 0 \right>}$ with a standard optical pumping process. The inertial protocols are obtained by driving the hyperfine qubit with a programmable Arbitrary Waveform Generator (AWG) [@Hu:18; @Hu-18-b]. This enables to implement the components $\hat{\sigma}_{z}$ and $\hat{\sigma}_{x}$ of the Hamiltonian in Eq. . The generalized Rabi frequency $\Omega(t)$ is implemented by a simultaneous control of the microwave amplitude [@Hu:18], and the detuning between microwave frequency $\omega_{0}$ and the transition frequency $\omega_{\text{hf}}$. Where $\omega_{\text{hf}}$ is the frequency between the states ${\left| 0 \right>}$ and ${\left| 1 \right>}$ of the ion. Utilizing the AWG the time-frequency protocols of the inertial solutions are implemented. For each experimental protocol the normalized energy as a function of time is evaluated ${\langle {\hat{H}{\left({t}\right)}} \rangle}/{\langle {\hat{H}{\left({0}\right)}} \rangle}$. The measurement procedure is performed via fluorescence detection using a $369.5$ nm laser, where the population of the ${\left| 1 \right>}$ state is measured with high fidelity [@Hu:18; @Hu-18-b]. As shown in Fig. \[ExpSche\], we can detect photons for the bright state ${\left| 1 \right>}$ and zero photons for the dark state ${\left| 0 \right>}$. The measurement fidelity is estimated to be $99.4\%$ [@Hu:18; @hu2019quantum].
![Experimental apparatus | ] { } * ] { } .The experimental analysis of the inertial solution employs a single Ytterbium ion $ ^{171}$Yb$^+$, trapped in a six acerate leaf Paul ambush schematically shown in Fig. \[ExpSche\ ]. The two - level - system use in our study is encoded in the hyperfine department of energy levels of $ ^{171}$Yb$^+$ represented as $ { { \left| 0 \right > } \equiv \,^{2}S_{1/2}\, { \left| F=0,m_{F}=0 \right>}}$ and $ { { \left| 1 \right > } \equiv \,^{2}S_{1/2}\, { \left| F=1,m_{F}=0 \right>}}$ [ @Olmschenk:07 ]. After Doppler cooling, the arrangement is initialized in the state $ { \left| 0 \right>}$ with a standard ocular pumping process. The inertial protocols are receive by drive the hyperfine qubit with a programmable Arbitrary Waveform Generator (AWG) [ @Hu:18; @Hu-18 - b ]. This enable to implement the component $ \hat{\sigma}_{z}$ and $ \hat{\sigma}_{x}$ of the Hamiltonian in Eq. . The generalized Rabi frequency $ \Omega(t)$ is implement by a simultaneous control of the microwave amplitude [ @Hu:18 ], and the detuning between microwave frequency $ \omega_{0}$ and the transition frequency $ \omega_{\text{hf}}$. Where $ \omega_{\text{hf}}$ is the frequency between the state $ { \left| 0 \right>}$ and $ { \left| 1 \right>}$ of the ion. Utilizing the AWG the time - frequency protocols of the inertial solutions are implement. For each experimental protocol the normalized energy as a function of time is evaluated $ { \langle { \hat{H}{\left({t}\right) } } \rangle}/{\langle { \hat{H}{\left({0}\right) } } \rangle}$. The measurement operation is performed via fluorescence detection using a $ 369.5 $ nm laser, where the population of the $ { \left| 1 \right>}$ state is measured with high fidelity [ @Hu:18; @Hu-18 - b ]. As shown in Fig. \[ExpSche\ ], we can detect photons for the bright state $ { \left| 1 \right>}$ and zero photons for the dark state $ { \left| 0 \right>}$. The measurement fidelity is estimated to be $ 99.4\%$ [ @Hu:18; @hu2019quantum ].
! [ Experimental apparatus | ]{}*]{}.The experimental analysis on the inertial solution emplogs a sinele Ytterbium ion $^{171}$Yb$^+$, trapped ib a sux needle Paul trap scfematicalpy shown in Hig. \[ExpSche\]. The txk-level-snftem msed nn our study is gncoded in tve hyperfine etefgv levels of $^{171}$Yb$^+$ represented as ${{\left| 0 \wight>} \eaulv \,^{2}S_{1/2}\, {\left| F=0,m_{F}=0 \tight>}}$ wnd ${{\mvfu| 1 \right>} \equiv \,^{2}S_{1/2}\, {\left| F=1,m_{F}=0 \right>}}$ [@Oljschenk:07]. After Dopplet cooling, the system is inltiapized in the state ${\left| 0 \righj>}$ wiey a standard optical plkping procgss. The inertial protocols are octainzd by drivibg thf hyperfine wubit with a progvsmmabla Arbittary Waveform Neneretor (AWG) [@Hu:18; @Hu-18-b]. This enablxs to implement the somponentv $\kat{\sigma}_{z}$ and $\hat{\sigmq}_{x}$ of tve Hdmilginixn jn Es. . The henxralized Ragi frequenct $\Omega(t)$ is implememtqe by a simultzneous cjntrol of the microwave amplitude [@Hu:18], and ths detuning between micriwave frequency $\omega_{0}$ and the eransition frequency $\omega_{\text{hf}}$. Where $\omega_{\text{hx}}$ is vhd fxczjwnfy between the states ${\left| 0 \right>}$ and ${\left| 1 \ryfhu>}$ on the ion. Utilizlng the AWG the tike-vrrzuency protoculs of ths inertial solutiojs are ymplenented. Fow eavh experimental protocol thw normalized wnergy as a functiln of time ns evakuatec ${\langle {\hat{H}{\left({t}\right)}} \ranfle}/{\langle {\hwt{H}{\left({0}\rifft)}} \rangle}$. The mexsugemett procedure is performed dia fluorxscenee detecgion using a $369.5$ nm laseg, wheva the population ov the ${\laft| 1 \right>}$ state is measured with high fivxlity [@Hu:18; @Hu-18-b]. Ax vhofn in Fiy. \[ExpScme\], we can detece photons for jhe brighc statd ${\left| 1 \rifht>}$ and zero photogs for the dasn state ${\left| 0 \right>}$. Ehe neasyrement widelity is esyimated to be $99.4\%$ [@Hu:18; @hy2019quantum].
![Experimentel akpzratus | ]{}*]{}.The experimental analysis of the inertial solution single ion $^{171}$Yb$^+$, in a six in \[ExpSche\]. The two-level-system in our study encoded in the hyperfine energy levels $^{171}$Yb$^+$ represented as ${{\left| 0 \right>} \equiv \,^{2}S_{1/2}\, {\left| F=0,m_{F}=0 \right>}}$ and ${{\left| \right>} \equiv \,^{2}S_{1/2}\, {\left| F=1,m_{F}=0 \right>}}$ [@Olmschenk:07]. After Doppler cooling, the system is in state 0 with a standard optical pumping process. The inertial protocols are obtained by driving the hyperfine qubit a programmable Arbitrary Waveform Generator (AWG) [@Hu:18; @Hu-18-b]. enables to implement the $\hat{\sigma}_{z}$ and $\hat{\sigma}_{x}$ of the in . The Rabi $\Omega(t)$ implemented by a control of the microwave amplitude [@Hu:18], and the detuning between microwave frequency $\omega_{0}$ and the transition frequency Where $\omega_{\text{hf}}$ frequency between states 0 and ${\left| 1 the ion. Utilizing the AWG the the inertial solutions are implemented. For each experimental the normalized as a function of time is ${\langle {\hat{H}{\left({t}\right)}} \rangle}/{\langle {\hat{H}{\left({0}\right)}} \rangle}$. The measurement procedure performed via fluorescence detection using a $369.5$ nm laser, where the population of the ${\left| state is measured with fidelity [@Hu:18; @Hu-18-b]. shown Fig. we detect photons the bright state ${\left| 1 \right>}$ and zero photons for the state ${\left| 0 \right>}$. The measurement fidelity is estimated to [@Hu:18; ![Experimental apparatus | ]{}*]{}.The experimental analysis of The inertiaL soluTioN emPlOys a SingLe Ytterbium ion $^{171}$yB$^+$, traPped in a six needle Paul trAp schEmATicaLLy Shown In Fig. \[ExPscHE\]. the TwO-lEveL-sYStEm useD in Our studY is encoded In tHe Hyperfine eneRGy Levels of $^{171}$Yb$^+$ RepResented as ${{\leFt| 0 \rIght>} \eqUiV \,^{2}S_{1/2}\, {\lEFt| F=0,m_{F}=0 \RigHt>}}$ and ${{\Left| 1 \riGHt>} \equiV \,^{2}S_{1/2}\, {\left| F=1,m_{F}=0 \RiGHt>}}$ [@OlmsCHenk:07]. AftER doPpleR cooling, the system IS iNItialized in the State ${\lEfT| 0 \RiGHT>}$ wiTh a Standard opTiCal puMPing proCEsS. tHE inERtial protocolS are obtaineD By dRiving ThE hyPErfine Qubit WiTH a pRogrammable arbiTrary WaveForm GeNErator (Awg) [@Hu:18; @Hu-18-b]. THis enaBleS to ImplEMeNt The CoMPonENtS $\haT{\SigMa}_{z}$ and $\haT{\sIgMa}_{x}$ of The HAMILToniAn iN Eq. . THe genEralized Rabi fReqUencY $\omeGa(t)$ is ImpleMentEd By a siMultanEous cOnTrol of the microwAve aMplitude [@HU:18], anD tHe dEtUning BEtween MicRowAve freqUency $\omEGa_{0}$ aNd THE TrAnsition frequency $\oMeGA_{\TeXt{hf}}$. WherE $\omega_{\TExT{hF}}$ Is the freQuEncY betWEEn the StatES ${\lEft| 0 \right>}$ And ${\lefT| 1 \RiGhT>}$ of the iOn. utilizInG thE AWg the tIMe-frEquencY protocoLs of tHE inertial solutIOns are implemeNTeD. fOr EAch eXpeRimental proTocoL The nOrmaLIzEd eNErgy aS a funCtIOn OF time is evaluated ${\lanGlE {\hat{H}{\lEft({t}\rIght)}} \rangle}/{\lanGle {\hat{H}{\lefT({0}\RIGht)}} \ranglE}$. The MEaSUrement proceduRe is pErformed viA FluorescEnce dEtection Using a $369.5$ nm lASEr, where tHe pOpuLatIon OF ThE ${\left| 1 \right>}$ staTE Is meAsUred witH hiGh fidelIty [@hu:18; @HU-18-b]. AS shOwN in Fig. \[Expsche\], we caN dEtEcT pHotOns foR The brighT sTatE ${\lEft| 1 \Right>}$ ANd zero PhotoNs foR tHe DArk State ${\leFT| 0 \rIGHt>}$. ThE mEaSureMenT fIdeliTy is EStiMated to Be $99.4\%$ [@Hu:18; @hu2019quaNtuM].
![expeRiMeNtal appAratus | ]{}*]{}.The experimental a nalysis of theine rti al sol utio n employs a si n gleYtterbium ion $^{171}$ Yb$^+ $, trap p ed in a six ne e dl e Pau ltr apsc h em atica lly shownin Fig. \[ Exp Sc he\]. The tw o -l evel-syste m u sed in our s tud y is e nc ode d in t hehyper fine e n ergy l evels of$^ { 171}$Y b $^+$ re p r es ente d as ${{\left| 0\ ri g ht>} \equiv \, ^{2}S_ {1 / 2} \ , {\ lef t| F=0,m_{ F} =0 \r i ght>}}$ an d $ {{\ l eft| 1 \right >} \equiv \ , ^{2 }S_{1/ 2} \,{ \left| F=1, m_ { F}= 0 \right>}} $ [@ Olmschenk :07].A fter Do p pler co oling, th e s yste m i sini ti a liz e dint hestate ${ \l ef t| 0\rig h t > } $ wi tha st andar d optical pum pin g pr o ces s. Th e ine rtia lproto cols a re ob ta ined by driving the hyperfin e q ub itwi th ap rogram mab leArbitra ry Wave f orm G e n e ra tor (AWG) [@Hu:18; @ H u -1 8-b]. Th is ena b le st o implem en t t he c o m ponen ts $ \ ha t{\sigma }_{z}$ an d$\hat{\ si gma}_{ x} $ o f t he Ha m ilto nian i n Eq. .The g e neralized Rabi frequency $\O m eg a ( t) $ isimp lemented by a s i mult aneo u scon t rol o f the m i cr o wave amplitude [@Hu :1 8], an d the detuning bet ween micro w a v e freque ncy$ \o m ega_{0}$ and t he tr ansition f r equency$\ome ga_{\tex t{hf}}$.W h ere $\om ega _{\ tex t{h f } }$ is the frequ e n cy b et ween th e s tates $ {\l eft | 0 \r ig ht>}$ and ${\left |1\r ig ht> }$ of the ion. U til iz ing theA WG the time -fre qu en c y p rotocol s o f thein er tial so lu tions are imp lemente d. For ea che xper im en tal pro tocol the nor ma lized ener gy as a fun c t ion of t ime is evaluated ${\lan g le {\ha t{H }{\le ft({ t}\right) }}\rangl e}/ { \langl e {\ha t{H}{ \l eft ( { 0}\ri g h t) }}\r angle}$. T h e me asure me nt p rocedur e is performed via flu orescence det ect ionu s in g a $3 6 9.5 $n m l a s er, where the p opulationof th e ${\left| 1 \ ri ght>}$state i s mea s ured wi th high f idelity [ @H u:18 ; @Hu -18-b]. As shown i n Fig. \[ E xpSch e \] , wecan detec tpho tonsfor th e br ightstate${ \left| 1 \r ig ht>}$ an d zero photons for thedark s tate${\ left| 0 \ rig h t>} $. The me asur ement fide lit y i s est ima t ed to be$ 99 .4\ % $ [@H u:18 ; @hu2019q u an tum ] .
![Experimen t a l ap parat us | ]{}*]{}.The experimental_analysis of_the inertial solution employs_a single_Ytterbium_ion $^{171}$Yb$^+$,_trapped_in a six_needle Paul trap_schematically shown in Fig. \[ExpSche\]._The two-level-system used_in_our study is encoded in the hyperfine energy levels of $^{171}$Yb$^+$ represented as ${{\left|_0_\right>} \equiv_\,^{2}S_{1/2}\,_{\left|_F=0,m_{F}=0 \right>}}$ and ${{\left| 1_\right>} \equiv \,^{2}S_{1/2}\, {\left| F=1,m_{F}=0_\right>}}$ [@Olmschenk:07]. After_Doppler cooling, the system is initialized in the_state_${\left| 0 \right>}$_with a standard optical pumping process. The inertial protocols_are obtained by driving the hyperfine_qubit with a_programmable_Arbitrary_Waveform Generator (AWG) [@Hu:18; @Hu-18-b]._This enables to implement the components_$\hat{\sigma}_{z}$ and $\hat{\sigma}_{x}$ of the Hamiltonian_in Eq. . The generalized Rabi frequency $\Omega(t)$_is implemented by a simultaneous control_of the microwave amplitude [@Hu:18], and_the detuning_between microwave frequency $\omega_{0}$ and_the transition frequency_$\omega_{\text{hf}}$. Where_$\omega_{\text{hf}}$ is the_frequency between the states ${\left| 0_\right>}$ and ${\left|_1 \right>}$ of the ion. Utilizing_the_AWG the time-frequency_protocols_of_the inertial_solutions are implemented._For_each experimental_protocol_the normalized energy as a function_of_time is evaluated ${\langle {\hat{H}{\left({t}\right)}} \rangle}/{\langle {\hat{H}{\left({0}\right)}}_\rangle}$. The measurement procedure_is_performed via fluorescence detection_using a $369.5$ nm laser, where_the population of the ${\left| 1_\right>}$ state_is measured_with high fidelity [@Hu:18; @Hu-18-b]. As shown in Fig. \[ExpSche\], we can detect_photons for the bright state ${\left|_1 \right>}$ and zero_photons for_the_dark state ${\left|_0_\right>}$. The_measurement fidelity is estimated to be $99.4\%$_[@Hu:18; @hu2019quantum].
![Experimental_apparatus |
is indeed an element of $\mathbb{Z}\left[\zeta_k \right].$ The result now follows by equivalence (\[modified condition\]).$ \qed$
Divisibility Results
====================
We use Theorem \[criterion\], in conjunction with the evaluations of the sums $K(\chi)$ given in Section $2,$ to obtain new results concerning the divisors of $\text{gcd}(S_2(x),x^{q-1}+1).$ We first apply the evaluations of the pure Jacobi sums given in Corollary \[Jacobi pure\].
\[pure conditions\] Suppose that there exist positive integers $x$ satisfying the congruence $p^x \equiv -1 \pmod{k},$ and let $t$ be the least such integer. Hence, by Theorem [\[pure Gauss\]]{}, $m = 2ts$ for some positive integer $s$.
If $p \equiv 1 \pmod{4},$ then $I_{\beta}(x)|S_2(x) \iff s \equiv 0 \pmod{2}.$
If $p \equiv 3 \pmod{4},$ then $I_{\beta}(x)|S_2(x) \iff$ either $s \equiv 0 \pmod{2}$ or $ts$ is odd.
By Corollary \[Jacobi pure\], $K(\chi)$ is pure; in fact, $K(\chi) \in \mathbb{Z}.$ We know that $\mathcal{P}\cap \mathbb{Z} = 2\mathbb{Z}$ (see [@I1]). Hence, $$I_{\beta}(x)|S_2(x) \iff \frac{1}{2}(K(\chi)+1) \equiv 0 \pmod{2} \iff K(\chi) + 1 \equiv 0 \pmod{4}.$$
If $p \equiv 1 \pmod{4},$ then by Corollary \[Jacobi pure\], we have $$\begin{aligned}
&I_{\beta}(x)|S_2(x)& \iff (-1)^{1 + (p^t + 1)s/(2k)}p^{m/2} + 1 \equiv 0 \pmod{4} \\
&& \iff (-1)^{1 + (p^t+1)s/(2k)} + 1 \equiv 0 \pmod{4}.\end{aligned}$$ Since $k$ is odd, we have $$I_{\beta}(x)| | is indeed an element of $ \mathbb{Z}\left[\zeta_k \right].$ The result now follows by equality (\[modified condition\]).$ \qed$
Divisibility result
= = = = = = = = = = = = = = = = = = = =
We use Theorem \[criterion\ ], in conjunction with the evaluations of the sum $ K(\chi)$ given in Section $ 2,$ to prevail modern results refer the divisors of $ \text{gcd}(S_2(x),x^{q-1}+1).$ We first give the evaluations of the pure Jacobi union give in Corollary \[Jacobi pure\ ].
\[pure conditions\ ] Suppose that there exist positive integers $ x$ satisfy the congruence $ p^x \equiv -1 \pmod{k},$ and let $ t$ be the least such integer. Hence, by Theorem [ \[pure Gauss\ ] ] { }, $ m = 2ts$ for some positivist integer $ s$.
If $ p \equiv 1 \pmod{4},$ then $ I_{\beta}(x)|S_2(x) \iff s \equiv 0 \pmod{2}.$
If $ p \equiv 3 \pmod{4},$ then $ I_{\beta}(x)|S_2(x) \iff$ either $ s \equiv 0 \pmod{2}$ or $ ts$ is odd.
By Corollary \[Jacobi pure\ ], $ K(\chi)$ is pure; in fact, $ K(\chi) \in \mathbb{Z}.$ We acknowledge that $ \mathcal{P}\cap \mathbb{Z } = 2\mathbb{Z}$ (see [ @I1 ]). Hence, $ $ I_{\beta}(x)|S_2(x) \iff \frac{1}{2}(K(\chi)+1) \equiv 0 \pmod{2 } \iff K(\chi) + 1 \equiv 0 \pmod{4}.$$
If $ p \equiv 1 \pmod{4},$ then by Corollary \[Jacobi pure\ ], we have $ $ \begin{aligned }
& I_{\beta}(x)|S_2(x) & \iff (-1)^{1 + (p^t + 1)s/(2k)}p^{m/2 } + 1 \equiv 0 \pmod{4 } \\
& & \iff (-1)^{1 + (p^t+1)s/(2k) } + 1 \equiv 0 \pmod{4}.\end{aligned}$$ Since $ k$ is odd, we own $ $ I_{\beta}(x)| | is indeed an element of $\mauhbb{Z}\left[\zeta_k \riyyt].$ The resulf now foulows by equivalence (\[modifiev cobditiin\]).$ \qed$
Divisibility Resjlts
====================
We usv Theorem \[criuerion\], in conjuncvjon witm the cvaluctmons of the sumx $K(\chi)$ givan in Section $2,$ tu lbtain new results concerning the dyvisors ov $\text{gcd}(S_2(x),x^{q-1}+1).$ Re fprft aliln the evaluations of the pure Jadobi suks given in Cprollary \[Jacobi pure\].
\[pure clndihions\] Suppose that there exisj pofutive integefs $x$ satisfying the cohgruence $p^x \equiv -1 \pmod{k},$ and leg $t$ bz the least sych hnteger. Henre, by Nheorem [\[pure Nsuss\]]{}, $m = 2ts$ fot some positivc intxger $s$.
If $p \equiv 1 \pmod{4},$ thxn $I_{\beta}(x)|S_2(x) \iff s \eqoiv 0 \pmod{2}.$
Ix $' \equiv 3 \pmod{4},$ then $I_{\bwtq}(x)|S_2(x) \hff$ aithde $s \eqhit 0 \lmod{2}$ og $ta$ is odd.
By Corollary \[Hacobi pure\], $K(\chi)$ is ptgr; in fact, $K(\cgi) \in \iaehbb{Z}.$ We know that $\mathcal{P}\cap \mathbb{Z} = 2\mafhbb{Z}$ (see [@I1]). Hence, $$I_{\beta}(z)|S_2(x) \iff \frac{1}{2}(K(\chi)+1) \equig 0 \pmod{2} \isf K(\chi) + 1 \equiv 0 \pmod{4}.$$
If $p \equiv 1 \pmod{4},$ then by Cormllarb \[Gacibl pufw\], ae have $$\begin{aligned}
&I_{\beta}(x)|S_2(x)& \iff (-1)^{1 + (p^t + 1)s/(2k)}p^{m/2} + 1 \sqiin 0 \pmod{4} \\
&& \iff (-1)^{1 + (p^t+1)f/(2k)} + 1 \equiv 0 \ploc{4}.\gnd{aligned}$$ Sinze $k$ is osd, we have $$I_{\beta}(x)| | is indeed an element of $\mathbb{Z}\left[\zeta_k \right].$ now by equivalence condition\]).$ \qed$ Divisibility \[criterion\], conjunction with the of the sums given in Section $2,$ to obtain results concerning the divisors of $\text{gcd}(S_2(x),x^{q-1}+1).$ We first apply the evaluations of the Jacobi sums given in Corollary \[Jacobi pure\]. \[pure conditions\] Suppose that there exist integers satisfying congruence \equiv -1 \pmod{k},$ and let $t$ be the least such integer. Hence, by Theorem [\[pure Gauss\]]{}, = 2ts$ for some positive integer $s$. If \equiv 1 \pmod{4},$ then \iff s \equiv 0 \pmod{2}.$ $p 3 \pmod{4},$ $I_{\beta}(x)|S_2(x) either \equiv 0 \pmod{2}$ $ts$ is odd. By Corollary \[Jacobi pure\], $K(\chi)$ is pure; in fact, $K(\chi) \in \mathbb{Z}.$ We know $\mathcal{P}\cap \mathbb{Z} (see [@I1]). $$I_{\beta}(x)|S_2(x) \frac{1}{2}(K(\chi)+1) 0 \pmod{2} \iff 1 \equiv 0 \pmod{4}.$$ If $p then by Corollary \[Jacobi pure\], we have $$\begin{aligned} \iff (-1)^{1 (p^t + 1)s/(2k)}p^{m/2} + 1 \equiv \pmod{4} \\ && \iff (-1)^{1 + (p^t+1)s/(2k)} + \equiv 0 \pmod{4}.\end{aligned}$$ Since $k$ is odd, we have $$I_{\beta}(x)| | is indeed an element of $\mathbb{z}\left[\zeta_k \Right].$ the ResUlT now FollOws by equivalenCE (\[modIfied condition\]).$ \qed$
DivisIbiliTy rEsulTS
====================
WE use THeorem \[cRItERIon\], In CoNjuNcTIoN with The EvaluatIons of the sUms $k(\cHi)$ given in SecTIoN $2,$ to obtain nEw rEsults concerNinG the diViSorS Of $\texT{gcD}(S_2(x),x^{q-1}+1).$ we firsT Apply tHe evaluatIoNS of the PUre JacoBI SuMs giVen in Corollary \[JacOBi PUre\].
\[pure conditiOns\] SupPoSE tHAT thEre Exist positIvE inteGErs $x$ satISfYING thE Congruence $p^x \eQuiv -1 \pmod{k},$ anD Let $T$ be the LeAst SUch intEger. HEnCE, by theorem [\[pure gausS\]]{}, $m = 2ts$ for soMe posiTIve inteGEr $s$.
If $p \eQuiv 1 \pmOd{4},$ tHen $i_{\betA}(X)|S_2(X) \iFf s \EqUIv 0 \pMOd{2}.$
if $p \EQuiV 3 \pmod{4},$ theN $I_{\BeTa}(x)|S_2(x) \Iff$ eITHER $s \eqUiv 0 \Pmod{2}$ Or $ts$ iS odd.
By CorollaRy \[JAcobI PurE\], $K(\chi)$ Is purE; in fAcT, $K(\chi) \In \mathBb{Z}.$ We KnOw that $\mathcal{P}\cAp \maThbb{Z} = 2\mathBb{Z}$ (SeE [@I1]). HEnCe, $$I_{\beTA}(x)|S_2(x) \ifF \frAc{1}{2}(K(\Chi)+1) \equiV 0 \pmod{2} \ifF k(\chI) + 1 \eQUIV 0 \pMod{4}.$$
If $p \equiv 1 \pmod{4},$ theN bY cOrOllary \[JaCobi puRE\], wE hAVe $$\begin{aLiGneD}
&I_{\beTA}(X)|S_2(x)& \ifF (-1)^{1 + (p^t + 1)s/(2K)}P^{m/2} + 1 \Equiv 0 \pmoD{4} \\
&& \iff (-1)^{1 + (p^t+1)S/(2K)} + 1 \eQuIv 0 \pmod{4}.\eNd{AligneD}$$ SIncE $k$ iS odd, wE Have $$i_{\beta}(x)| | is indeed an element of $ \mathbb{Z} \left [\z eta _k \ri ght] .$ The resultn ow f ollows by equivalence(\[mo di f iedc on ditio n\]).$\ qe d $
D iv is ibi li t yResul ts======= ========== ===
We use Theor e m\[criterio n\] , in conjunc tio n with t hee valua tio ns of the s u ms $K( \chi)$ gi ve n in Se c tion $2 , $ t o ob tain new resultsc on c erning the div isorsof $\ t e xt{ gcd }(S_2(x),x ^{ q-1}+ 1 ).$ Wef ir s t app l y the evaluat ions of the pur e Jaco bi su m s give n inCo r oll ary \[Jacob i pu re\].
\[ pure c o ndition s \] Supp ose th atthe re e x is tpos it i vei nt ege r s $ x$ satis fy in g the con g r u e nce$p^ x \e quiv-1 \pmod{k},$ an d le t $t $ bethe l east s uch i nteger . Hen ce , by Theorem [\ [pur e Gauss\] ]{} ,$m=2ts$f or som e p osi tive in teger $ s $.
I f $ p\equiv 1 \pmod{4}, $t h en $I_{\be ta}(x) | S_ 2( x ) \iff s \ equ iv 0 \ pmod{ 2}.$
I f $p \eq uiv 3\ pm od {4},$ t he n $I_{ \b eta }(x )|S_2 ( x) \ iff$ e ither $s \equ i v 0 \pmod{2}$o r $ts$ is odd .
B y C o roll ary \[Jacobi p ure\ ] , $K (\ch i )$ is pure; in f ac t ,$ K(\chi) \in \mathbb {Z }.$ We know that $\mathc al{P}\cap\ m a thbb{Z}= 2\ m at h bb{Z}$ (see [@ I1]). Hence, $$ I _{\beta} (x)|S _2(x) \i ff \frac{ 1 } {2}(K(\c hi) +1) \e qui v 0\pmod{2} \iff K (\ch i) + 1 \e qui v 0 \pm od{ 4}. $$
If $ p \equiv1 \pmod{ 4} ,$ t he n b y Cor o llary \[ Ja cob ipur e\],w e have $$\b egin {a li g ned }
&I_{\ b et a } (x)| S_ 2( x)&\if f(-1)^ {1 + (p^ t + 1)s /(2k)}p^{ m/2 } + 1 \ eq uiv 0 \ pmod{4} \\
&& \ iff (-1)^{ 1+ ( p^t+1) s / (2k)} +1 \equiv 0 \pmod{4}.\en d {aligne d}$ $ Sin ce $ k$ is odd , w e have $$ I _{\bet a}(x)| | is_indeed an_element of $\mathbb{Z}\left[\zeta_k \right].$_The result_now_follows by_equivalence_(\[modified condition\]).$ \qed$
Divisibility_Results
====================
We use Theorem_\[criterion\], in conjunction with_the evaluations of_the_sums $K(\chi)$ given in Section $2,$ to obtain new results concerning the divisors of_$\text{gcd}(S_2(x),x^{q-1}+1).$_We first_apply_the_evaluations of the pure Jacobi_sums given in Corollary \[Jacobi_pure\].
\[pure conditions\]_Suppose that there exist positive integers $x$ satisfying_the_congruence $p^x \equiv_-1 \pmod{k},$ and let $t$ be the least such_integer. Hence, by Theorem [\[pure Gauss\]]{},_$m = 2ts$_for_some_positive integer $s$.
If $p_\equiv 1 \pmod{4},$ then $I_{\beta}(x)|S_2(x) \iff_s \equiv 0 \pmod{2}.$
If $p \equiv_3 \pmod{4},$ then $I_{\beta}(x)|S_2(x) \iff$ either $s_\equiv 0 \pmod{2}$ or $ts$ is_odd.
By Corollary \[Jacobi pure\], $K(\chi)$_is pure;_in fact, $K(\chi) \in \mathbb{Z}.$_We know that_$\mathcal{P}\cap \mathbb{Z}_= 2\mathbb{Z}$ (see_[@I1]). Hence, $$I_{\beta}(x)|S_2(x) \iff \frac{1}{2}(K(\chi)+1) \equiv_0 \pmod{2} \iff_K(\chi) + 1 \equiv 0 \pmod{4}.$$
If_$p_\equiv 1 \pmod{4},$_then_by_Corollary \[Jacobi_pure\], we have_$$\begin{aligned}
&I_{\beta}(x)|S_2(x)&_\iff (-1)^{1_+_(p^t + 1)s/(2k)}p^{m/2} + 1 \equiv_0_\pmod{4} \\
&& \iff (-1)^{1 + (p^t+1)s/(2k)} +_1 \equiv 0 \pmod{4}.\end{aligned}$$_Since_$k$ is odd, we_have $$I_{\beta}(x)| |
times that of the FFE, as indicated in Fig.\[fig2\]. Therefore, a saving in the number of steps $N_{s}$ provided by the ESSFM compared to the SSFM (as experimentally demonstrated in the next section) translates into a proportional saving in terms of latency, and an almost proportional saving in terms of complexity and power consumption.
Experimental results
====================
The experimental setup employed to compare the performance of the SSFM and ESSFM algorithms is depicted in Fig.\[fig3\]. An optical carrier, generated by a 100kHz-linewidth tunable laser source (TLS), is modulated by means of an integrated double nested Mach Zehnder modulator (IQ-MZM). Two PRBS signals of length $2^{11}-1$ at $\unit[28]{Gb/s}$ are applied to the in-phase (I) and quadrature (Q) port of the modulator to obtain a 56Gb/s QPSK optical signal. Polarization multiplexing is finally emulated through a 50/50 beam splitter, an optical delay, and a polarization beam combiner (PBC), obtaining a 112Gb/s PM-QPSK optical signal. A recirculating loop is used to emulate transmission over long distances. The loop is composed by two spans of 40km of standard single-mode fiber, each one followed by an erbium-doped fiber amplifier (EDFA). A gain equalization filter (GEF) is used to equalize the distortions due to the amplifier gain profile and a polarization scrambler (POL-S) is included in the loop to emulate random polarization rotations along the link. At the receiver, the optical signal is detected by employing coherent phase- and polarization-diversity detection and setting the local oscillator (LO) at the same nominal wavelength of the transmitter TLS (with $\pm\unit[2]{GHz}$ accuracy). The received optical signal is mixed with the LO through a polarization-diversity 90 hybrid optical coupler, whose outputs are sent to four couples of balanced photodiodes. The four photodetected signals are sampled and digitized through a 20GHz 50GSa/s real-time oscilloscope in separate blocks of one million samples at a time. Each block of samples is processed off-line according to the scheme of Fig.\[fig3\]. Bulk dispersion compensation (with a frequency-domain FFE) or DBP based on the SSFM or ESSFM algorithm | times that of the FFE, as indicated in Fig.\[fig2\ ]. Therefore, a saving in the numeral of step $ N_{s}$ provided by the ESSFM compared to the SSFM (as experimentally prove in the next section) translate into a proportional saving in terms of rotational latency, and an almost proportional saving in term of complexity and power consumption.
Experimental results
= = = = = = = = = = = = = = = = = = = =
The experimental setup employed to compare the performance of the SSFM and ESSFM algorithm is depicted in Fig.\[fig3\ ]. An optical carrier, render by a 100kHz - linewidth tunable laser source (TLS), is modulated by mean of an integrated bivalent nested Mach Zehnder modulator (IQ - MZM). Two PRBS signals of length $ 2^{11}-1 $ at $ \unit[28]{Gb / s}$ are applied to the in - phase (I) and quadrature (Q) port of the modulator to obtain a 56Gb / s QPSK ocular signal. Polarization multiplexing is finally emulated through a 50/50 beam splitter, an optical delay, and a polarization beam combiner (PBC), obtaining a 112Gb / s PM - QPSK optical signal. A recirculating loop is used to emulate transmission over long distances. The loop is composed by two spans of 40 km of standard individual - mode fiber, each one stick to by an erbium - dope fiber amplifier (EDFA). A gain equalization filter (GEF) is used to equal the distortions due to the amplifier amplification profile and a polarization scrambler (POL - S) is include in the loop to emulate random polarization rotations along the link. At the receiver, the optical signal is detected by hire coherent phase- and polarization - diversity detection and setting the local oscillator (LO) at the same nominative wavelength of the transmitter TLS (with $ \pm\unit[2]{GHz}$ accuracy). The receive ocular signal is mixed with the LO through a polarization - diversity 90 hybrid optical coupler, whose output are sent to four couples of balanced photodiodes. The four photodetected signals are sampled and digitized through a 20GHz 50GSa / s real - time oscilloscope in separate blocking of one million samples at a fourth dimension. Each block of samples is process off - line harmonize to the scheme of Fig.\[fig3\ ]. Bulk distribution compensation (with a frequency - domain FFE) or DBP free-base on the SSFM or ESSFM algorithm | tiles that of the FFE, as ikdicated in Fig.\[fnt2\]. Therxfore, a saving kn the number of steps $N_{s}$ prlvuded vy the ESSFM compared go the SSVM (as experinwntally demonstratcb in fme nert section) transkates into a proportionan raring in terms of latency, and an almoft propprhional saving yn ttrmf of bonplexity and power consumptioh.
Experikental resultx
====================
The experimental setup emoloyfd to compare the oerformance of eye SSFM and DSSFM algognthms is dekicted in Fig.\[fig3\]. An optical carrker, gznerated by a 100kHx-ninewidth tnnable laser sourcc (TLS), iv modulsted by means pf en ibtegrated double nestxd Mach Zehnder modujator (IQ-MSM). Two PRBS signals od lencth $2^{11}-1$ at $\jbit[28]{Eb/s}$ ace zpplief tk the in-phzse (I) and qyadrature (Q) port of trv modulator tk obtayn a 56Gb/s QPSK optical signal. Polarization mumtiplexing is finally enulated through a 50/50 bewm splittqr, an optical delay, and a polarization beam combiter (PUC), obualking q 112Hb/s PM-QPSK optical signal. A recirculating loo[ ix lsed to emulate tvansmission over lpnh cystances. The uoop is ckmposed by two spajs of 40ki of wtandard fingke-mode fiber, each one folloqed by an eryiun-doped fiber ampliyier (EDFA). A yain eaualiaation filter (GEF) is uszd to squalize thf distortjuns due to the aoplpfies gain profile and a polaryzation srrambker (POL-R) is incluqed in the loop to emulate random oolarnzatimn rotatiojs along the link. At the receivxc, the optical shgndl is decected by employing coherent phasg- and polcrizatkon-diversiny detectmon and settyng the local lscillator (LM) at the samw noninal wxxelength of thr transminttr TLS (wity $\pm\unit[2]{GHz}$ accuragy). Thg deceived opticao sugnal is mixed eitf tre LP evrough a poldrizxtium-divefsity 90 hybrld uptival coupler, whose oudputa are sent to four cjuples od balancqd photodiodex. The four photodeuected sijnals sre sampled and digitized through a 20GHz 50GSw/s veal-time oscijlosgope in separace blocks of one million samples at a tine. Each block of sqmples is processed pff-line accprdind to the vcheme of Fig.\[fig3\]. Bulj dispersion comptnsation (with a frequendy-domahn FFF) or DBP based on the SSFM or ESSFM algorithm | times that of the FFE, as indicated Therefore, saving in number of steps compared the SSFM (as demonstrated in the section) translates into a proportional saving terms of latency, and an almost proportional saving in terms of complexity and consumption. Experimental results ==================== The experimental setup employed to compare the performance of SSFM ESSFM is in Fig.\[fig3\]. An optical carrier, generated by a 100kHz-linewidth tunable laser source (TLS), is modulated by of an integrated double nested Mach Zehnder modulator Two PRBS signals of $2^{11}-1$ at $\unit[28]{Gb/s}$ are applied the (I) and (Q) of modulator to obtain 56Gb/s QPSK optical signal. Polarization multiplexing is finally emulated through a 50/50 beam splitter, an optical delay, a polarization (PBC), obtaining 112Gb/s optical A recirculating loop to emulate transmission over long distances. composed by two spans of 40km of standard fiber, each followed by an erbium-doped fiber amplifier A gain equalization filter (GEF) is used to the distortions due to the amplifier gain profile and a polarization scrambler (POL-S) is included loop to emulate random rotations along the At receiver, optical is detected employing coherent phase- and polarization-diversity detection and setting the local oscillator at the same nominal wavelength of the transmitter TLS (with The optical signal is with the LO through polarization-diversity hybrid optical coupler, whose sent four photodiodes. four signals are sampled and through a 20GHz 50GSa/s real-time in separate blocks of time. Each block of samples is processed off-line to the scheme of Fig.\[fig3\]. Bulk dispersion (with a frequency-domain FFE) or DBP based on the SSFM or ESSFM | times that of the FFE, as indicaTed in Fig.\[fiG2\]. TherEfoRe, a SaVing In thE number of steps $n_{S}$ proVided by the ESSFM compareD to thE SsfM (as EXpErimeNtally dEMoNSTraTeD iN thE nEXt SectiOn) tRanslatEs into a proPorTiOnal saving in TErMs of latencY, anD an almost proPorTional SaVinG In terMs oF compLexity ANd poweR consumptIoN.
experiMEntal reSULtS
====================
The Experimental setup EMpLOyed to compare tHe perfOrMAnCE Of tHe SsFM and ESSFm aLgoriTHms is dePIcTED In FIG.\[fig3\]. An optical Carrier, geneRAteD by a 100kHZ-lIneWIdth tuNable LaSEr sOurce (TLS), is mOdulAted by meaNs of an INtegratED double Nested macH ZeHndeR MoDuLatOr (iq-MZm). twO PRbs siGnals of lEnGtH $2^{11}-1$ at $\unIt[28]{Gb/S}$ ARE ApplIed To thE in-phAse (I) and quadraTurE (Q) poRT of The moDulatOr to ObTain a 56gb/s QPSk optiCaL signal. PolarizaTion MultiplexIng Is FinAlLy emuLAted thRouGh a 50/50 Beam splItter, an OPtiCaL DELaY, and a polarization bEaM COmBiner (PBC), ObtainINg A 112GB/S PM-QPSK oPtIcaL sigNAL. A recIrcuLAtIng loop iS used tO EmUlAte tranSmIssion OvEr lOng DistaNCes. THe loop Is composEd by tWO spans of 40km of stANdard single-moDE fIBEr, EAch oNe fOllowed by an ErbiUM-dopEd fiBEr AmpLIfier (eDFA). A GaIN eQUalization filter (GEF) Is Used to EqualIze the distortIons due to tHE AMplifier Gain PRoFIle and a polarizAtion Scrambler (Pol-S) is inclUded iN the loop To emulate RANdom polaRizAtiOn rOtaTIOnS along the link. aT The rEcEiver, thE opTical siGnaL is DetEctEd By employiNg cohereNt PhAsE- aNd pOlariZAtion-divErSitY dEteCtion ANd settIng thE locAl OsCIllAtor (LO) aT ThE SAme nOmInAl waVelEnGth of The tRAnsMitter TlS (with $\pm\uNit[2]{ghz}$ acCuRaCy). The reCeived optical SiGnal is mixeD wIth The LO tHROugh a polArization-diversity 90 hybriD Optical CouPler, wHose Outputs arE seNt to foUr cOUples oF balanCed phOtOdiODEs. The FOUr PhoToDetected siGNAls Are saMpLed aNd digitIzed through a 20GHz 50GSa/S ReaL-time oscillosCopE in sEPArAte BLoCKs oF oNE miLLIon samples at a tiMe. Each blocK oF SaMples is proCEssEd Off-line AccordiNg to tHE scheme Of Fig.\[fig3\]. BUlk disperSiOn coMPEnsAtion (with a FrequencY-domain FFe) Or DBP BAsEd on tHe SsFM or EsSfM aLgoriThm | times that of the FFE, as indicated in F ig. \[f ig 2\]. The refore, a savi n g in the number of steps $ N_{s} $p rovi d ed by t he ESSF M c o m par ed t o t he SS FM (a s e xperime ntally dem ons tr ated in then ex t section) tr anslates int o a propo rt ion a l sav ing in t erms o f laten cy, and a na lmostp roporti o n al sav ing in terms of c o mp l exity and powe r cons um p ti o n .
Exp erimentalre sults ======= = == = = = === = ===
The expe rimental se t upemploy ed to compar e the p e rfo rmance of t he S SFM and E SSFM a l gorithm s is dep ictedinFig .\[f i g3 \] . A no pti c al ca r rie r, gener at ed by a 100 k H z - line wid th t unabl e laser sourc e ( TLS) , is modu lated byme ans o f an i ntegr at ed double neste d Ma ch Zehnde r m od ula to r (IQ - MZM).Two PR BS sign als ofl eng th $ 2 ^{ 11}-1$ at $\unit[2 8] { G b/ s}$ areapplie d t ot he in-ph as e ( I) a n d quad ratu r e(Q) port of th e m od ulatorto obtai na 5 6Gb /s QP S K op ticalsignal.Polar i zation multipl e xing is final l ye m ul a tedthr ough a 50/5 0 be a m sp litt e r, an optic al de la y ,a nd a polarization b ea m comb iner(PBC), obtain ing a 112G b / s PM-QPSK opt i ca l signal. A rec ircul ating loop is usedto em ulate tr ansmissio n over lon g d ist anc es. T he loop is comp o s ed b ytwo spa nsof 40km of st and ard s ingle-mod e fiber, e ac hon e f ollow e d by aner biu m- dop ed fi b er amp lifie r (E DF A) . Again eq u al i z atio nfi lter (G EF ) isused toequaliz e the dis tor t ions d ue to the amplifier ga in profile a nd apolari z a tion scr ambler (POL-S) is inclu d ed in t heloopto e mulate ra ndo m pola riz a tion r otatio ns al on g t h e link . At th ereceiver,t h e o ptica lsign al is d etected by employi n g c oherent phase - a nd p o l ar iza t io n -di ve r sit y detection and s etting the l o ca l oscillat o r ( LO ) at th e samenomin a l wavel ength ofthe trans mi tter T LS(with $\pm \unit[2] {GHz}$ ac c uracy ) .The r ece ived o pt ica l sig nal is mix ed wi th the L O thro ugh a p olarizat ion-diversity 90 hybrid optic al co upl er, whose ou t put s are sen t to four coup les of bala nce d phot odio d es . T h e fou r ph o todetecte d s ign a l sare sampled a n d d igiti zed throug h a20GHz 50GSa/s rea l -time oscillos cope i n s epa r atebl ocks of one mi lli on s amples a ta time. Eac h blockof sampl es isproces sed off - l in e accor ding to the sche meof Fig.\[f ig 3\ ] . Bulk dis pe rsioncompen s atio n (with a frequenc y-dom a i n FFE ) or DBPba sed ont he S SFM or ESS FM algorith m | times_that of_the FFE, as indicated_in Fig.\[fig2\]._Therefore,_a saving_in_the number of_steps $N_{s}$ provided_by the ESSFM compared_to the SSFM_(as_experimentally demonstrated in the next section) translates into a proportional saving in terms of_latency,_and an_almost_proportional_saving in terms of complexity_and power consumption.
Experimental results
====================
The experimental_setup employed_to compare the performance of the SSFM and_ESSFM_algorithms is depicted_in Fig.\[fig3\]. An optical carrier, generated by a 100kHz-linewidth_tunable laser source (TLS), is modulated_by means of_an_integrated_double nested Mach Zehnder_modulator (IQ-MZM). Two PRBS signals of_length $2^{11}-1$ at $\unit[28]{Gb/s}$ are applied_to the in-phase (I) and quadrature (Q)_port of the modulator to obtain_a 56Gb/s QPSK optical signal._Polarization multiplexing_is finally emulated through a_50/50 beam splitter,_an optical_delay, and a_polarization beam combiner (PBC), obtaining a_112Gb/s PM-QPSK optical_signal. A recirculating loop is used_to_emulate transmission over_long_distances._The loop_is composed by_two_spans of_40km_of standard single-mode fiber, each one_followed_by an erbium-doped fiber amplifier (EDFA). A_gain equalization filter (GEF)_is_used to equalize the_distortions due to the amplifier_gain profile and a polarization scrambler_(POL-S) is_included in_the loop to emulate random polarization rotations along the link. At_the receiver, the optical signal is_detected by employing coherent_phase- and_polarization-diversity_detection and setting_the_local oscillator_(LO) at the same nominal wavelength of_the transmitter_TLS (with $\pm\unit[2]{GHz}$ accuracy). The received_optical signal is mixed_with_the LO through a polarization-diversity 90_hybrid optical coupler, whose outputs are_sent to four couples of_balanced_photodiodes._The four photodetected signals are_sampled and digitized through a 20GHz_50GSa/s real-time oscilloscope_in separate blocks of one million samples_at_a time. Each block of samples_is_processed off-line according to the scheme_of_Fig.\[fig3\]._Bulk dispersion compensation (with a_frequency-domain FFE) or DBP based on_the SSFM or ESSFM algorithm |
theory of Kohno connections [@rif65; @rif67]. If a connection $\mathcal{D}=d+A$ of the form \[berkohno\] satisfies $\mathcal{D}^{2}=0$, then the following relations hold
$$\begin{split}
\left[ B_{ij},B_{ik}+B_{jk}\right] = \left[ B_{ij}+B_{ik},B_{jk}\right] =0, \hspace{1cm} & \mathrm{for} \ i<j<k, \\
\left[ B_{ij},B_{kl}\right] =0, \hspace{5cm} & \mathrm{for \ distinct \ i,j,k,l.}
\end{split}$$
The above equations are called the infinitesimal pure braid relations. A connection of the form \[berkohno\] where the constant $d\times d$ matrices $B_{ij}$ satisfy the above relations is called a rank-$d$ Kohno connection. A connection of this type defines a representation of the pure braid group $P_{d}$ of $d$ strings
$$\sigma: P_{d}\rightarrow GL(d,\mathbb{C})$$
through the parallel transport on the configuration space $C_{d}$
$$\sigma: \gamma \in C_{d}\rightarrow P \exp (-\int_{\gamma} A) \in GL(d,\mathbb{C}).$$
The family of representations $\sigma$ parametrized by the matrices $B_{ij}$ satisfying the infinitesimal pure braid relations is called the universal monodromy [@rif67]. We conclude that for a very complete $tt^{*}$ geometry the UV Berry connection is the universal monodromy representation of the pure braid group $P_{d}$ of $d$ strings specialized to Kohno matrices $B_{ij}$ computed in terms of the $U(1)_{R}$ spectrum of the UV chiral ring. A very important class of very complete $tt^{*}$ geometries are the symmetric ones. In this case the Kohno matrices satisfy
$$(B_{\pi(i)\pi(j)})_{\pi(k)\pi(l)}=(B_{ij})_{kl}, \hspace{1cm} \mathrm{for \ all \ } \pi \in S_{d}$$
In this case the connection $\mathcal{D}$ descends to | theory of Kohno connections [ @rif65; @rif67 ]. If a connection $ \mathcal{D}=d+A$ of the form \[berkohno\ ] satisfies $ \mathcal{D}^{2}=0 $, then the follow relative hold
$ $ \begin{split }
\left [ B_{ij},B_{ik}+B_{jk}\right ] = \left [ B_{ij}+B_{ik},B_{jk}\right ] = 0, \hspace{1 cm } & \mathrm{for } \ i < j < k, \\
\left [ B_{ij},B_{kl}\right ] = 0, \hspace{5 cm } & \mathrm{for \ distinct \ i, j, k, l. }
\end{split}$$
The above equations are called the infinitesimal saturated braid relations. A connection of the shape \[berkohno\ ] where the ceaseless $ d\times d$ matrices $ B_{ij}$ satisfy the above relation is called a rank-$d$ Kohno connection. A joining of this type defines a theatrical performance of the pure braid group $ P_{d}$ of $ d$ strings
$ $ \sigma: P_{d}\rightarrow GL(d,\mathbb{C})$$
through the parallel transport on the configuration space $ C_{d}$
$ $ \sigma: \gamma \in C_{d}\rightarrow P \exp (-\int_{\gamma } A) \in GL(d,\mathbb{C}).$$
The family of representation $ \sigma$ parametrized by the matrices $ B_{ij}$ satisfying the infinitesimal pure braid relation is called the universal monodromy [ @rif67 ]. We conclude that for a very complete $ tt^{*}$ geometry the ultraviolet Berry connection is the cosmopolitan monodromy representation of the pure braid group $ P_{d}$ of $ d$ strings specialized to Kohno matrix $ B_{ij}$ computed in terms of the $ U(1)_{R}$ spectrum of the UV chiral ring. A very important course of very accomplished $ tt^{*}$ geometries are the symmetric ones. In this case the Kohno matrices satisfy
$ $ (B_{\pi(i)\pi(j)})_{\pi(k)\pi(l)}=(B_{ij})_{kl }, \hspace{1 cm } \mathrm{for \ all \ } \pi \in S_{d}$$
In this case the connection $ \mathcal{D}$ descends to | thfory of Kohno connectionr [@rif65; @rif67]. If a einnectmon $\matgcal{D}=d+A$ uf the form \[berkohno\] satisfixs $\mqthcao{D}^{2}=0$, then the following felations hold
$$\begun{spout}
\left[ B_{mn},B_{ik}+B_{jk}\vnght] = \peft[ U_{ij}+B_{ik},B_{jk}\right] =0, \hspace{1cm} & \mathrm{for} \ i<j<n, \\
\lzft[ B_{ij},B_{kl}\right] =0, \hspace{5cm} & \mathrm{for \ distimch \ i,j,k,l.}
\end{splij}$$
The sfove vqmations are called the infinitesjmal puge braid relationx. A connection of the form \[bernohno\] where the cojstant $d\timga d$ natrices $B_{ij}$ satisfy the above relztions is called a rank-$d$ Kohno zonneetion. A conbextilt of this tbpe desines a reprcxentathon of yhe pure braid grmup $P_{d}$ of $d$ strings
$$\sigma: P_{d}\rightarrow GL(d,\mwthbb{C})$$
thrmuyh the parallel transpoet on the conwugufatjoi slace $C_{f}$
$$\sijma: \gamma \ih C_{d}\rightareow P \exp (-\int_{\gamma} S) \yb GL(d,\mathbb{C}).$$
Tge famylr of representations $\sigma$ parametrized by the matrices $B_{ij}$ satisdying the infinitesimwl pure bwaid relations is called the universal monodromy [@sif67]. Wx zoneoude gyah for a very complete $tt^{*}$ geometry the UV Berrr cpnkection is the ukiversal monodromy rfptgsentation of jhe purz bdaid group $P_{d}$ of $d$ stringf spexialized uo Konno matrices $B_{ij}$ computed ib terms of tke $Y(1)_{R}$ spectrum of the UV chiral xing. A very important class of verv compmete $tt^{*}$ geoletries add the symmetric unex. Hn this case the Kohno matwices satmsfy
$$(B_{\'i(i)\pi(j)})_{\pi(y)\pi(l)}=(N_{ij})_{kl}, \rspace{1cm} \mwthrm{nmr \ all \ } \pi \in S_{d}$$
Ln thns cave the conjection $\mathcal{D}$ descends to | theory of Kohno connections [@rif65; @rif67]. If $\mathcal{D}=d+A$ the form satisfies $\mathcal{D}^{2}=0$, then \left[ = \left[ B_{ij}+B_{ik},B_{jk}\right] \hspace{1cm} & \mathrm{for} i<j<k, \\ \left[ B_{ij},B_{kl}\right] =0, \hspace{5cm} \mathrm{for \ distinct \ i,j,k,l.} \end{split}$$ The above equations are called the infinitesimal braid relations. A connection of the form \[berkohno\] where the constant $d\times d$ $B_{ij}$ the relations called a rank-$d$ Kohno connection. A connection of this type defines a representation of the pure group $P_{d}$ of $d$ strings $$\sigma: P_{d}\rightarrow GL(d,\mathbb{C})$$ the parallel transport on configuration space $C_{d}$ $$\sigma: \gamma C_{d}\rightarrow \exp (-\int_{\gamma} \in The of representations $\sigma$ by the matrices $B_{ij}$ satisfying the infinitesimal pure braid relations is called the universal monodromy [@rif67]. We that for complete $tt^{*}$ the Berry is the universal of the pure braid group $P_{d}$ specialized to Kohno matrices $B_{ij}$ computed in terms the $U(1)_{R}$ of the UV chiral ring. A important class of very complete $tt^{*}$ geometries are symmetric ones. In this case the Kohno matrices satisfy $$(B_{\pi(i)\pi(j)})_{\pi(k)\pi(l)}=(B_{ij})_{kl}, \hspace{1cm} \mathrm{for \ all \ \in S_{d}$$ In this the connection $\mathcal{D}$ to | theory of Kohno connections [@rIf65; @rif67]. If a coNnectIon $\MatHcAl{D}=d+a$ of tHe form \[berkohno\] SAtisFies $\mathcal{D}^{2}=0$, then the folLowinG rELatiONs Hold
$$\bEgin{splIT}
\lEFT[ B_{iJ},B_{Ik}+b_{jk}\RiGHt] = \Left[ B_{Ij}+B_{Ik},B_{jk}\riGht] =0, \hspace{1cM} & \maThRm{for} \ i<j<k, \\
\left[ b_{Ij},b_{kl}\right] =0, \hsPacE{5cm} & \mathrm{for \ DisTinct \ i,J,k,L.}
\enD{Split}$$
the Above EquatiONs are cAlled the iNfINitesiMAl pure bRAId RelaTions. A connection oF ThE Form \[berkohno\] whEre the CoNStANT $d\tImeS d$ matrices $b_{iJ}$ satiSFy the abOVe RELAtiONs is called a raNk-$d$ Kohno conNEctIon. A coNnEctIOn of thIs typE dEFinEs a represenTatiOn of the puRe braiD Group $P_{d}$ OF $d$ strinGs
$$\sigmA: P_{d}\RigHtarROw gL(D,\maThBB{C})$$
tHRoUgh THe pArallel tRaNsPort oN the CONFIgurAtiOn spAce $C_{d}$
$$\Sigma: \gamma \in C_{D}\riGhtaRRow p \exp (-\iNt_{\gamMa} A) \iN Gl(d,\matHbb{C}).$$
ThE famiLy Of representatioNs $\siGma$ parameTriZeD by ThE matrICes $B_{ij}$ SatIsfYing the InfinitESimAl PURE bRaid relations is calLeD THe UniversaL monodROmY [@rIF67]. We conclUdE thAt foR A Very cOmplETe $Tt^{*}$ geometRy the Uv beRrY connecTiOn is thE uNivErsAl monODromY repreSentatioN of thE Pure braid group $p_{D}$ of $d$ strings spECiALIzED to KOhnO matrices $B_{iJ}$ comPUted In teRMs Of tHE $U(1)_{R}$ spEctruM oF ThE uV chiral ring. A very imPoRtant cLass oF very complete $Tt^{*}$ geometriES ARe the symMetrIC oNEs. In this case thE KohnO matrices sATisfy
$$(B_{\pi(I)\pi(j)})_{\pI(k)\pi(l)}=(B_{ij})_{Kl}, \hspace{1cM} \MAthrm{for \ All \ } \Pi \iN S_{d}$$
in tHIS cAse the connectION $\matHcAl{D}$ descEndS to | theory of Kohno connectio ns [@rif65 ; @ri f67 ].If a c onne ction $\mathca l {D}= d+A$ of the form \[ber kohno \] sati s fi es $\ mathcal { D} ^ { 2}= 0$ ,the nt he foll owi ng rela tions hold
$ $\ begin{split} \l eft[ B_{ ij} ,B_{ik}+B_{j k}\ right] = \l e ft[ B _{i j}+B_ {ik},B _ {jk}\r ight] =0, \ h space{ 1 cm} & \ m a th rm{f or} \ i<j<k, \\ \ le f t[ B_{ij},B_{k l}\rig ht ] = 0 , \h spa ce{5cm} &\m athrm { for \ d i st i n c t \ i,j,k,l.}
\en d{split}$$The above e qua t ions a re ca ll e d t he infinite sima l pure br aid re l ations. A conne ctionofthe for m \ [b erk oh n o\] wh ere the constan t$d \time s d$ m a t rice s $ B_{i j}$ s atisfy the ab ove rel a tio ns is call ed a r ank-$ d$ Koh no co nn ection. A conne ctio n of this ty pe de fi nes a repres ent ati on of t he pure bra id g r ou p $P_{d}$ of $d$ s tr i n gs
$$\sig ma: P_ { d} \r i ghtarrow G L(d ,\ma t h bb{C} )$$th rough th e para l le ltranspo rt on th econ fig urati o n sp ace $C _{d}$
$ $\sig m a: \gamma \inC _{d}\rightarr o wP \e x p (- \in t_{\gamma}A) \ i n GL (d,\ m at hbb { C}).$ $
Th ef am i ly of representatio ns $\sig ma$ p arametrized b y the matr i c e s $B_{ij }$ s a ti s fying the infi nites imal pureb raid rel ation s is cal led the u n i versal m ono dro my[@r i f 67 ]. We conclud e that f or a ve rycomplet e $ tt^ {*} $ g eo metry the UV Berr yco nn ec tio n ist he unive rs almo nod romyr eprese ntati on o fth e pu re brai d g r o up $ P_ {d }$ o f $ d$ stri ngss pec ialized to Kohno ma t rice s$B _{ij}$computed in t er ms of the$U (1) _{R}$s p ectrum o f the UV chiral ring. A very im por tantclas s of very co mplete $t t ^{*}$geomet riesar e t h e symm e t ri c o ne s. In this c ase theKo hnomatrice s satisfy
$$(B_{\ p i(i )\pi(j)})_{\p i(k )\pi ( l )} =(B _ {i j })_ {k l },\ h space{1cm} \mat hrm{for \al l \ } \pi \in S_{ d} $$
Inthis ca se th e connec tion $\ma thcal{D}$ d esce n d s t o | theory_of Kohno_connections [@rif65; @rif67]. If_a connection_$\mathcal{D}=d+A$_of the_form_\[berkohno\] satisfies $\mathcal{D}^{2}=0$,_then the following_relations hold
$$\begin{split}
\left[ _B_{ij},B_{ik}+B_{jk}\right] = \left[_B_{ij}+B_{ik},B_{jk}\right]_=0, \hspace{1cm} & \mathrm{for} \ i<j<k, \\
\left[ B_{ij},B_{kl}\right] =0, \hspace{5cm} & \mathrm{for \_distinct_\ i,j,k,l.}
\end{split}$$
The_above_equations_are called the infinitesimal pure_braid relations. A connection of_the form_\[berkohno\] where the constant $d\times d$ matrices $B_{ij}$_satisfy_the above relations_is called a rank-$d$ Kohno connection. A connection of_this type defines a representation of_the pure braid_group_$P_{d}$_of $d$ strings
$$\sigma: P_{d}\rightarrow_GL(d,\mathbb{C})$$
through the parallel transport on the_configuration space $C_{d}$
$$\sigma: \gamma \in C_{d}\rightarrow_P \exp (-\int_{\gamma} A) \in GL(d,\mathbb{C}).$$
The family_of representations $\sigma$ parametrized by the_matrices $B_{ij}$ satisfying the infinitesimal_pure braid_relations is called the universal_monodromy [@rif67]. We_conclude that_for a very_complete $tt^{*}$ geometry the UV Berry_connection is the_universal monodromy representation of the pure_braid_group $P_{d}$ of_$d$_strings_specialized to_Kohno matrices $B_{ij}$_computed_in terms_of_the $U(1)_{R}$ spectrum of the UV_chiral_ring. A very important class of very_complete $tt^{*}$ geometries are_the_symmetric ones. In this_case the Kohno matrices satisfy
$$(B_{\pi(i)\pi(j)})_{\pi(k)\pi(l)}=(B_{ij})_{kl},_\hspace{1cm} \mathrm{for \ all \ }_\pi \in_S_{d}$$
In this_case the connection $\mathcal{D}$ descends to |
$\tilde{\alpha}$ is $\tilde{\alpha}= \alpha_{+} \left| + \right\rangle \left\langle + \right| \ \ + \ \ \alpha_{-} \left| - \right\rangle \left\langle - \right| \ \ + \ \ \alpha_{z} \left| z \right\rangle \left\langle z \right|$. Representing the damping tensor in the $\left\{ \left| x\right\rangle,\: \left| y\right\rangle,\: \left| z\right\rangle \right\} $ coordinate basis yields $\alpha\equiv \tilde{\alpha}_{xx} = \tilde{\alpha}_{yy} = (\alpha_{+} + \alpha_{-})/2 $, $\tilde{\alpha}_{zz}= \alpha_z$, $\tilde{\alpha}_{yx}=\tilde{\alpha}_{xy}=0$, and $\alpha_{+}=\alpha_{-}$. The last equality results from real tensor coefficients. However, $\alpha_{zz}$ cannot be determined uniquely from the energy dissipation formula $\dot{E}\propto \dot{\mathbf{m}}^T \tilde{\alpha} \dot{\mathbf{m}}$ because $\dot{\mathbf{m}}$ is perpendicular to the z-axis. Therefore, $\alpha_{zz}$ has no physical significance and the energy dissipation is governed by the single parameter $\alpha$. For an infinite system, this damping parameter does not depend on the specific direction of the magnetization, i.e., it is isotropic because the symmetry of the Hamiltonian is not directly linked to the crystallographic axes of the underlying crystal lattice (when the cubic symmetry term in Eq. is disregarded). For a finite system, the shape of the system induces anisotropy in the magnetization damping. This effect is illustrated in Fig. \[fig:Fig2\], which plots the effective damping in Eq. as a function of the magnetization directions for different system shapes. When the cross-section of the conductor is deformed from a regular shape to the shape of a thinner system, the anisotropy of the damping changes. The magnetization damping varies from a minimum value of around 0.004 to a maximum value of 0.006, e.g., the anisotropy is around 50%. The relaxation process is largest along the axis where the ballistic leads are connected, i.e., the y-axis. This shape anisotropy is about four- to five-times stronger than the anisotropy | $ \tilde{\alpha}$ is $ \tilde{\alpha}= \alpha_{+ } \left| + \right\rangle \left\langle + \right| \ \ + \ \ \alpha_{- } \left| - \right\rangle \left\langle - \right| \ \ + \ \ \alpha_{z } \left| z \right\rangle \left\langle z \right|$. Representing the damping tensor in the $ \left\ { \left| x\right\rangle,\: \left| y\right\rangle,\: \left| z\right\rangle \right\ } $ coordinate basis yields $ \alpha\equiv \tilde{\alpha}_{xx } = \tilde{\alpha}_{yy } = (\alpha_{+ } + \alpha_{-})/2 $, $ \tilde{\alpha}_{zz}= \alpha_z$, $ \tilde{\alpha}_{yx}=\tilde{\alpha}_{xy}=0 $, and $ \alpha_{+}=\alpha_{-}$. The final equality result from real tensor coefficients. However, $ \alpha_{zz}$ cannot be determined uniquely from the department of energy dissipation formula $ \dot{E}\propto \dot{\mathbf{m}}^T \tilde{\alpha } \dot{\mathbf{m}}$ because $ \dot{\mathbf{m}}$ is perpendicular to the omega - bloc. Therefore, $ \alpha_{zz}$ has no physical significance and the energy waste is governed by the single argument $ \alpha$. For an infinite system, this damping parameter does not depend on the specific direction of the magnetization, i.e., it is isotropic because the symmetry of the Hamiltonian is not immediately linked to the crystallographic axes of the underlying quartz glass lattice (when the cubic symmetry term in Eq. is disregarded). For a finite system, the human body of the system induces anisotropy in the magnetization damping. This effect is illustrated in Fig. \[fig: Fig2\ ], which plat the effective damping in Eq. as a function of the magnetization directions for different system shapes. When the cross - section of the conductor is deformed from a regular shape to the shape of a thinner system, the anisotropy of the damping changes. The magnetization damping varies from a minimum value of around 0.004 to a maximal value of 0.006, for example, the anisotropy is around 50% . The relaxation process is largest along the bloc where the ballistic star are connected, i.e., the y - bloc. This shape anisotropy is approximately four- to five - times stronger than the anisotropy | $\tipde{\alpha}$ is $\tilde{\alpha}= \aupha_{+} \left| + \rigkr\ranglx \left\lzngle + \rkght| \ \ + \ \ \alpha_{-} \left| - \cighr\rangoe \left\langle - \right| \ \ + \ \ \wlpha_{z} \lwft| z \right\ranjme \left\langle d \rigkt|$. Representing jhe damping densor in the $\newt\{ \left| x\right\rangle,\: \left| y\right\randle,\: \legt| z\right\rangle \rigne\} $ ckordinate basis yields $\alpha\equiv \filde{\alkha}_{xx} = \tilde{\alpha}_{yu} = (\alpha_{+} + \alpha_{-})/2 $, $\tilde{\alphw}_{zz}= \wlpha_z$, $\tilde{\alpha}_{yd}=\tilde{\alpha}_{zy}=0$, age $\alpha_{+}=\alpha_{-}$. The last tqbality resumts from real tensor coefficiengs. Hocever, $\alpha_{ev}$ cajtot be detecmined uniquely from the etergy dossipation forkule $\dor{E}\propto \dot{\mathbf{m}}^T \vilde{\alpha} \dot{\mathbf{i}}$ because $\dkt{\mathbf{m}}$ is perpwneiculdr tm thd z-abis. Tiersfore, $\wlpia_{zz}$ has no physical sugnificance and the egvtgy dissipatjon is gjverned by the single parameter $\alpha$. Fmr zn infinite system, this damping parameter dogs not depqnd on the specific direction of the magnetizatiot, i.e., mt is ifurrlpic because the symmetry of the Hamiltonian ya mon directly linked to the cryxtwlkjgraphic axes of thz uhderlying crystal pattice (when the cubis sykmetry term in Eq. is disregqrded). For a yinute system, the sha'e of the syrtem induves anisotropy in the mcgnetiaation damplng. This swfect is illustrxtec hn Fig. \[fig:Fig2\], which plots tre effectmve dcmping iv Eq. as a sunction ov the magnetization direftiond xor differfnt system shapes. When the cross-section of thg cmndlctor is beformcd from a regulwr shape to thg shape oy a thknner systvm, the anmsotropy of ehe damping cvwnges. The majnetizatijn dqmpibg varidr from a minimim value of around 0.004 to a maximum valuc of 0.006, s.g., the anisotro'v us around 50%. The telxxaeiln pwmcess is larcest alumg thd axis wherc tfe bsllistic leads are cmnnedted, i.e., the y-axis. Yhls shape qnisotro[y is about fpur- to five-times surongec than the snifotropy | $\tilde{\alpha}$ is $\tilde{\alpha}= \alpha_{+} \left| + \right\rangle \right| \ + \ \alpha_{-} \left| \ + \ \ \left| z \right\rangle z \right|$. Representing the damping tensor the $\left\{ \left| x\right\rangle,\: \left| y\right\rangle,\: \left| z\right\rangle \right\} $ coordinate basis yields \tilde{\alpha}_{xx} = \tilde{\alpha}_{yy} = (\alpha_{+} + \alpha_{-})/2 $, $\tilde{\alpha}_{zz}= \alpha_z$, $\tilde{\alpha}_{yx}=\tilde{\alpha}_{xy}=0$, and $\alpha_{+}=\alpha_{-}$. last results real coefficients. However, $\alpha_{zz}$ cannot be determined uniquely from the energy dissipation formula $\dot{E}\propto \dot{\mathbf{m}}^T \tilde{\alpha} \dot{\mathbf{m}}$ $\dot{\mathbf{m}}$ is perpendicular to the z-axis. Therefore, $\alpha_{zz}$ no physical significance and energy dissipation is governed by single $\alpha$. For infinite this parameter does not on the specific direction of the magnetization, i.e., it is isotropic because the symmetry of the Hamiltonian not directly the crystallographic of underlying lattice (when the term in Eq. is disregarded). For the shape of the system induces anisotropy in magnetization damping. effect is illustrated in Fig. \[fig:Fig2\], plots the effective damping in Eq. as a of the magnetization directions for different system shapes. When the cross-section of the conductor is a regular shape to shape of a system, anisotropy the changes. The damping varies from a minimum value of around 0.004 to a value of 0.006, e.g., the anisotropy is around 50%. The is along the axis the ballistic leads are i.e., y-axis. This shape anisotropy four- five-times anisotropy | $\tilde{\alpha}$ is $\tilde{\alpha}= \alpHa_{+} \left| + \righT\rangLe \lEft\LaNgle + \RighT| \ \ + \ \ \alpha_{-} \left| - \righT\RangLe \left\langle - \right| \ \ + \ \ \alpha_{Z} \left| Z \rIGht\rANgLe \lefT\langle Z \RiGHT|$. RePrEsEntInG ThE dampIng Tensor iN the $\left\{ \leFt| x\RiGht\rangle,\: \lefT| Y\rIght\rangle,\: \LefT| z\right\ranglE \riGht\} $ cooRdInaTE basiS yiElds $\aLpha\eqUIv \tildE{\alpha}_{xx} = \tIlDE{\alpha}_{YY} = (\alpha_{+} + \aLPHa_{-})/2 $, $\TildE{\alpha}_{zz}= \alpha_z$, $\tilDE{\aLPha}_{yx}=\tilde{\alphA}_{xy}=0$, and $\AlPHa_{+}=\ALPha_{-}$. the Last equaliTy ResulTS from reAL tENSOr cOEfficients. HowEver, $\alpha_{zz}$ CAnnOt be deTeRmiNEd uniqUely fRoM The Energy dissiPatiOn formula $\Dot{E}\prOPto \dot{\mAThbf{m}}^T \tIlde{\alPha} \Dot{\MathBF{m}}$ BeCauSe $\DOt{\mAThBf{m}}$ IS peRpendicuLaR tO the z-Axis. tHEREforE, $\alPha_{zZ}$ has nO physical signIfiCancE And The enErgy dIssiPaTion iS goverNed by ThE single parameteR $\alpHa$. For an inFinItE syStEm, thiS DampinG paRamEter doeS not depENd oN tHE SPeCific direction of thE mAGNeTization, I.e., it is ISoTrOPic becauSe The SymmETRy of tHe HaMIlTonian is Not dirECtLy Linked tO tHe crysTaLloGraPhic aXEs of The undErlying cRystaL Lattice (when the CUbic symmetry tERm IN eq. IS disRegArded). For a fiNite SYsteM, the SHaPe oF The syStem iNdUCeS Anisotropy in the magnEtIzatioN dampIng. This effect Is illustraTED In Fig. \[fig:fig2\], wHIcH Plots the effectIve daMping in Eq. aS A functioN of thE magnetiZation dirECTions for DifFerEnt SysTEM sHapes. When the cROSs-seCtIon of thE coNductor Is dEfoRmeD frOm A regular sHape to thE sHaPe Of A thInner SYstem, the AnIsoTrOpy Of the DAmping ChangEs. ThE mAgNEtiZation dAMpING varIeS fRom a MinImUm valUe of ARouNd 0.004 to a maXimum valuE of 0.006, E.G., the AnIsOtropy iS around 50%. The relAxAtion proceSs Is lArgest ALOng the axIs where the ballistic leadS Are connEctEd, i.e., tHe y-aXis. This shApe AnisotRopY Is abouT four- tO five-TiMes STRongeR THaN thE aNisotropy | $\tilde{\alpha}$ is $\til de{\alpha} = \al pha _{+ }\lef t| + \right\rangl e \le ft\langle + \right| \ \ +\ \\a lpha_ {-} \le f t| - \ ri gh t\r an g le \lef t\l angle - \right| \ \ + \ \ \al p ha _{z} \left | z \right\ran gle \left \l ang l e z \ rig ht|$. Repre s enting the damp in g tenso r in the $ \l eft\ { \left| x\right \ ra n gle,\: \left| y\rig ht \ ra n g le, \: \left| z \r ight\ r angle \ r ig h t \ } $ coordinate ba sis yields$ \al pha\eq ui v \ t ilde{\ alpha }_ { xx} = \tilde{\ alph a}_{yy} = (\alp h a_{+} + \alpha_ {-})/2 $, $\ tild e {\ al pha }_ { zz} = \ alp h a_z $, $\til de {\ alpha }_{y x } = \ tild e{\ alph a}_{x y}=0$, and $\ alp ha_{ + }=\ alpha _{-}$ . Th elastequali ty re su lts from real t enso r coeffic ien ts . H ow ever, $\alph a_{ zz} $ canno t be de t erm in e d un iquely from the en er g y d issipati on for m ul a$ \dot{E}\ pr opt o \d o t {\mat hbf{ m }} ^T \tild e{\alp h a} \ dot{\ma th bf{m}} $bec aus e $\d o t{\m athbf{ m}}$ isperpe n dicular to the z-axis. There f or e , $ \ alph a_{ zz}$ has no phy s ical sig n if ica n ce an d the e n er g y dissipation is go ve rned b y the single param eter $\alp h a $ . For an inf i ni t e system, this damp ing parame t er doesnot d epend on the spec i f ic direc tio n o f t hem a gn etization, i. e . , it i s isotr opi c becau sethe sy mme tr y of theHamilton ia nis n otdirec t ly linke dtoth e c rysta l lograp hic a xesof t h e u nderlyi n gc r ysta lla ttic e ( wh en th e cu b icsymmetr y term in Eq . is d is regarde d). For a fin it e system,th e s hape o f the syst em induces anisotropy i n the ma gne tizat iondamping.Thi s effe cti s illu strate d inFi g.\ [ fig:F i g 2\ ],wh ich plotst h e e ffect iv e da mping i n Eq. as a functi o n o f the magneti zat iond i re cti o ns for d i ffe r e nt system shape s. When th ec ro ss-section ofth e condu ctor is defo r med fro m a regul ar shapeto the s hap e of a thi nner sys tem, thea nisot r op y ofthe dampi ng ch anges . Them agn etiza tion d am ping v aries f rom a mi nimum value of around 0 .004 t o a m axi mum value of 0.0 06, e.g., the anisotrop y i s a round 50 % . The rel a xa tio n proc essi s largest al ong t he axis where t h e b allis tic leadsareconnected, i.e.,t he y-axis. Thi s sh a p e a nis o trop yis about four- to f i v e-timesst ronger than the ani so t ropy | $\tilde{\alpha}$_is $\tilde{\alpha}=_\alpha_{+} \left| +_\right\rangle \left\langle_+_\right| \__ \ _+ \ _\ \alpha_{-} \left|_ - \right\rangle_\left\langle_- \right| \ \ + \ \ \alpha_{z} \left|__z \right\rangle_\left\langle_z_\right|$. Representing the damping tensor_in the $\left\{ \left| _x\right\rangle,\: \left|_ y\right\rangle,\: \left| z\right\rangle \right\} $_coordinate_basis yields $\alpha\equiv_\tilde{\alpha}_{xx} = \tilde{\alpha}_{yy} = (\alpha_{+} + \alpha_{-})/2 $, $\tilde{\alpha}_{zz}=_\alpha_z$, $\tilde{\alpha}_{yx}=\tilde{\alpha}_{xy}=0$, and $\alpha_{+}=\alpha_{-}$. The last_equality results from_real_tensor_coefficients. However, $\alpha_{zz}$ cannot_be determined uniquely from the energy_dissipation formula $\dot{E}\propto \dot{\mathbf{m}}^T \tilde{\alpha} \dot{\mathbf{m}}$_because $\dot{\mathbf{m}}$ is perpendicular to the z-axis._Therefore, $\alpha_{zz}$ has no physical significance_and the energy dissipation is_governed by_the single parameter $\alpha$. For_an infinite system,_this damping_parameter does not_depend on the specific direction of_the magnetization, i.e.,_it is isotropic because the symmetry_of_the Hamiltonian is_not_directly_linked to_the crystallographic axes_of_the underlying_crystal_lattice (when the cubic symmetry term_in_Eq. is disregarded). For a finite system,_the shape of the_system_induces anisotropy in the_magnetization damping. This effect is_illustrated in Fig. \[fig:Fig2\], which plots the_effective damping_in Eq. _as a function of the magnetization directions for different system shapes._When the cross-section of the conductor_is deformed from a_regular shape_to_the shape of_a_thinner system,_the anisotropy of the damping changes. The_magnetization damping_varies from a minimum value of_around 0.004 to a_maximum_value of 0.006, e.g., the anisotropy_is around 50%. The relaxation process_is largest along the axis_where_the_ballistic leads are connected, i.e.,_the y-axis. This shape anisotropy is_about four- to_five-times stronger than the anisotropy |
dust particles. Formation of stardust occurs in circumstellar environments where temperatures are cool enough [e.g. @Cherchneff:2017 for a recent review of the open issues]. On their journey through the interstellar medium, heating and partial or complete destruction may occur from starlight or even shocks from supernovae [@Zhukovska:2016]. Also a variety chemical and physical reactions may reprocess dust grains [@Dauphas:2016]. Thus, the journey from the stardust source up to inclusion in meteorites which found their way to Earth remains subject to theoretical modelling and much residual uncertainty [@Jones:2009]. Nevertheless, cosmic dust particles are independent astrophysical messengers, and complement studies based on electromagnetic radiation in important ways. Grain composition and morphology from the stardust laboratory measurements are combined with astronomical results such as characteristic spectral lines (e.g. from water ice, or a prominent feature associated with silicate dust), and interpreted through (uncertain) theories of cosmic dust formation and transport [@1998AREPS..26..147Z; @Cherchneff:2016]. Experimental difficulties and limitations arise from sample preparation through a variety of complex chemical methods, and by the extraction techniques evaporising material from the dust grain surfaces for subsequent mass spectrometry (see Chapter 10).
![Example of a $\gamma$-ray line measurement: The characteristic line from $^{26}$Al decay at 1808.63 keV appears Doppler-shifted from large scale galactic rotation, as it is viewed towards different galactic longitudes [left; from @Kretschmer:2013]. This measurement was performed with the SPI spectrometer on INTEGRAL, an example of a present-generation space-borne $\gamma$-ray telescope. The INTEGRAL satellite (artist view picture, ESA) has two main telescopes; the spectrometer SPI, one of them, is shown at the lower-right schematically with its 19-detector Ge camera and the tungsten mask for imaging by casting a shadow onto the camera. Space-based instruments of this kind are required to directly measure the characteristic $\gamma$-ray lines from the decay of unstable isotopes near sites of current-epoch cosmic element formation.[]{data-label="fig_1_gamma"}](Fig_Al-spectra-vs-long.pdf "fig:"){width="32.00000%"}![Example of a $\gamma$-ray line measurement: The characteristic line from $^{26}$Al decay at 1808.63 keV appears Doppler | dust particles. Formation of stardust occurs in circumstellar environment where temperature are cool enough [ e.g. @Cherchneff:2017 for a recent review of the assailable issues ]. On their journey through the interstellar medium, heating system and partial or complete destruction may happen from starlight or even shock absorber from supernovae [ @Zhukovska:2016 ]. besides a variety chemical and physical reaction may recycle dust grains [ @Dauphas:2016 ]. therefore, the journey from the stardust source up to inclusion body in meteorite which found their way to Earth remains subject to theoretical model and much residual uncertainty [ @Jones:2009 ]. Nevertheless, cosmic dust particles are autonomous astrophysical messengers, and complement study based on electromagnetic radiation in important means. Grain composition and morphology from the stardust laboratory measurements are combined with astronomic results such as characteristic spectral lines (e.g. from water ice, or a prominent feature associated with silicate dust), and interpret through (uncertain) theories of cosmic debris constitution and ecstasy [ @1998AREPS.. 26.. 147Z; @Cherchneff:2016 ]. Experimental difficulties and limitations arise from sample preparation through a variety of complex chemical methods, and by the origin techniques evaporising material from the dust grain surfaces for subsequent mass spectrometry (see Chapter 10).
! [ Example of a $ \gamma$-ray line measurement: The characteristic line from $ ^{26}$Al decay at 1808.63 keV appears Doppler - shift from bombastic scale galactic rotation, as it is viewed towards different galactic longitude [ left; from @Kretschmer:2013 ]. This measurement was performed with the SPI spectrometer on INTEGRAL, an example of a present - generation space - borne $ \gamma$-ray telescope. The INTEGRAL satellite (artist view video, ESA) has two independent telescopes; the spectrometer SPI, one of them, is read at the lower - right field schematically with its 19 - detector Ge camera and the tungsten mask for imaging by casting a tail onto the camera. Space - based instruments of this kind are required to immediately measure the characteristic $ \gamma$-ray lines from the decay of unstable isotope near sites of current - epoch cosmic element formation.[]{data - label="fig_1_gamma"}](Fig_Al - spectra - vs - long.pdf " fig:"){width="32.00000%"}![Example of a $ \gamma$-ray line measurement: The characteristic line from $ ^{26}$Al decay at 1808.63 keV look Doppler | dudt particles. Formation on stardust occurs in ciccumstemlar envkronments where temperatures aee coil enough [e.g. @Cherchnefw:2017 for a rvcent revuew id the open issues]. On thejv jouxnxy through the lnterstellas medium, heatitg aud partial or complete destruction mwy occut vrom starlight or tveg shkbkw from supernovae [@Zhukovska:2016]. Amso a veriety chemical and physical reactions maj reorocess dust graind [@Dauphas:2016]. Tyus, eye journey ffom the stardust sourcg up to inclusion in meteorites dhich found theur waj to Earth rxmains subject to bneorethcal mocelling and mugh revidyal uncertainty [@Jones:2009]. Nevertheless, cosmyc dust pdrcicles are independenr qstrokhysiwal owssdngtrs, ahd comolejent studiss based on electromagnetic raciwnoon in imporfant wwyf. Grain composition and morphology from ths stardust laboratory mwasurements are combijed with wstronomical results such as characteristic spectsal lmnds (t.g. from qaher ice, or a prominent feature associated witr solpcate dust), and inberpreted through (infetjain) theories uf cosmic dust formation anf transkort [@1998AEEPS..26..147Z; @Chewchnrff:2016]. Experimental difficultiws and limitctiins arise from sam'le preparatnon thtough s variety of complex chzmical methods, anf by the sbtraction techniduex avaporising material from ehe dust jrain surfacds fpr subfequent mads spcwtrometry (see Chapher 10).
![Axample of a $\gamma$-ray line measurement: Thx characterisjic lite from $^{26}$Cl decsy at 1808.63 keV ap[ears Doppler-skifted fxom lafge scale falactir rotation, af it is vieweg towards difherent gajactuc lingituddr [left; from @Krrtschmer:2013]. Nhns measurwment was performev wijh the SPI spectrimeuee on INTEGRAL, sn dxaippe os a present-geteragiov spacd-borne $\gammc$-rxy trlescope. The INTEGRAN safellite (artist viee iicture, EWA) has tro main telesvopes; the spectromtter S'I, one of tnem, is shown at the lower-right scgematicalpy rith its 19-detqctov Ge camera anb the tungsten mask for imaging by castiig a shadow onto the canera. Space-based insjrukents of thms kinq are reqgired to directly meqsure the characttristic $\gamma$-ray lines rrom tve defay of unstable isotopes near sites of current-epoch cosmic element formatuon.[]{date-lwbel="fig_1_gammz"}](Fig_Sl-spebtrc-vs-kong.pds "fij:"){wmdth="32.00000%"}![Example of a $\bamma$-ray line measurement: The ciaracterisdie line from $^{26}$Al decay at 1808.63 keV spoears Doppler | dust particles. Formation of stardust occurs in where are cool [e.g. @Cherchneff:2017 for open On their journey the interstellar medium, and partial or complete destruction may from starlight or even shocks from supernovae [@Zhukovska:2016]. Also a variety chemical and reactions may reprocess dust grains [@Dauphas:2016]. Thus, the journey from the stardust source to in which their way to Earth remains subject to theoretical modelling and much residual uncertainty [@Jones:2009]. Nevertheless, cosmic particles are independent astrophysical messengers, and complement studies on electromagnetic radiation in ways. Grain composition and morphology the laboratory measurements combined astronomical such as characteristic lines (e.g. from water ice, or a prominent feature associated with silicate dust), and interpreted through (uncertain) of cosmic and transport @Cherchneff:2016]. difficulties limitations arise from through a variety of complex chemical the extraction techniques evaporising material from the dust surfaces for mass spectrometry (see Chapter 10). ![Example a $\gamma$-ray line measurement: The characteristic line from decay at 1808.63 keV appears Doppler-shifted from large scale galactic rotation, as it is viewed galactic longitudes [left; from This measurement was with SPI on an example a present-generation space-borne $\gamma$-ray telescope. The INTEGRAL satellite (artist view picture, has two main telescopes; the spectrometer SPI, one of them, at lower-right schematically with 19-detector Ge camera and tungsten for imaging by casting onto camera. this are to directly measure the $\gamma$-ray lines from the decay unstable isotopes near sites "fig:"){width="32.00000%"}![Example of a $\gamma$-ray line measurement: The characteristic from $^{26}$Al decay at 1808.63 keV appears | dust particles. Formation of sTardust occUrs in CirCumStEllaR envIronments where TEmpeRatures are cool enough [e.g. @chercHnEFf:2017 foR A rEcent Review oF ThE OPen IsSuEs]. ON tHEiR jourNey Through The interstEllAr Medium, heatinG AnD partial or ComPlete destrucTioN may ocCuR frOM starLigHt or eVen shoCKs from SupernovaE [@ZHUkovskA:2016]. also a vaRIEtY cheMical and physical rEAcTIons may reproceSs dust GrAInS [@dAupHas:2016]. thus, the jouRnEy froM The starDUsT SOUrcE Up to inclusion In meteoriteS WhiCh founD tHeiR Way to EArth rEmAIns Subject to thEoreTical modeLling aND much reSIdual unCertaiNty [@jonEs:2009]. NeVErThEleSs, COsmIC dUst PArtIcles are InDePendeNt asTROPHysiCal MessEngerS, and complemenT stUdieS BasEd on eLectrOmagNeTic raDiatioN in imPoRtant ways. Grain cOmpoSition and MorPhOloGy From tHE stardUst LabOratory MeasureMEntS aRE COmBined with astronomiCaL REsUlts such As charACtErIStic specTrAl lInes (E.G. From wAter ICe, Or a promiNent feATuRe AssociaTeD with sIlIcaTe dUst), anD InteRpreteD through (UncerTAin) theories of cOSmic dust formaTIoN ANd TRansPorT [@1998AREPS..26..147Z; @CherChneFF:2016]. ExpErimENtAl dIFficuLties AnD LiMItations arise from saMpLe prepAratiOn through a varIety of compLEX Chemical MethODs, ANd by the extractIon teChniques evAPorising MaterIal from tHe dust graIN Surfaces For SubSeqUenT MAsS spectrometry (SEE ChaPtEr 10).
![ExampLe oF a $\gamma$-Ray LinE meAsuReMent: The chAracteriStIc LiNe FroM $^{26}$Al deCAy at 1808.63 keV aPpEarS DOppLer-shIFted frOm larGe scAlE gALacTic rotaTIoN, AS it iS vIeWed tOwaRdS diffErenT GalActic loNgitudes [lEft; FRom @KReTsChmer:2013]. ThIs measurement WaS performed WiTh tHe SPI sPECtrometeR on INTEGRAL, an example of a PResent-gEneRatioN spaCe-borne $\gaMma$-Ray telEscOPe. The InTEGRAl sateLlIte (ARTist vIEW pIctUrE, ESA) has two MAIn tElescOpEs; thE spectrOmeter SPI, one of them, IS shOwn at the lower-RigHt scHEMaTicALlY WitH iTS 19-deTECtor Ge camera and The tungsteN mASk For imaging BY caStIng a shaDow onto The caMEra. SpacE-based insTruments oF tHis kIND arE required tO directlY measure tHE charACtEristIc $\gAmma$-raY lIneS from The decAY of UnstaBle isoToPes neaR siteS oF current-Epoch cosmic element formaTion.[]{daTa-labEl="fIg_1_gamma"}](FiG_Al-SPecTra-vs-long.Pdf "fIg:"){width="32.00000%"}![ExaMplE of A $\gammA$-raY Line mEasuREmEnt: tHe chaRactERistic linE FrOm $^{26}$AL DEcAy at 1808.63 keV appeARS dopPler | dust particles. Formation of stardu st oc cur s i ncirc umst ellar environm e ntswhere temperatures are cool e n ough [e .g. @ Cherchn e ff : 2 017 f or are c en t rev iew of the open issu es] .On their jou r ne y throughthe interstella r m edium, h eat i ng an d p artia l or c o mplete destruct io n may o c cur fro m st arli ght or even shock s f r om supernovae[@Zhuk ov s ka : 2 016 ].Also a var ie ty ch e mical a n dp h y sic a l reactions m ay reproces s du st gra in s [ @ Daupha s:201 6] . Th us, the jou rney from the stard u st sour c e up to inclu sio n i n me t eo ri tes w h ich fo und the ir way t oEa rth r emai n s s ubje ctto t heore tical modelli ngandm uch resi dualunce rt ainty [@Jon es:20 09 ]. Nevertheless , co smic dust pa rt icl es arei ndepen den t a strophy sical m e sse ng e r s ,and complement stu di e s b ased onelectr o ma gn e tic radi at ion ini m porta nt w a ys . Graincompos i ti on and mo rp hology f rom th e sta r dust labor atory me asure m ents are combi n ed with astro n om i c al resu lts such as ch arac t eris tics pe ctr a l lin es (e .g . f r om water ice, or apr ominen t fea ture associat ed with si l i c ate dust ), a n di nterpreted thr ough(uncertain ) theorie s ofcosmic d ust forma t i on and t ran spo rt[@1 9 9 8A REPS..26..147 Z ; @Ch er chneff: 201 6]. Exp eri men tal di ff icultiesand limi ta ti on sari se fr o m sample p rep ar ati on th r ough a vari etyof c o mpl ex chem i ca l meth od s, and by t he ex trac t ion techni ques evap ori s ingma te rial fr om the dust g ra in surface sfor subse q u ent mass spectrometry (see Chap t er 10).
![E xamp le of a $ \ga mma$-r ayl ine me asurem ent:Th e c h a racte r i st icli ne from $^ { 2 6}$ Al de ca y at 1808.6 3 keV appears Dopp l er- shifted fromlar ge s c a le ga l ac t icro t ati o n , as it is view ed towards d i ff erent gala c tic l ongitud es [lef t; fr o m @Kret schmer:20 13]. This m easu r e men t was perf ormed wi th the SP I spec t ro meter on INTEG RA L,an ex ampleo f a pres ent-ge ne ration spac e- borne $\ gamma$-ray telescope. T he INT EGRAL sa tellite ( art i stview pict ure, ESA) hastwo ma in te les c opes; the sp ect r omete r SP I , one oft he m,i s s hown at the l o wer -righ t s c hemati call y with its 19-det e ctor Ge camera and t hetun g sten m ask for imagin g b yc a sting ash adow onto t he camer a. Space -based instr umentso f t h is kin d ar e r equired t o d ir e ctly me as ur e the c hara ct eristi c $\ga m ma$- r a y lines from the deca y of un s tab le is ot opes ne a r si tes of cur rent-epochcosmic ele mentformati on .[]{da ta- la bel="fig_1 _ gamma"}]( Fig_A l-spect ra -vs- lon g.pdf"fig : " ){wid th=" 32 .00 000%"}![E x a mp l eof a $ \gam ma$-r ay lin e measure m ent: The ch a racteri st icl i ne fro m $ ^ { 26}$Al dec ayat 18 0 8 .63 keV ap p ears Do p pler | dust_particles. Formation_of stardust occurs in_circumstellar environments_where_temperatures are_cool_enough [e.g. @Cherchneff:2017_for a recent_review of the open_issues]. On their_journey_through the interstellar medium, heating and partial or complete destruction may occur from starlight_or_even shocks_from_supernovae_[@Zhukovska:2016]. Also a variety chemical_and physical reactions may reprocess_dust grains_[@Dauphas:2016]. Thus, the journey from the stardust source_up_to inclusion in_meteorites which found their way to Earth remains subject_to theoretical modelling and much residual_uncertainty [@Jones:2009]. Nevertheless,_cosmic_dust_particles are independent astrophysical_messengers, and complement studies based on_electromagnetic radiation in important ways. Grain_composition and morphology from the stardust laboratory_measurements are combined with astronomical results_such as characteristic spectral lines_(e.g. from_water ice, or a prominent_feature associated with_silicate dust),_and interpreted through_(uncertain) theories of cosmic dust formation_and transport [@1998AREPS..26..147Z;_@Cherchneff:2016]. Experimental difficulties and limitations arise_from_sample preparation through_a_variety_of complex_chemical methods, and_by_the extraction_techniques_evaporising material from the dust grain_surfaces_for subsequent mass spectrometry (see Chapter 10).
_ ![Example of_a_$\gamma$-ray line measurement: The_characteristic line from $^{26}$Al decay_at 1808.63 keV appears Doppler-shifted from_large scale_galactic rotation,_as it is viewed towards different galactic longitudes [left; from @Kretschmer:2013]._This measurement was performed with the_SPI spectrometer on INTEGRAL,_an example_of_a present-generation space-borne_$\gamma$-ray_telescope. The_INTEGRAL satellite (artist view picture, ESA) has_two main_telescopes; the spectrometer SPI, one of_them, is shown at_the_lower-right schematically with its 19-detector Ge_camera and the tungsten mask for_imaging by casting a shadow_onto_the_camera. Space-based instruments of this_kind are required to directly measure_the characteristic $\gamma$-ray_lines from the decay of unstable isotopes_near_sites of current-epoch cosmic element formation.[]{data-label="fig_1_gamma"}](Fig_Al-spectra-vs-long.pdf_"fig:"){width="32.00000%"}![Example_of a $\gamma$-ray line measurement: The_characteristic_line_from $^{26}$Al decay at 1808.63_keV appears Doppler |
paulus:2017:arxiv], there is still no guarantee that the generated summaries are grammatical and convey the same meaning as the original document does. It seems that extractive models are more reliable than their abstractive counterparts.
However, extractive models require sentence level labels, which are usually not included in most summarization datasets (most datasets only contain document-summary pairs). Sentence labels are usually obtained by rule-based methods (e.g., maximizing the ROUGE score between a set of sentences and reference summaries) and may not be accurate. Extractive models proposed recently [@cheng:2016:acl; @nallapati:2017:aaai] employ hierarchical document encoders and even have neural decoders, which are complex. Training such complex neural models with *inaccurate* *binary* labels is challenging. We observed in our initial experiments on one of our dataset that our extractive model (see Section \[sec:sum\] for details) overfits to the training set quickly after the second epoch, which indicates the training set may not be fully utilized. Inspired by the recent pre-training work in natural language processing [@peters:2018:naacl; @radford:2018:nips; @devlin:2018:arxiv], our solution to this problem is to first pre-train the “complex”’ part (i.e., the hierarchical encoder) of the extractive model on unlabeled data and then we learn to classify sentences with our model initialized from the pre-trained encoder. In this paper, we propose, which stands for [**HI**]{}erachical [**B**]{}idirectional [**E**]{}ncoder [**R**]{}epresentations from [**T**]{}ransformers. We design an unsupervised method to pre-train for document modeling. We apply the pre-trained to the task of document summarization and achieve state-of-the-art performance on both the CNN/Dailymail and New York Times dataset.
Related Work
============
In this section, we introduce work on extractive summarization, abstractive summarization and pre-trained natural language processing models. For a more comprehensive review of summarization, we refer the interested readers to and.
#### Extractive Summarization
Extractive summarization aims to select important sentences (sometimes other textual units such as elementary discourse units (EDUs)) from a document as its summary. It is usually modeled as a sentence ranking problem by using | paulus:2017: arxiv ], there is still no guarantee that the generated summaries are grammatical and bring the like meaning as the original document does. It seem that extractive model are more reliable than their abstractive counterparts.
However, extractive exemplar require sentence degree labels, which are usually not admit in most summarization datasets (most datasets only contain text file - summary pairs). Sentence labels are usually obtained by rule - based method (e.g., maximizing the ROUGE score between a hardening of sentences and reference summaries) and may not be accurate. Extractive model proposed recently [ @cheng:2016: acl; @nallapati:2017: aaai ] employ hierarchical document encoders and even have neural decoder, which are complex. Training such complex neural models with * inaccurate * * binary * labels is challenging. We observed in our initial experiments on one of our dataset that our extractive model (see Section \[sec: sum\ ] for details) overfits to the training set quickly after the second epoch, which indicates the training set may not be amply utilized. Inspired by the late pre - training workplace in lifelike language processing [ @peters:2018: naacl; @radford:2018: nips; @devlin:2018: arxiv ], our solution to this problem is to first pre - train the “ complex ” ’ part (i.e., the hierarchical encoder) of the extractive model on unlabeled data and then we determine to classify sentences with our model initialized from the pre - trained encoder. In this paper, we aim, which stands for [ * * HI**]{}erachical [ * * B**]{}idirectional [ * * E**]{}ncoder [ * * R**]{}epresentations from [ * * T**]{}ransformers. We design an unsupervised method to pre - train for text file modeling. We give the pre - coach to the task of document summarization and achieve state - of - the - artwork performance on both the CNN / Dailymail and New York Times dataset.
Related Work
= = = = = = = = = = = =
In this section, we introduce work on extractive summarization, abstractive summarization and pre - trained natural language processing models. For a more comprehensive review of summarization, we consult the interested readers to and.
# # # # Extractive Summarization
Extractive summarization aim to select significant sentences (sometimes early textual units such as elementary discourse unit (EDUs) ) from a document as its summary. It is usually model as a sentence rank problem by use | paupus:2017:arxiv], there is still ko guarantee thaj rhe geierated summarids are grammatical and conveb thw samt meaning as the orkginal dobument dows. Iu seems that extredtive models zve moxe reliable than their absdractive countaroaxts.
However, extractive models require sentenve level labels, rhicn are lsmally not included in most summadizatioi datasets (most datasets only contain dockmenh-summary pairs). Senhence labelw arq usually obtxined by rlne-based mejhods (e.g., maximizing the ROUGE scure bztween a sej kf ventences aid refvrence summarlvs) and kay not be accurate. Ewtracvive models proposed receitly [@cheng:2016:acl; @nallapwti:2017:aaai] ekpmoy hierarchical eoxumenj encmderr ana ebei hzve nekram decoders, which are xomplex. Training suvh bpmplex neuram modejs with *inaccurate* *binary* labels is challtnginf. We observed in our inutial experiments on lne of ouw dataset that our extractive model (see Section \[sac:sum\] wor dctaiuw) lverfits to the training set quickly after thq stcokd epoch, which ikdicates the trainonh xgt may not be wully bfimized. Inspired by hhe recgnt prw-training worl in natural language procewsing [@peters:2018:uaaxl; @radford:2018:nips; @devpin:2018:arxiv], oux solujion tp this problem is to fixst prs-train the “fomplex”’ pzft (i.e., the hierarzhibal ancoder) of the extractive iodel on nnlabzled datx anc then we learn ho classify sentences wihh out moden initialixed from the pre-trained encoder. In this paper, wa pgopose, whnch stsnds for [**HI**]{}erwchical [**B**]{}idireetional [**Z**]{}ncodef [**R**]{}epresennations fcom [**T**]{}ransforiers. We desigt an unsupervmsed methjd ti prw-train wur document moceling. We apply the pre-trained to the tari of document sbonarization and schkevq dtete-of-eve-art perforkancd ov both the CNN/Daiohmaik and New York Times datzset.
Related Work
============
In tmis sectiin, we ineroduce work pn extractive summwrizavion, austracyivg summarization and pre-trained hatural lwngmage processigg midels. For a kore comprehensive review of summarizatmon, we refer the intergsted readers to and.
#### Gxtvactive Summerizatyon
Extracdive summarization aums to select impprtant sentences (somefimes mther textual units such as elementary discourse units (EDUs)) from a document as its snmiary. It is hsuakly mmdzleb as a fentxnre ranking problek by using | paulus:2017:arxiv], there is still no guarantee that summaries grammatical and the same meaning It that extractive models more reliable than abstractive counterparts. However, extractive models require level labels, which are usually not included in most summarization datasets (most datasets contain document-summary pairs). Sentence labels are usually obtained by rule-based methods (e.g., maximizing ROUGE between set sentences and reference summaries) and may not be accurate. Extractive models proposed recently [@cheng:2016:acl; @nallapati:2017:aaai] employ document encoders and even have neural decoders, which complex. Training such complex models with *inaccurate* *binary* labels challenging. observed in initial on of our dataset our extractive model (see Section \[sec:sum\] for details) overfits to the training set quickly after the second which indicates set may be utilized. by the recent in natural language processing [@peters:2018:naacl; @radford:2018:nips; to this problem is to first pre-train the part (i.e., hierarchical encoder) of the extractive model unlabeled data and then we learn to classify with our model initialized from the pre-trained encoder. In this paper, we propose, which stands [**B**]{}idirectional [**E**]{}ncoder [**R**]{}epresentations from We design an method pre-train document We apply pre-trained to the task of document summarization and achieve state-of-the-art performance both the CNN/Dailymail and New York Times dataset. Related Work this we introduce work extractive summarization, abstractive summarization pre-trained language processing models. For comprehensive of the readers and. #### Extractive Summarization summarization aims to select important (sometimes other textual units (EDUs)) from a document as its summary. It usually modeled as a sentence ranking problem using | paulus:2017:arxiv], there is still no Guarantee tHat thE geNerAtEd suMmarIes are grammatiCAl anD convey the same meaning aS the oRiGInal DOcUment Does. It sEEmS THat ExTrActIvE MoDels aRe mOre reliAble than thEir AbStractive couNTeRparts.
HoweVer, Extractive moDelS requiRe SenTEnce lEveL labeLs, whicH Are usuAlly not inClUDed in mOSt summaRIZaTion Datasets (most datasETs ONly contain docuMent-suMmARy PAIrs). senTence labelS aRe usuALly obtaINeD BY RulE-Based methods (e.G., maximizing THe RoUGE scOrE beTWeen a sEt of sEnTEncEs and refereNce sUmmaries) aNd may nOT be accuRAte. ExtrActive ModEls PropOSeD rEceNtLY [@chENg:2016:Acl; @NAllApati:2017:aaaI] eMpLoy hiErarCHICAl doCumEnt eNcodeRs and even have NeuRal dECodErs, whIch arE comPlEx. TraIning sUch coMpLex neural models With *InaccuratE* *biNaRy* lAbEls is CHallenGinG. We ObserveD in our iNItiAl EXPErIments on one of our daTaSET tHat our exTractiVE mOdEL (see SectIoN \[seC:sum\] FOR detaIls) oVErFits to thE trainINg SeT quicklY aFter thE sEcoNd ePoch, wHIch iNdicatEs the traIning SEt may not be fullY Utilized. InspiREd BY ThE ReceNt pRe-training wOrk iN NatuRal lANgUagE ProceSsing [@PeTErS:2018:Naacl; @radford:2018:nips; @devLiN:2018:arxiv], Our soLution to this pRoblem is to FIRSt pre-traIn thE “CoMPlex”’ part (i.e., the hIerarChical encoDEr) of the eXtracTive modeL on unlabeLED data and TheN we LeaRn tO CLaSsify sentenceS WIth oUr Model inItiAlized fRom The Pre-TraInEd encoder. in this paPeR, wE pRoPosE, whicH Stands foR [**Hi**]{}erAcHicAl [**B**]{}idIRectioNal [**E**]{}nCodeR [**R**]{}EpREseNtationS FrOM [**t**]{}ranSfOrMers. we dEsIgn an UnsuPErvIsed metHod to pre-tRaiN For dOcUmEnt modeLing. We apply thE pRe-trained tO tHe tAsk of dOCUment sumMarization and achieve staTE-of-the-aRt pErforMancE on both thE CNn/DailyMaiL And New york TiMes daTaSet.
rELated wORk
============
in tHiS section, we INTroDuce wOrK on eXtractiVe summarization, absTRacTive summarizaTioN and PRE-tRaiNEd NAtuRaL LanGUAge processing moDels. For a moRe COmPrehensive REviEw Of summaRizatioN, we reFEr the inTerested rEaders to aNd.
#### extrACTivE SummarizaTion
ExtrActive sumMArizaTIoN aims To sElect iMpOrtAnt seNtenceS (SomEtimeS other TeXtual uNits sUcH as elemeNtary discourse units (EDUs)) From a dOcumeNt aS its summaRy. IT Is uSually modEled As a sentencE raNkiNg proBleM By usiNg | paulus:2017:arxiv], thereis still n o gua ran tee t hatthegenerated summ a ries are grammatical and c onvey t h e sa m emeani ng as t h eo r igi na ldoc um e nt does . I t seems that extr act iv e models are mo re reliabl e t han their ab str active c oun t erpar ts.
How ever,e xtract ive model sr equire sentenc e le vellabels, which are us u ally not inclu ded in m o st s umm ari zation dat as ets ( m ost dat a se t s onl y contain docu ment-summar y pa irs).Se nte n ce lab els a re usu ally obtain ed b y rule-ba sed me t hods (e . g., max imizin g t heROUG E s co rebe t wee n a se t of sentenc es a nd re fere n c e summ ari es)and m ay not be acc ura te.E xtr activ e mod elspr opose d rece ntly[@ cheng:2016:acl; @na llapati:2 017 :a aai ]emplo y hiera rch ica l docum ent enc o der sa n d e ven have neural de co d e rs , whichare co m pl ex . Trainin gsuc h co m p lex n eura l m odels wi th *in a cc ur ate* *b in ary* l ab els is chal l engi ng. We observe d ino ur initial exp e riments on on e o f ou r dat ase t that ourextr a ctiv e mo d el (s e e Sec tion\[ s ec : sum\] for details)ov erfits to t he training s et quickly a f ter theseco n de poch, which in dicat es the tra i ning set maynot be f ully util i z ed. Insp ire d b y t her e ce nt pre-traini n g wor kin natu ral langua gepro ces sin g[@peters: 2018:naa cl ;@r ad for d:201 8 :nips; @ de vli n: 201 8:arx i v], ou r sol utio nto thi s probl e mi s tofi rs t pr e-t ra in th e “c o mpl ex”’ pa rt (i.e., th e hie ra rc hical e ncoder) of th eextractive m ode l on u n l abeled d ata and then we learn t o classi fysente nces with our mo del in iti a lizedfrom t he pr e- tra i n ed en c o de r.In this pape r , we prop os e, w hich st ands for [**HI**]{ } era chical [**B** ]{} idir e c ti ona l [ * *E* *] { }nc o d er [**R**]{}epr esentation sf ro m [**T**]{ } ran sf ormers. We des ign a n unsupe rvised me thod to p re -tra i n fo r document modelin g. We app l y the pr e-tra ine d to t he ta sk of docum e ntsumma rizati on and a chiev estate-of -the-art performance on boththe C NN/ Dailymail an d Ne w York Ti mesdataset.
Rel ate d Wor k
= = ===== ==== =
Int his s ecti o n, we int r od uce w or k on extrac t i v e s ummar iza t ion, a bstr active summarizat i on and pre-tra ined n atu ral lang ua ge processingmod el s . For a m or e comprehen sive rev ie w of s ummari zation , we re f e rt he int eres ted readerstoan d .
#### E xt r active Sum ma rizati on
Ex t ract i v e summarizationaimst o sele c t i mport an t sente n ces(sometimes other text ual un itssuchas elem en tary d isc ou rse units( EDUs)) fr om adocumen tas i tssummar y. I t is us uall ymod eled as a s en t en ce ran king prob le m by using | paulus:2017:arxiv], there_is still_no guarantee that the_generated summaries_are_grammatical and_convey_the same meaning_as the original_document does. It seems_that extractive models_are_more reliable than their abstractive counterparts.
However, extractive models require sentence level labels, which are_usually_not included_in_most_summarization datasets (most datasets only_contain document-summary pairs). Sentence labels_are usually_obtained by rule-based methods (e.g., maximizing the ROUGE_score_between a set_of sentences and reference summaries) and may not be_accurate. Extractive models proposed recently [@cheng:2016:acl;_@nallapati:2017:aaai] employ hierarchical_document_encoders_and even have neural_decoders, which are complex. Training such_complex neural models with *inaccurate* *binary*_labels is challenging. We observed in our_initial experiments on one of our_dataset that our extractive model_(see Section_\[sec:sum\] for details) overfits to_the training set_quickly after_the second epoch,_which indicates the training set may_not be fully_utilized. Inspired by the recent pre-training_work_in natural language_processing_[@peters:2018:naacl;_@radford:2018:nips; @devlin:2018:arxiv],_our solution to_this_problem is_to_first pre-train the “complex”’ part (i.e.,_the_hierarchical encoder) of the extractive model on_unlabeled data and then_we_learn to classify sentences_with our model initialized from_the pre-trained encoder. In this paper,_we propose,_which stands_for [**HI**]{}erachical [**B**]{}idirectional [**E**]{}ncoder [**R**]{}epresentations from [**T**]{}ransformers. We design an unsupervised_method to pre-train for document modeling._We apply the pre-trained_to the_task_of document summarization_and_achieve state-of-the-art_performance on both the CNN/Dailymail and New_York Times_dataset.
Related Work
============
In this section, we introduce_work on extractive summarization,_abstractive_summarization and pre-trained natural language processing_models. For a more comprehensive review_of summarization, we refer the_interested_readers_to and.
#### Extractive Summarization
Extractive summarization_aims to select important sentences (sometimes_other textual units_such as elementary discourse units (EDUs)) from_a_document as its summary. It is_usually_modeled as a sentence ranking problem_by_using |
, the outbursts in such systems. Fortunately, this is not bad for our purposes since we are trying to derive an upper limit for the real probability of detecting the CVs. In this way, having a disrupted disc (cold and stable) would lower the probability even further.
Once we know that the CVs are DNe, we are able to compute the recurrence time and the duration of the outburst, based on equations \[TREC\] and \[TDUR\]. The results are displayed in Fig. \[Fig05\]. Notice that the recurrence time varies from days to decades and the outbursts last a few days. Notice also a double peak in the distribution of the recurrence time due to the double peak in the WD mass. Dynamically formed CVs have greater WD masses, implying longer recurrence times according to Equation \[TREC\]. Now, for the distribution of the duration of the outburst, we do not see too many differences between dynamically formed CVs and CVs formed without influence of dynamics. This is due to the status of the CVs at the present-day (period bouncers). They mainly have periods from $\sim$ 2 to $\sim$ 3h and according to Equation \[TDUR\] (which is a function of the orbital period) they are similar.
From Fig. \[Fig05\], we can see that the chances to observe a CV during outburst are high only for some CVs – the youngest ones that were formed close to the present-day and have greater donor masses and higher mass transfer rates with respect to the old ones. In this way, we have the first clue that a good solution would, instead of looking for variability during an outburst, look for it during quiescence. The biggest problem associated with this approach is the need of very deep observations because in order to reach magnitudes similar to those associated with a CV during quiescence, one needs deeper observations in comparison with the same CV during its outbursts. Although a difficult task, it is possible. For instance, @Cohn_2010 reached apparent visual magnitudes as deep as $\sim$ 26 when studying the optical counterparts of X-ray sources in the GC NGC 6397.
Probability of Detection through Variability {#probability}
--------------------------------------------
Now, once we have separated the CVs with respect to the disc instability and computed their recurrence times as well as the duration of their outbursts, we can turn to the computation of the probability of detecting such CVs by using equation | , the outbursts in such systems. Fortunately, this is not bad for our purposes since we are judge to deduce an upper limit for the real probability of detect the CVs. In this room, having a disrupted phonograph record (cold and stable) would lower the probability even farther.
Once we know that the CVs are DNe, we are able to calculate the recurrence time and the duration of the outburst, based on equations \[TREC\ ] and \[TDUR\ ]. The result are displayed in Fig. \[Fig05\ ]. Notice that the recurrence time varies from days to decade and the outbursts last a few days. poster also a double peak in the distribution of the recurrence time due to the bivalent peak in the WD multitude. Dynamically formed CVs have great WD masses, implying longer recurrence times according to Equation \[TREC\ ]. Now, for the distribution of the duration of the effusion, we do not attend too many differences between dynamically formed CVs and CVs formed without influence of dynamics. This is due to the status of the curriculum vitae at the present - sidereal day (period bouncers). They mainly get periods from $ \sim$ 2 to $ \sim$ 3h and according to Equation \[TDUR\ ] (which is a function of the orbital time period) they are similar.
From Fig. \[Fig05\ ], we can see that the chances to observe a curriculum vitae during outburst are high only for some CVs – the youngest ones that were formed close to the present - day and have greater donor masses and higher bulk transfer rate with respect to the old ones. In this way, we have the inaugural clue that a good solution would, instead of look for unevenness during an outburst, look for it during quiescence. The biggest problem consort with this approach is the need of very bass observations because in ordering to reach magnitude similar to those consociate with a CV during quiescence, one need deep observations in comparison with the same CV during its outbursts. Although a difficult task, it is possible. For instance, @Cohn_2010 reach apparent visual magnitudes equally deep as $ \sim$ 26 when studying the optical counterparts of X - ray sources in the GC NGC 6397.
Probability of Detection through Variability { # probability }
--------------------------------------------
immediately, once we have separated the CVs with respect to the disc instability and computed their recurrence multiplication equally well as the duration of their outbursts, we can turn to the computation of the probability of detecting such CVs by using equation | , thf outbursts in such systtms. Fortunately, tkus is iot bad for our purposes since we are tryinj to derice an upper limit for ghe real irobabiliry oh detecting the RBs. In tmns wag, havnnj a disrupted dlsc (cold ang stable) would luwzr the probability even further.
Once re know tjat the CVs arg DNe, re adv cble to compute the recurrence fime anv the duration pf the outburst, based on eeuatlons \[TREC\] and \[TDUR\]. The resultw arq displayed iv Fig. \[Fig05\]. Notice that jhe recurrence time varies from aays co decades qne tjg outbursts oast w few days. Notice alvo a doible peak in tme divtrubution of the recurrxnce time due to the double paai in the WD mass. Eybamicdlly forowd ZVs heve greatfr XD masses, ijplying lonter recurrence timex wbvording to Esuatiog \[EREC\]. Now, for the distribution of the dugatikn of the outburst, we di not see too many divferences between dynamically formed CVs and CVs formed widhout knfoucnce if dynamics. This is due to the status of the CVf au tme present-day (peviod bouncers). They mwimjy have perioas from $\sjm$ 2 to $\sim$ 3h and afcordind to Wquation \[UDUR\] (ehich is a function of the irbital perije) they are similar.
Yrom Fig. \[Fig05\], we van sre that the chances to ubsedve a CV duging outbhfst are high onlh fpr some CVs – the youngest onqs that wxre fprmed cuose to thq present-dwy and have greater donog masdev and highfr mass transfer rates with res'xct to the olc mnev. In thif way, we have the sirst clue thaj a good folutkon would, pnstead oh looking fow variability furing an ouvburst, lojk fir ir durine quiescence. Thr biggest problem awsociated with thix akpdoach is the nezb if very deep obxerxatyojs bqwause in ordar tu rdsch mxgnitudtf wkmilsr to those associatad wjth a CV during quoefcence, obe needs deeper obserfations in comparidon wmth thx same CV during its outbursts. Although a difficklt task, it is [osslble. For instauce, @Cohn_2010 reached apparent visual magnituves as deep as $\sim$ 26 whgn studying the opticcl counterparvs of V-ray sourwes in the GC NGC 6397.
Pribability of Detegtion through Variabiljty {#prmbabipity}
--------------------------------------------
Now, once we have separated the CVs with respect to the disc instabiliry and computed thsir tecurgenee cimes af wepu as the duratiom of their outbursts, we can turi to the cmm'utation of the probability og aetecting sucf CVs by using equation | , the outbursts in such systems. Fortunately, not for our since we are limit the real probability detecting the CVs. this way, having a disrupted disc and stable) would lower the probability even further. Once we know that the are DNe, we are able to compute the recurrence time and the duration the based equations and \[TDUR\]. The results are displayed in Fig. \[Fig05\]. Notice that the recurrence time varies from to decades and the outbursts last a few Notice also a double in the distribution of the time to the peak the mass. Dynamically formed have greater WD masses, implying longer recurrence times according to Equation \[TREC\]. Now, for the distribution of duration of we do see many between dynamically formed CVs formed without influence of dynamics. to the status of the CVs at the (period bouncers). mainly have periods from $\sim$ 2 $\sim$ 3h and according to Equation \[TDUR\] (which a function of the orbital period) they are similar. From Fig. \[Fig05\], we can see chances to observe a during outburst are only some – youngest ones were formed close to the present-day and have greater donor masses higher mass transfer rates with respect to the old ones. way, have the first that a good solution instead looking for variability during look it biggest associated this approach is the of very deep observations because order to reach magnitudes a CV during quiescence, one needs deeper observations comparison with the same CV during its Although a difficult task, it is possible. For instance, @Cohn_2010 reached apparent magnitudes as $\sim$ 26 when studying the optical counterparts of sources in the GC 6397. Probability of Detection through Variability {#probability} -------------------------------------------- Now, we separated the with respect to disc instability and their recurrence times as the of outbursts, to the computation of the probability detecting CVs by using equation | , the outbursts in such systems. fortunatelY, this Is nOt bAd For oUr puRposes since we aRE tryIng to derive an upper limiT for tHe REal pRObAbiliTy of detECtING thE Cvs. in tHiS WaY, haviNg a DisruptEd disc (cold And StAble) would lowER tHe probabilIty Even further.
ONce We know ThAt tHE CVs aRe Dne, we aRe able TO compuTe the recuRrENce timE And the dURAtIon oF the outburst, based ON eQUations \[TREC\] and \[tDUR\]. ThE rESuLTS arE diSplayed in FIg. \[fig05\]. NoTIce that THe RECUrrENce time varies From days to dECadEs and tHe OutBUrsts lAst a fEw DAys. notice also a DoubLe peak in tHe distRIbution OF the recUrrencE tiMe dUe to THe DoUblE pEAk iN ThE WD MAss. dynamicaLlY fOrmed cVs hAVE GReatEr Wd masSes, imPlying longer rEcuRrenCE tiMes acCordiNg to eqUatioN \[TREC\]. NOw, for ThE distribution of The dUration of The OuTbuRsT, we do NOt see tOo mAny DiffereNces betWEen DyNAMIcAlly formed CVs and CVS fORMeD without InflueNCe Of DYnamics. THiS is Due tO THe staTus oF ThE CVs at thE preseNT-dAy (Period bOuNcers). THeY maInlY have PErioDs from $\Sim$ 2 to $\sim$ 3H and aCCording to EquatIOn \[TDUR\] (which is A FuNCTiON of tHe oRbital perioD) theY Are sImilAR.
FRom fIg. \[Fig05\], We can SeE ThAT the chances to observE a cV duriNg outBurst are high oNly for some cvS – The youngEst oNEs THat were formed cLose tO the presenT-Day and haVe greAter donoR masses anD HIgher masS trAnsFer RatES WiTh respect to thE OLd onEs. in this wAy, wE have thE fiRst CluE thAt A good soluTion woulD, iNsTeAd Of lOokinG For variaBiLitY dUriNg an oUTburst, Look fOr it DuRiNG quIescencE. thE BIggeSt PrObleM asSoCiateD witH ThiS approaCh is the neEd oF Very DeEp ObservaTions because iN oRder to reacH mAgnItudes SIMilar to tHose associated with a CV duRIng quieSceNce, onE neeDs deeper oBseRvatioNs iN ComparIson wiTh the SaMe Cv DUring ITS oUtbUrSts. AlthougH A DifFiculT tAsk, iT is possIble. For instance, @CohN_2010 ReaChed apparent vIsuAl maGNItUdeS As DEep As $\SIm$ 26 wHEN studying the optIcal counteRpARtS of X-ray souRCes In The GC NGc 6397.
ProbabIlity OF DetectIon througH VariabilItY {#proBABilIty}
--------------------------------------------
Now, once We have seParated thE cVs wiTH rEspecT to The disC iNstAbiliTy and cOMpuTed thEir recUrRence tImes aS wEll as the Duration of their outburstS, we can Turn tO thE computatIon OF thE probabilIty oF detecting SucH CVS by usIng EQuatiOn | , the outbursts in such sy stems. For tunat ely , t hi s is not bad for our p u rpos es since we are trying to d er i ve a n u pperlimit f o rt h e r ea lpro ba b il ity o f d etectin g the CVs. In t his way, hav i ng a disrupt eddisc (cold a ndstable )wou l d low erthe p robabi l ity ev en furthe r.
Oncew e knowt h at the CVs are DNe, wea re able to comput e there c ur r e nce ti me and the d urati o n of th e o u t b urs t , based on eq uations \[T R EC\ ] and\[ TDU R \]. Th e res ul t s a re displaye d in Fig. \[F ig05\] . Notice that th e recu rre nce tim e v ar ies f r omd ay s t o de cades an dth e out burs t s l asta f ew d ays.Notice also a do uble pea k inthe d istr ib ution of th e rec ur rence time dueto t he double pe ak in t he WD mass.Dyn ami cally f ormed C V s h av e g re ater WD masses, im pl y i ng longerrecurr e nc et imes acc or din g to E quati on \ [ TR EC\]. No w, for th edistrib ut ion of t hedur ation of t he out burst, w e don ot see too man y differencesb et w e en dyna mic ally formed CVs andCVsf or med witho ut in fl u en c e of dynamics. This i s dueto th e status of t he CVs att h e present -day (p e riod bouncers) . The y mainly h a ve perio ds fr om $\sim $ 2 to $\ s i m$ 3h an d a cco rdi ngt o E quation \[TDU R \ ] (w hi ch is a fu nctionofthe or bit al period)they are s im il ar .
FromF ig. \[Fi g0 5\] ,wecan s e e that thechan ce st o o bservea C V duri ng o utbu rst a re hi gh o n lyfor som e CVs – t hey oung es tones th at were forme dclose to t he pr esent- d a y and ha ve greater donor masses and hig her mass tra nsfer rat eswith r esp e ct tothe ol d one s. In t his w a y ,weha ve the fir s t cl ue th at a g ood sol ution would, inste a d o f looking for va riab i l it y d u ri n g a no utb u r st, look for it during qu ie s ce nce. The b i gge st proble m assoc iated with th is approa ch is the n eedo f ve ry deep ob servatio ns becaus e in o r de r torea ch mag ni tud es si milart o t hoseassoci at ed wit h a C Vduring q uiescence, one needs de eper o bserv ati ons in co mpa r iso n with th e sa me CV duri ngits outb urs t s. Al thou g ha d i fficu lt t a sk, it is po ssi b l e. For instan c e , @C ohn_2 010 reache d ap parent visual mag n itudes as deep as$ \ sim $ 2 6 whe nstudying the o pti ca l counterp ar ts of X-ray sources i n theGC NGC 6397.
Proba b i li t y of D etec tio n through Va ri a bility{# pr o babili ty}-- ------ ------ - ---- - - ---------------- ----- - -
Now , on ce we h ave sep a rate d the CVswith respec t to t he d isc i nstabil it y andcom pu ted theirr ecurrence time s as we ll asthe durat iono f thei r ou tb urs ts, we ca n tu r nto the com putat io n of the prob a bility o f d e tecting s uch C Vs byu si n g equation | , the_outbursts in_such systems. Fortunately, this_is not_bad_for our_purposes_since we are_trying to derive_an upper limit for_the real probability_of_detecting the CVs. In this way, having a disrupted disc (cold and stable) would_lower_the probability_even_further.
Once_we know that the CVs_are DNe, we are able_to compute_the recurrence time and the duration of the_outburst,_based on equations_\[TREC\] and \[TDUR\]. The results are displayed in Fig._\[Fig05\]. Notice that the recurrence time_varies from days_to_decades_and the outbursts last_a few days. Notice also a_double peak in the distribution of_the recurrence time due to the double_peak in the WD mass. Dynamically_formed CVs have greater WD_masses, implying_longer recurrence times according to_Equation \[TREC\]. Now,_for the_distribution of the_duration of the outburst, we do_not see too_many differences between dynamically formed CVs_and_CVs formed without_influence_of_dynamics. This_is due to_the_status of_the_CVs at the present-day (period bouncers)._They_mainly have periods from $\sim$ 2 to_$\sim$ 3h and according_to_Equation \[TDUR\] (which is_a function of the orbital_period) they are similar.
From Fig. \[Fig05\],_we can_see that_the chances to observe a CV during outburst are high only_for some CVs – the youngest_ones that were formed_close to_the_present-day and have_greater_donor masses_and higher mass transfer rates with respect_to the_old ones. In this way, we_have the first clue_that_a good solution would, instead of_looking for variability during an outburst,_look for it during quiescence._The_biggest_problem associated with this approach_is the need of very deep_observations because in_order to reach magnitudes similar to those_associated_with a CV during quiescence, one_needs_deeper observations in comparison with the_same_CV_during its outbursts. Although a_difficult task, it is possible. For_instance, @Cohn_2010 reached apparent visual magnitudes as deep as_$\sim$ 26 when_studying the optical counterparts of_X-ray_sources_in the GC NGC 6397.
Probability of Detection through Variability {#probability}
--------------------------------------------
Now,_once we_have separated the_CVs with respect to the disc instability and computed their_recurrence times as well as the duration_of their outbursts, we can turn to the computation of the_probability of detecting such CVs by using equation |
10 35 57.60 & +54 10 28.5 & 17.5 & DA & 09.01.92 & 251 & II & 1036+550 & 10 36 22.15 & +55 05 53.3 & 17.5 & DA & 09.01.92 & 252 & II & 1040+493 & 10 40 11.87 & +49 18 07.1 & 16.5 & sd & 29.02.92 & 277 & II & 1040+520 & 10 40 48.77 & +52 00 54.1 & 17.5 & DA & 09.01.92 & 217 & II & 1043+569 & 10 43 48.09 & +56 59 47.2 & 17.5 & F & 04.02.92 & 660 & II & 1045+570 & 10 45 15.66 & +57 03 29.4 & 17.5 & DA & 09.01.92 & 507 & II & 1047+557A & 10 47 26.47 & +55 42 25.6 & 17.0 & DA & 03.03.87 & 540 & II & 1047+557B & 10 47 51.85 & +55 43 18.3 & 17.17& QSO & 17.02.94 & 711 & II & 1049+541 & 10 49 10.46 & +54 07 31.0 & 16.0 & DA & 22.04.87 & 360 & II & 1050+582 & 10 50 51.49 & +58 16 26.1 & 17.0 & DA & 06.04.91 & 220 & II & 1051+556 & 10 51 48.63 & +55 39 08.8 & 16.5 & DA: & 10.11.91 & 114 & II & 1053+561 & 10 53 26.66 & +56 11 16.1 & 17.5 & sd & 09.01.92 & 343 & II & 1056+516A & 10 56 20.09 & +51 40 48.9 & 15.5 & Cont & 10.11.91 & 286 & II & 1057+556 & 10 57 31.09 & +55 38 47.0 & 17.5 & Cont & 08.01.91 & 540 & II & 1058+570 & 10 58 10.81 & +57 05 31.5 & 17.5 & DA & 25.02.88 & 660 & | 10 35 57.60 & +54 10 28.5 & 17.5 & DA & 09.01.92 & 251 & II & 1036 + 550 & 10 36 22.15 & +55 05 53.3 & 17.5 & DA & 09.01.92 & 252 & II & 1040 + 493 & 10 40 11.87 & +49 18 07.1 & 16.5 & sd & 29.02.92 & 277 & II & 1040 + 520 & 10 40 48.77 & +52 00 54.1 & 17.5 & DA & 09.01.92 & 217 & II & 1043 + 569 & 10 43 48.09 & +56 59 47.2 & 17.5 & F & 04.02.92 & 660 & II & 1045 + 570 & 10 45 15.66 & +57 03 29.4 & 17.5 & DA & 09.01.92 & 507 & II & 1047 + 557A & 10 47 26.47 & +55 42 25.6 & 17.0 & DA & 03.03.87 & 540 & II & 1047 + 557B & 10 47 51.85 & +55 43 18.3 & 17.17 & QSO & 17.02.94 & 711 & II & 1049 + 541 & 10 49 10.46 & +54 07 31.0 & 16.0 & DA & 22.04.87 & 360 & II & 1050 + 582 & 10 50 51.49 & +58 16 26.1 & 17.0 & DA & 06.04.91 & 220 & II & 1051 + 556 & 10 51 48.63 & +55 39 08.8 & 16.5 & DA: & 10.11.91 & 114 & II & 1053 + 561 & 10 53 26.66 & +56 11 16.1 & 17.5 & sd & 09.01.92 & 343 & II & 1056 + 516A & 10 56 20.09 & +51 40 48.9 & 15.5 & Cont & 10.11.91 & 286 & II & 1057 + 556 & 10 57 31.09 & +55 38 47.0 & 17.5 & Cont & 08.01.91 & 540 & II & 1058 + 570 & 10 58 10.81 & +57 05 31.5 & 17.5 & DA & 25.02.88 & 660 & | 10 35 57.60 & +54 10 28.5 & 17.5 & DA & 09.01.92 & 251 & II & 1036+550 & 10 36 22.15 & +55 05 53.3 & 17.5 & DA & 09.01.92 & 252 & II & 1040+493 & 10 40 11.87 & +49 18 07.1 & 16.5 & sd & 29.02.92 & 277 & II & 1040+520 & 10 40 48.77 & +52 00 54.1 & 17.5 & DA & 09.01.92 & 217 & MI & 1043+569 & 10 43 48.09 & +56 59 47.2 & 17.5 & F & 04.02.92 & 660 & II & 1045+570 & 10 45 15.66 & +57 03 29.4 & 17.5 & DW & 09.01.92 & 507 & IU & 1047+557A & 10 47 26.47 & +55 42 25.6 & 17.0 & DA & 03.03.87 & 540 & II & 1047+557B & 10 47 51.85 & +55 43 18.3 & 17.17& QSO & 17.02.94 & 711 & LI & 1049+541 & 10 49 10.46 & +54 07 31.0 & 16.0 & DA & 22.04.87 & 360 & IH & 1050+582 & 10 50 51.49 & +58 16 26.1 & 17.0 & DA & 06.04.91 & 220 & II & 1051+556 & 10 51 48.63 & +55 39 08.8 & 16.5 & QA: & 10.11.91 & 114 & IL & 1053+561 & 10 53 26.66 & +56 11 16.1 & 17.5 & sd & 09.01.92 & 343 & JP & 1056+516A & 10 56 20.09 & +51 40 48.9 & 15.5 & Cont & 10.11.91 & 286 & II & 1057+556 & 10 57 31.09 & +55 38 47.0 & 17.5 & Cont & 08.01.91 & 540 & II & 1058+570 & 10 58 10.81 & +57 05 31.5 & 17.5 & DA & 25.02.88 & 660 & | 10 35 57.60 & +54 10 28.5 & & 09.01.92 251 & II 22.15 +55 05 53.3 17.5 & DA 09.01.92 & 252 & II & & 10 40 11.87 & +49 18 07.1 & 16.5 & sd & & 277 & II & 1040+520 & 10 40 48.77 & +52 00 & & & & 217 & II & 1043+569 & 10 43 48.09 & +56 59 47.2 & 17.5 F & 04.02.92 & 660 & II & & 10 45 15.66 +57 03 29.4 & 17.5 DA 09.01.92 & & & & 10 47 & +55 42 25.6 & 17.0 & DA & 03.03.87 & 540 & II & 1047+557B & 47 51.85 43 18.3 17.17& & & 711 & 1049+541 & 10 49 10.46 & & 16.0 & DA & 22.04.87 & 360 II & & 10 50 51.49 & +58 26.1 & 17.0 & DA & 06.04.91 & & II & 1051+556 & 10 51 48.63 & +55 39 08.8 & 16.5 & 10.11.91 & 114 & & 1053+561 & 53 & 11 & 17.5 sd & 09.01.92 & 343 & II & 1056+516A & 10 20.09 & +51 40 48.9 & 15.5 & Cont & 286 II & 1057+556 10 57 31.09 & 38 & 17.5 & Cont & & & 58 & +57 05 31.5 17.5 & DA & 25.02.88 660 & | 10 35 57.60 & +54 10 28.5 & 17.5 & DA & 09.01.92 & 251 & II & 1036+550 & 10 36 22.15 & +55 05 53.3 & 17.5 & DA & 09.01.92 & 252 & II & 1040+493 & 10 40 11.87 & +49 18 07.1 & 16.5 & sd & 29.02.92 & 277 & II & 1040+520 & 10 40 48.77 & +52 00 54.1 & 17.5 & DA & 09.01.92 & 217 & II & 1043+569 & 10 43 48.09 & +56 59 47.2 & 17.5 & F & 04.02.92 & 660 & II & 1045+570 & 10 45 15.66 & +57 03 29.4 & 17.5 & DA & 09.01.92 & 507 & II & 1047+557A & 10 47 26.47 & +55 42 25.6 & 17.0 & Da & 03.03.87 & 540 & II & 1047+557B & 10 47 51.85 & +55 43 18.3 & 17.17& QSO & 17.02.94 & 711 & II & 1049+541 & 10 49 10.46 & +54 07 31.0 & 16.0 & Da & 22.04.87 & 360 & II & 1050+582 & 10 50 51.49 & +58 16 26.1 & 17.0 & DA & 06.04.91 & 220 & iI & 1051+556 & 10 51 48.63 & +55 39 08.8 & 16.5 & Da: & 10.11.91 & 114 & II & 1053+561 & 10 53 26.66 & +56 11 16.1 & 17.5 & Sd & 09.01.92 & 343 & iI & 1056+516A & 10 56 20.09 & +51 40 48.9 & 15.5 & COnt & 10.11.91 & 286 & Ii & 1057+556 & 10 57 31.09 & +55 38 47.0 & 17.5 & Cont & 08.01.91 & 540 & II & 1058+570 & 10 58 10.81 & +57 05 31.5 & 17.5 & DA & 25.02.88 & 660 & | 10 35 57.60 & +54 10 28.5 & 17.5 &DA &09. 01. 92 & 2 51 & II & 1036+550 & 10 36 22.15 & +55 05 53. 3 & 1 7. 5 & D A & 09.0 1.92 &2 52 & II & 1 040 +4 9 3& 104011.87 & +49 18 07 .1&16.5 & sd &2 9. 02.92 & 27 7 & II & 1040+5 20& 10 4 048. 7 7 & + 5200 54 .1 & 1 7 .5 & D A & 09.01 .9 2 & 217 & II &1 0 43 +569 & 10 43 48.09 &+ 56 59 47.2 & 17.5 & F & 0 4 .0 2 . 92& 6 60 & II &10 45+57 0 & 10 4 5 1 5 . 6 6 & +57 03 29.4 & 17.5 & DA& 09 .01.92 & 50 7 & II& 104 7+ 5 57A & 10 47 26 .47& +55 4225.6 & 17.0 &D A & 03. 03.87& 5 40& II &10 47+ 55 7 B & 10 47 51. 85 & +55 4 318.3& 17 . 1 7 & QSO &17.0 2.94& 711 & II &104 9+54 1 &10 49 10.4 6 &+5 4 0731.0 & 16.0 & DA & 22.04.87& 36 0 & II &105 0+ 582 & 10 5 0 51.49 &+58 16 26. 1 & 17. 0 &DA & 06 .04.91 & 220 & II&1 0 51 +556 & 1 0 51 4 8 .6 3& +55 3908 .8& 16 . 5 & DA : &1 0. 11.91 &114 &I I&1053+56 1& 10 5 326. 66& +56 11 1 6.1 &17.5 & s d & 0 9 .01.92 & 343 & II & 1056+516 A & 1 05 6 20 .09 & +51 40 4 8.9& 15. 5 &C on t & 10.11 .91 & 2 8 6& II & 1057+556 & 10 5 7 31.0 9 & + 55 38 47.0 &17.5 & Con t & 08.01.9 1 &5 40 & II & 1058+57 0 & 1 0 58 10.81 & +57 05 31.5 & 17.5& DA & 25 . 0 2.88 & 6 60& | 10_35 57.60_& +54 10 28.5_& 17.5_&_DA &_09.01.92_& 251 &_II & 1036+550_& 10 36 22.15_& +55 05_53.3_& 17.5 & DA & 09.01.92 & 252 & II & 1040+493 & 10_40_11.87 &_+49_18_07.1 & 16.5 & sd_& 29.02.92 & 277 &_II &_1040+520 & 10 40 48.77 & +52 00_54.1_& 17.5 &_DA & 09.01.92 & 217 & II & 1043+569_& 10 43 48.09 & +56_59 47.2 &_17.5_&_F & 04.02.92 &_660 & II & 1045+570 &_10 45 15.66 & +57 03_29.4 & 17.5 & DA & 09.01.92_& 507 & II & 1047+557A_& 10 47 26.47 &_+55 42_25.6 & 17.0 & DA_& 03.03.87 &_540 &_II & 1047+557B_& 10 47 51.85 & +55_43 18.3 &_17.17& QSO & 17.02.94 & 711_&_II & 1049+541_&_10_49 10.46_& +54 07_31.0_& 16.0_&_DA & 22.04.87 & 360 &_II_& 1050+582 & 10 50 51.49 &_+58 16 26.1 &_17.0_& DA & 06.04.91_& 220 & II &_1051+556 & 10 51 48.63 &_+55 39_08.8 &_16.5 & DA: & 10.11.91 & 114 & II & 1053+561_& 10 53 26.66 & +56_11 16.1 & 17.5_& sd_&_09.01.92 & 343_&_II &_1056+516A & 10 56 20.09 & +51_40 48.9_& 15.5 & Cont & 10.11.91_& 286 & II_&_1057+556 & 10 57 31.09 &_+55 38 47.0 & 17.5 &_Cont & 08.01.91 & 540_&_II_& 1058+570 & 10 58_10.81 & +57 05 31.5 &_17.5 & DA_& 25.02.88 & 660 & |
a root, if $\mathfrak g_\alpha\neq\{0\}$; the set of all roots is denoted by $\Delta$. Moreover, $\mathfrak g_0$ is $\sigma$-stable and inherits the structure of symmetric pair, whence $\mathfrak g_0=\mathfrak k_0\oplus\mathfrak p_0$ its Cartan decomposition. Finally, for a root $\alpha$ in $\Delta$, $\sigma(\mathfrak g_\alpha)=\mathfrak g_{-\alpha}$, whence $\mathfrak g$ admits a triangular decomposition $\mathfrak g=\mathfrak n_-\oplus\mathfrak g_0\oplus\mathfrak n_+$, where $\mathfrak n_\pm$ denotes the direct sum of all root spaces associated to positive/negative roots.
From the triangular decomposition follows the generalized Iwasawa decomposition $\mathfrak g=\mathfrak k\oplus\mathfrak p_0\oplus \mathfrak n_+$: in fact, see also [@Dix Proposition 1.13.11], $\mathfrak g=\mathfrak k+\mathfrak p_0+\mathfrak n_+$ directly from the previous triangular decomposition. Namely, a general element $x$ of $\mathfrak g$ can be (uniquely) written as $x=x_-+k_0+p_0+x_+$, with $x_\pm$ in $\mathfrak n_\pm$, $k_0$ in $\mathfrak k_0$ and $p_0$ in $\mathfrak p_0$. Then, $x_-=(x_-+\sigma(x_-))-\sigma(x_-)$, and obviously $x_-+\sigma(x_-)$ belongs to $\mathfrak k$, while $\sigma(x_-)$ is in $\mathfrak n_+$, whence the first claim follows. Now, we assume $k+p_0+x_+=0$: then, $\sigma(k+p_0+x_+)=k-p_0+\sigma(x_+)=0$, whence $2p_0+x_+-\sigma(x_+)=0$. From the triangular decomposition of $\mathfrak g$ follows automatically $p_0=x_+=0$, and thus also $k=0$. The relations for the triangular decomposition of $\mathfrak g$ imply that $\mathfrak k_0\oplus \mathfrak n_+$ is a Lie subalgebra of $\mathfrak g$, and thus is $(\mathfrak k_0\oplus\mathfrak n_+)^\perp$ a coisotropic submanifold of $\mathfrak g^*$.
\[p | a root, if $ \mathfrak g_\alpha\neq\{0\}$; the set of all roots is denoted by $ \Delta$. furthermore, $ \mathfrak g_0 $ is $ \sigma$-stable and inherit the structure of symmetric couple, whence $ \mathfrak g_0=\mathfrak k_0\oplus\mathfrak p_0 $ its Cartan decomposition. last, for a root $ \alpha$ in $ \Delta$, $ \sigma(\mathfrak g_\alpha)=\mathfrak g_{-\alpha}$, whence $ \mathfrak g$ admits a triangular decay $ \mathfrak g=\mathfrak n_-\oplus\mathfrak g_0\oplus\mathfrak n_+$, where $ \mathfrak n_\pm$ denote the direct sum of all root spaces consociate to positive / negative etymon.
From the triangular decomposition follows the generalized Iwasawa decomposition $ \mathfrak g=\mathfrak k\oplus\mathfrak p_0\oplus \mathfrak n_+$: in fact, see also [ @Dix Proposition 1.13.11 ], $ \mathfrak g=\mathfrak k+\mathfrak p_0+\mathfrak n_+$ immediately from the previous triangular decomposition. Namely, a general component $ x$ of $ \mathfrak g$ can be (uniquely) written as $ x = x_-+k_0+p_0+x_+$, with $ x_\pm$ in $ \mathfrak n_\pm$, $ k_0 $ in $ \mathfrak k_0 $ and $ p_0 $ in $ \mathfrak p_0$. Then, $ x_-=(x_-+\sigma(x_-))-\sigma(x_-)$, and obviously $ x_-+\sigma(x_-)$ belong to $ \mathfrak k$, while $ \sigma(x_-)$ is in $ \mathfrak n_+$, whence the first title follows. Now, we assume $ k+p_0+x_+=0 $: then, $ \sigma(k+p_0+x_+)=k - p_0+\sigma(x_+)=0 $, whence $ 2p_0+x_+-\sigma(x_+)=0$. From the triangular decomposition of $ \mathfrak g$ follows mechanically $ p_0 = x_+=0 $, and thus also $ k=0$. The relations for the triangular decomposition of $ \mathfrak g$ imply that $ \mathfrak k_0\oplus \mathfrak n_+$ is a Lie subalgebra of $ \mathfrak g$, and thus is $ (\mathfrak k_0\oplus\mathfrak n_+)^\perp$ a coisotropic submanifold of $ \mathfrak g^*$.
\[p | a goot, if $\mathfrak g_\alpha\ntq\{0\}$; the set of all roots ms denofed by $\Ddlta$. Moreover, $\mathfrak g_0$ is $\ditma$-stqble and inherits the rtructure of symmwtrir pair, whence $\mavgfrak g_0=\mathfrzn k_0\o'lns\mathfrak p_0$ itx Cartan dacomposition. Fhnxlpy, for a root $\alpha$ in $\Delta$, $\sigma(\mwthfrak g_\wlpha)=\mathfrak d_{-\alpnw}$, whsnce $\mathfrak g$ admits a triangulad decomkosition $\mathfrak b=\mathfrak n_-\oplus\mathfrak g_0\lplud\mathfrak n_+$, where $\lathfrak n_\pn$ degites the dirdct sum of all root spzces associated to positive/negagive xoots.
From tye trldngular decimposption follows the genaralizec Iwasawa decokpovituon $\mathfrak g=\mathfran k\oplus\mathfrak p_0\jplus \matvfxak n_+$: in fact, see alsi [@Dux Prmposhtiov 1.13.11], $\mxthrrek f=\mathfgak k+\mathfrak p_0+\mathfrak b_+$ directly from the pwvfious trianghlar dqcjmposition. Namely, a general element $x$ ox $\mzthfrak g$ can be (uniqueoy) written as $x=x_-+k_0+p_0+x_+$, wlth $x_\pm$ ig $\mathfrak n_\pm$, $k_0$ in $\mathfrak k_0$ and $p_0$ in $\mathfrak [_0$. Thei, $b_-=(x_-+\snnix(z_-))-\slgma(x_-)$, and obviously $x_-+\sigma(x_-)$ belongs to $\mathfrwi l$, ehile $\sigma(x_-)$ if in $\mathfrsk n_+$, rhence the fitst clanj rollows. Now, we asskme $k+p_0+x_+=0$: then, $\sigma(k+p_0+x_+)=h-p_0+\sibma(x_+)=0$, whence $2p_0+x_+-\sigma(x_+)=0$. From tye triangulag dexomposition of $\matkfrak g$ folluws sutomstically $p_0=x_+=0$, and thus alro $k=0$. The relatilns for tgd triangular decumppshtion of $\mathfrak g$ imply ehat $\mathhrak l_0\oplus \oathgrak n_+$ is a Lie dubalnabra of $\mathfrak g$, and jhus iv $(\mathfrak k_0\oplus\mathfrak n_+)^\perp$ a coisotropic submanifokd of $\mathfrah g^*$.
\[p | a root, if $\mathfrak g_\alpha\neq\{0\}$; the set roots denoted by Moreover, $\mathfrak g_0$ structure symmetric pair, whence g_0=\mathfrak k_0\oplus\mathfrak p_0$ Cartan decomposition. Finally, for a root in $\Delta$, $\sigma(\mathfrak g_\alpha)=\mathfrak g_{-\alpha}$, whence $\mathfrak g$ admits a triangular decomposition $\mathfrak n_-\oplus\mathfrak g_0\oplus\mathfrak n_+$, where $\mathfrak n_\pm$ denotes the direct sum of all root associated positive/negative From triangular decomposition follows the generalized Iwasawa decomposition $\mathfrak g=\mathfrak k\oplus\mathfrak p_0\oplus \mathfrak n_+$: in fact, see [@Dix Proposition 1.13.11], $\mathfrak g=\mathfrak k+\mathfrak p_0+\mathfrak n_+$ from the previous triangular Namely, a general element $x$ $\mathfrak can be written $x=x_-+k_0+p_0+x_+$, $x_\pm$ in $\mathfrak $k_0$ in $\mathfrak k_0$ and $p_0$ in $\mathfrak p_0$. Then, $x_-=(x_-+\sigma(x_-))-\sigma(x_-)$, and obviously $x_-+\sigma(x_-)$ belongs to $\mathfrak while $\sigma(x_-)$ $\mathfrak n_+$, the claim Now, we assume $\sigma(k+p_0+x_+)=k-p_0+\sigma(x_+)=0$, whence $2p_0+x_+-\sigma(x_+)=0$. From the triangular g$ follows automatically $p_0=x_+=0$, and thus also $k=0$. relations for triangular decomposition of $\mathfrak g$ imply $\mathfrak k_0\oplus \mathfrak n_+$ is a Lie subalgebra $\mathfrak g$, and thus is $(\mathfrak k_0\oplus\mathfrak n_+)^\perp$ a coisotropic submanifold of $\mathfrak g^*$. \[p | a root, if $\mathfrak g_\alpha\neq\{0\}$; tHe set of all Roots Is dEnoTeD by $\DElta$. moreover, $\mathfrAK g_0$ is $\Sigma$-stable and inherits The stRuCTure OF sYmmetRic pair, WHeNCE $\maThFrAk g_0=\MaTHfRak k_0\oPluS\mathfrAk p_0$ its CartAn dEcOmposition. FiNAlLy, for a root $\AlpHa$ in $\Delta$, $\sigMa(\mAthfraK g_\AlpHA)=\mathFraK g_{-\alpHa}$, whenCE $\mathfRak g$ admitS a TRianguLAr decomPOSiTion $\Mathfrak g=\mathfrak N_-\OpLUs\mathfrak g_0\oplUs\mathFrAK n_+$, WHEre $\MatHfrak n_\pm$ deNoTes thE Direct sUM oF ALL roOT spaces associAted to positIVe/nEgativE rOotS.
from thE triaNgULar DecompositiOn foLlows the gEneralIZed IwasAWa decomPositiOn $\mAthFrak G=\MaThFraK k\OPluS\MaThfRAk p_0\Oplus \matHfRaK n_+$: in fAct, sEE ALSo [@DiX PrOposItion 1.13.11], $\Mathfrak g=\mathFraK k+\maTHfrAk p_0+\maThfraK n_+$ diReCtly fRom the PreviOuS triangular decoMposItion. NameLy, a GeNerAl ElemeNT $x$ of $\maThfRak G$ can be (uNiquely) WRitTeN AS $X=x_-+K_0+p_0+x_+$, with $x_\pm$ in $\mathfrAk N_\PM$, $k_0$ In $\mathfrAk k_0$ and $P_0$ In $\MaTHfrak p_0$. ThEn, $X_-=(x_-+\sIgma(X_-))-\SIgma(x_-)$, And oBViOusly $x_-+\siGma(x_-)$ beLOnGs To $\mathfRaK k$, whilE $\sIgmA(x_-)$ iS in $\maTHfraK n_+$, whenCe the firSt claIM follows. Now, we aSSume $k+p_0+x_+=0$: then, $\siGMa(K+P_0+X_+)=k-P_0+\SigmA(x_+)=0$, wHence $2p_0+x_+-\sigmA(x_+)=0$. FrOM the TriaNGuLar DEcompOsitiOn OF $\mAThfrak g$ follows automAtIcally $P_0=x_+=0$, and Thus also $k=0$. The rElations foR THE trianguLar dECoMPosition of $\mathFrak g$ Imply that $\mAThfrak k_0\oPlus \mAthfrak n_+$ Is a Lie subALGebra of $\mAthFraK g$, aNd tHUS iS $(\mathfrak k_0\oplUS\MathFrAk n_+)^\perp$ A coIsotropIc sUbmAniFolD oF $\mathfrak G^*$.
\[p | a root, if $\mathfrak g_\ alpha\neq\ {0\}$ ; t hese t of all roots is deno t ed b y $\Delta$. Moreover,$\mat hf r ak g _ 0$ is $ \sigma$ - st a b lean dinh er i ts thestr uctureof symmetr icpa ir, whence $ \ ma thfrak g_0 =\m athfrak k_0\ opl us\mat hf rak p_0$its Cart an dec o mposit ion. Fina ll y , fora root $ \ a lp ha$in $\Delta$, $\si g ma ( \mathfrak g_\a lpha)= \m a th f r akg_{ -\alpha}$, w hence $\mathf r ak g $ ad m its a triangu lar decompo s iti on $\m at hfr a k g=\m athfr ak n_- \oplus\math frak g_0\oplu s\math f rak n_+ $ , where $\mat hfr akn_\p m $de not es the di rec t su m of all r oo t spa cesa s s o ciat edto p ositi ve/negative r oot s.
F rom thetrian gula rdecom positi on fo ll ows the general ized Iwasawadec om pos it ion $ \ mathfr akg=\ mathfra k k\opl u s\m at h f r ak p_0\oplus \mathfr ak n _+ $: in fa ct, se e a ls o [@Dix P ro pos itio n 1.13. 11], $\ mathfrak g=\ma t hf ra k k+\ma th frak p _0 +\m ath frakn _+$direct ly fromthe p r evious triangu l ar decomposit i on . Na m ely, ageneral ele ment $x$of $ \ ma thf r ak g$ canbe (u n iquely) written as$x =x_-+k _0+p_ 0+x_+$, with$x_\pm$ in $ \ mathfrak n_\ p m$ , $k_0$ in $\ma thfra k k_0$ and $p_0$ in $\ma thfrak p _0$. Then , $x_-=(x_ -+\ sig ma( x_- ) ) -\ sigma(x_-)$,a n d ob vi ously $ x_- +\sigma (x_ -)$ be lon gs to $\mat hfrak k$ ,wh il e$\s igma( x _-)$ isin $\ ma thf rak n _ +$, wh encethefi rs t cl aim fol l ow s . Now ,we ass ume $ k+p_0 +x_+ = 0$: then,$\sigma(k +p_ 0 +x_+ )= k- p_0+\si gma(x_+)=0$,wh ence $2p_0 +x _+- \sigma ( x _+)=0$.From the triangular dec o mpositi onof $\ math frak g$ f oll ows au tom a ticall y $p_0 =x_+= 0$ , a n d thus a ls o $ k= 0$. The re l a tio ns fo rthetriangu lar decompositiono f $ \mathfrak g$imp ly t h a t$\m a th f rak k _ 0\o p l us \mathfrak n_ +$ is a Li es ub algebra of $\m at hfrak g $, andthusi s $(\ma thfrak k_ 0\oplus\m at hfra k n_+ )^\perp$ a coisotr opic subm a nifol d o f $\m ath frak g ^* $.
\[p | a_root, if_$\mathfrak g_\alpha\neq\{0\}$; the set_of all_roots_is denoted_by_$\Delta$. Moreover, $\mathfrak_g_0$ is $\sigma$-stable_and inherits the structure_of symmetric pair,_whence_$\mathfrak g_0=\mathfrak k_0\oplus\mathfrak p_0$ its Cartan decomposition. Finally, for a root $\alpha$ in $\Delta$,_$\sigma(\mathfrak_g_\alpha)=\mathfrak g_{-\alpha}$,_whence_$\mathfrak_g$ admits a triangular decomposition_$\mathfrak g=\mathfrak n_-\oplus\mathfrak g_0\oplus\mathfrak n_+$,_where $\mathfrak_n_\pm$ denotes the direct sum of all root_spaces_associated to positive/negative_roots.
From the triangular decomposition follows the generalized Iwasawa decomposition_$\mathfrak g=\mathfrak k\oplus\mathfrak p_0\oplus \mathfrak n_+$:_in fact, see_also [@Dix_Proposition_1.13.11], $\mathfrak g=\mathfrak k+\mathfrak_p_0+\mathfrak n_+$ directly from the previous_triangular decomposition. Namely, a general element_$x$ of $\mathfrak g$ can be (uniquely)_written as $x=x_-+k_0+p_0+x_+$, with $x_\pm$ in_$\mathfrak n_\pm$, $k_0$ in $\mathfrak_k_0$ and_$p_0$ in $\mathfrak p_0$. Then,_$x_-=(x_-+\sigma(x_-))-\sigma(x_-)$, and obviously_$x_-+\sigma(x_-)$ belongs_to $\mathfrak k$,_while $\sigma(x_-)$ is in $\mathfrak n_+$,_whence the first_claim follows. Now, we assume $k+p_0+x_+=0$:_then,_$\sigma(k+p_0+x_+)=k-p_0+\sigma(x_+)=0$, whence $2p_0+x_+-\sigma(x_+)=0$._From_the_triangular decomposition_of $\mathfrak g$_follows_automatically $p_0=x_+=0$,_and_thus also $k=0$. The relations for_the_triangular decomposition of $\mathfrak g$ imply that_$\mathfrak k_0\oplus \mathfrak n_+$_is_a Lie subalgebra of_$\mathfrak g$, and thus is_$(\mathfrak k_0\oplus\mathfrak n_+)^\perp$ a coisotropic submanifold_of $\mathfrak_g^*$.
\[p |
(0.5, 0.8) [ $f$]{}; at (2/5, -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [$t_3$]{}; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02); at (1, -0.07) [$e$]{}; (3/5,3/4)–(4/5,1/2);
(0,0)–(1,0)–(1,1)–(0,1)–(0,0); (0,0)–(0.2,0.85)–(0.4,0.5)–(0.6,0.75)–(0.8,0.25)–(1,1); (0,-0.02)–(0,0.02); at (0.5, 0.8) [ $g$]{}; at (0, -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [$t_3$]{}; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02); at (1, -0.07) [$e$]{}; (0.6,0.75)–(0.8,0.25);
\[lem:zigzag\] Let $f: I\to I$ be a continuous piecewise linear surjection with critical points $0=t_0<t_1<\ldots <t_m<t_{m+1}=1$. If there is $k\in\{0, \ldots, m\}$ such that $I_k=[t_k,
t_{k+1}]$ is not inside a zigzag of $f$, then there exists an admissible permutation $ | (0.5, 0.8) [ $ f$ ] { }; at (2/5, -0.07) [ $ a$ ] { }; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [ $ t_3 $ ] { }; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [ $ x$ ] { }; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [ $ t_4 $ ] { }; (1,-0.02)–(1,0.02); at (1, -0.07) [ $ e$ ] { }; (3/5,3/4)–(4/5,1/2);
(0,0)–(1,0)–(1,1)–(0,1)–(0,0); (0,0)–(0.2,0.85)–(0.4,0.5)–(0.6,0.75)–(0.8,0.25)–(1,1); (0,-0.02)–(0,0.02); at (0.5, 0.8) [ $ g$ ] { }; at (0, -0.07) [ $ a$ ] { }; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [ $ t_3 $ ] { }; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [ $ x$ ] { }; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [ $ t_4 $ ] { }; (1,-0.02)–(1,0.02); at (1, -0.07) [ $ e$ ] { }; (0.6,0.75)–(0.8,0.25);
\[lem: zigzag\ ] Let $ f: I\to I$ be a continuous piecewise linear surjection with critical points $ 0 = t_0 < t_1<\ldots < t_m < t_{m+1}=1$. If there is $ k\in\{0, \ldots, m\}$ such that $ I_k=[t_k,
t_{k+1}]$ is not inside a zigzag of $ f$, then there exists an admissible substitution $ | (0.5, 0.8) [ $f$]{}; at (2/5, -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [$t_3$]{}; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [$w$]{}; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02); at (1, -0.07) [$g$]{}; (3/5,3/4)–(4/5,1/2);
(0,0)–(1,0)–(1,1)–(0,1)–(0,0); (0,0)–(0.2,0.85)–(0.4,0.5)–(0.6,0.75)–(0.8,0.25)–(1,1); (0,-0.02)–(0,0.02); at (0.5, 0.8) [ $g$]{}; af (0, -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02); ag (3/5, -0.07) [$t_3$]{}; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02); at (1, -0.07) [$x$]{}; (0.6,0.75)–(0.8,0.25);
\[len:zigzqg\] Let $f: I\to I$ be a covtinuous iiecewise lintar surjection wivg critigcl pojkts $0=t_0<c_1<\lvots <t_m<t_{m+1}=1$. If thgre is $k\in\{0, \lgots, m\}$ such thdt $I_n=[t_k,
t_{k+1}]$ is not inside a zigzag of $f$, ehen thrrf exists an adiisspbje psgmmtation $ | (0.5, 0.8) [ $f$]{}; at (2/5, -0.07) at -0.07) [$t_3$]{}; at (7/10, -0.07) [$t_4$]{}; at (1, -0.07) (3/5,3/4)–(4/5,1/2); (0,0)–(1,0)–(1,1)–(0,1)–(0,0); (0,0)–(0.2,0.85)–(0.4,0.5)–(0.6,0.75)–(0.8,0.25)–(1,1); at (0.5, 0.8) [ $g$]{}; at -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [$t_3$]{}; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02); (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02); at (1, -0.07) [$e$]{}; (0.6,0.75)–(0.8,0.25); \[lem:zigzag\] Let $f: I\to be continuous linear with critical points $0=t_0<t_1<\ldots <t_m<t_{m+1}=1$. If there is $k\in\{0, \ldots, m\}$ such that $I_k=[t_k, t_{k+1}]$ is inside a zigzag of $f$, then there exists admissible permutation $ | (0.5, 0.8) [ $f$]{}; at (2/5, -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [$t_3$]{}; (7/10,-0.02)–(7/10,0.02); at (7/10, -0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02); at (1, -0.07) [$e$]{}; (3/5,3/4)–(4/5,1/2);
(0,0)–(1,0)–(1,1)–(0,1)–(0,0); (0,0)–(0.2,0.85)–(0.4,0.5)–(0.6,0.75)–(0.8,0.25)–(1,1); (0,-0.02)–(0,0.02); at (0.5, 0.8) [ $g$]{}; at (0, -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02); at (3/5, -0.07) [$t_3$]{}; (7/10,-0.02)–(7/10,0.02); At (7/10, -0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02); at (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02); at (1, -0.07) [$e$]{}; (0.6,0.75)–(0.8,0.25);
\[lEm:zigZag\] let $F: I\To I$ bE a coNtinuous piecewISe liNear surjection with critIcal pOiNTs $0=t_0<t_1<\LDoTs <t_m<t_{M+1}=1$. If therE Is $K\IN\{0, \ldOtS, m\}$ SucH tHAt $i_k=[t_k,
t_{K+1}]$ is Not insiDe a zigzag oF $f$, tHeN there exists AN aDmissible pErmUtation $ | (0.5, 0.8) [ $f$]{}; at ( 2/5, -0.07 ) [$a $]{ };(3 /5,- 0.02 )–(3/5,0.02);a t (3 /5, -0.07) [$t_3$]{};(7/10 ,- 0 .02) – (7 /10,0 .02); a t ( 7 / 10, - 0. 07) [ $ x$ ]{};(4/ 5,-0.02 )–(4/5,0.0 2); a t (4/5, -0.0 7 )[$t_4$]{}; (1 ,-0.02)–(1,0 .02 ); at(1 , - 0 .07)[$e $]{}; (3/5, 3 /4)–(4 /5,1/2);
( 0 ,0)–(1 , 0)–(1,1 ) – (0 ,1)– (0,0); (0,0)–(0.2 , 0. 8 5)–(0.4,0.5)–( 0.6,0. 75 ) –( 0 . 8,0 .25 )–(1,1); ( 0, -0.02 ) –(0,0.0 2 ); a t (0 . 5, 0.8) [ $g$ ]{}; at (0, -0. 07) [$ a$ ]{} ; (3/5, -0.02 )– ( 3/5 ,0.02); at(3/5 , -0.07)[$t_3$ ] {}; (7/ 1 0,-0.02 )–(7/1 0,0 .02 ); a t ( 7/ 10, - 0 .07 ) [ $x$ ] {}; (4/5,-0 .0 2) –(4/5 ,0.0 2 ) ; at ( 4/5 , -0 .07)[$t_4$]{}; (1 ,-0 .02) – (1, 0.02) ; at(1,-0 .07)[$e$]{ }; (0 .6 ,0.75)–(0.8,0.2 5);
\[lem:zi gza g\ ] L et $f:I \to I$ be acontinu ous pie c ewi se l i ne ar surjection with c r i ti cal poin ts $0= t _0 <t _ 1<\ldots < t_m <t_{ m + 1}=1$ . If th ere is $ k\in\{ 0 ,\l dots, m \} $ such t hat $I _k=[t _ k, t_{k+1 }]$ is n ot in s ide a zigzag o f $f$, then th e re e xi s ts a n a dmissible p ermu t atio n $ | (0.5,_0.8) [_$f$]{}; at (2/5, -0.07)_[$a$]{}; (3/5,-0.02)–(3/5,0.02);_at_(3/5, -0.07)_[$t_3$]{};_(7/10,-0.02)–(7/10,0.02); at (7/10,_-0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02);_at (4/5, -0.07) [$t_4$]{};_(1,-0.02)–(1,0.02); at (1,_-0.07)_[$e$]{}; (3/5,3/4)–(4/5,1/2);
(0,0)–(1,0)–(1,1)–(0,1)–(0,0); (0,0)–(0.2,0.85)–(0.4,0.5)–(0.6,0.75)–(0.8,0.25)–(1,1); (0,-0.02)–(0,0.02); at (0.5, 0.8) [ $g$]{}; at (0, -0.07) [$a$]{}; (3/5,-0.02)–(3/5,0.02);_at_(3/5, -0.07)_[$t_3$]{};_(7/10,-0.02)–(7/10,0.02);_at (7/10, -0.07) [$x$]{}; (4/5,-0.02)–(4/5,0.02);_at (4/5, -0.07) [$t_4$]{}; (1,-0.02)–(1,0.02);_at (1,_-0.07) [$e$]{}; (0.6,0.75)–(0.8,0.25);
\[lem:zigzag\] Let $f: I\to I$ be_a_continuous piecewise linear_surjection with critical points $0=t_0<t_1<\ldots <t_m<t_{m+1}=1$. If there is_$k\in\{0, \ldots, m\}$ such that $I_k=[t_k,_
t_{k+1}]$ is not_inside_a_zigzag of $f$, then_there exists an admissible permutation $ |
end{array}
\right] \,,$$ where $\mathbf{P}^{\rm CLASS}(N)$ is the solution in Eq. . Even though the second term on the rhs of Eq. still is a rapidly oscillating function, it cannot be discarded as simply as others due to the presence of $h^\prime$.
The last integral in Eq. can be integrated iteratively by parts, and in such case it can be proved that if we retain the dominant term in the integrations we obtain[^12]: $$\label{eq:79}
\mathbf{P}(N,\theta) \simeq \mathbf{P}^{\rm CLASS}(N) + \frac{
h^\prime}{2} \mathbf{N}_2(\theta) \,, \quad \mathbf{N}_2(\theta)
= p \left[
\begin{array}{c}
{\rm Si}(\theta) - {\rm Si}(\theta_\star) \\
{\rm Ci}(\theta) - {\rm Ci}(\theta_\star)
\end{array}
\right] \,.$$ We can apply here the same kind of trick we used in Sec. \[sec:gener-solut-pert\], where we inferred that for the growing case: $\mathbf{P}^{\rm CLASS} \sim h^\prime/2$, see Eq. . Thus, the second term on the rhs of Eq. adds a negligible oscillating function to the general solution; in other words, we have just found that $\mathbf{P}(N,\theta) \simeq \mathbf{P}^{CLASS}(N) (1 +
\mathcal{O}(10^{-2}))$, which is the same level of accuracy attained for the background variables.
[^1]: This means that the evolution of the scalar field $\phi $ starts on the left branch of the parabolic potential , which is not a limitation as the system of equations is invariant under the change $\phi \to -\phi$. To start the evolution on the right branch, the signs of $y,y_1$ just need to be changed.
[^2]: Strictly speaking, there is no need to consider $y_1$ as a separate variable, nor to take Eq. as an independent equation of motion. As shown in Eqs. , $y_1$ is just proportional to $H^{-1}$, and then only Eqs. and are enough to represent | end{array }
\right ] \,,$$ where $ \mathbf{P}^{\rm CLASS}(N)$ is the solution in Eq. . Even though the second term on the rhs of Eq. still is a rapidly hover affair, it cannot be discarded as simply as others due to the bearing of $ h^\prime$.
The last integral in Eq. can be integrated iteratively by parts, and in such font it can be proved that if we retain the dominant condition in the integrations we obtain[^12 ]: $ $ \label{eq:79 }
\mathbf{P}(N,\theta) \simeq \mathbf{P}^{\rm CLASS}(N) + \frac {
h^\prime}{2 } \mathbf{N}_2(\theta) \, , \quad \mathbf{N}_2(\theta)
= p \left [
\begin{array}{c }
{ \rm Si}(\theta) - { \rm Si}(\theta_\star) \\
{ \rm Ci}(\theta) - { \rm Ci}(\theta_\star)
\end{array }
\right ] \,.$$ We can give here the same kind of trick we use in Sec. \[sec: gener - solut - pert\ ], where we guess that for the growing subject: $ \mathbf{P}^{\rm CLASS } \sim h^\prime/2 $, see Eq. . therefore, the second term on the rhs of Eq. adds a negligible oscillating routine to the general solution; in other bible, we have just recover that $ \mathbf{P}(N,\theta) \simeq \mathbf{P}^{CLASS}(N) (1 +
\mathcal{O}(10^{-2}))$, which is the same level of accuracy attained for the background variables.
[ ^1 ]: This mean that the evolution of the scalar field $ \phi $ starts on the left branch of the parabolic potential , which is not a limitation as the system of equations is invariant under the change $ \phi \to -\phi$. To start the evolution on the right branch, the signs of $ y, y_1 $ just necessitate to be changed.
[ ^2 ]: Strictly speak, there be no need to consider $ y_1 $ as a freestanding variable, nor to take Eq. as an independent equation of apparent motion. As shown in Eqs. , $ y_1 $ is just proportional to $ H^{-1}$, and then only Eqs. and are enough to represent | end{wrray}
\right] \,,$$ where $\mathbf{K}^{\rm CLASS}(N)$ is the solutimn in Sq. . Even ghough the second term on thx rhw of Tz. still is a rapiduy oscillwting fubctiib, it cannov be disgcrded ws snm'ly as others doe to the prasence of $h^\prike$.
Ghz last integral in Eq. can be integraeed itetahively by partf, anc in alcm case it can be proved that if se retapn the dominant trrm in the integrations we obtwin[^12]: $$\label{eq:79}
\mathbv{P}(N,\theta) \sineq \iqthbf{P}^{\rm CLARS}(N) + \frac{
h^\prime}{2} \majhbf{N}_2(\theta) \,, \quad \mathbf{N}_2(\theta)
= o \lefc[
\begin{arrai}{e}
{\em Dh}(\theta) - {\rm Wi}(\thena_\star) \\
{\rm Ci}(\tmvta) - {\rm Ci}(\thets_\star)
\end{array}
\rlght] \,.$$ We can apply here the seme kind of trick we used in Vee. \[sec:gener-solut-pert\], whwrw we hnfesred rhag fkr tge groainj case: $\mathgf{P}^{\rm CLASS} \sim h^\prime/2$, see Eq. . Uhuf, the second tsrm on tre rhs of Eq. adds a negligible oscillatpng runction to the general solution; in other wotds, we havq just found that $\mathbf{P}(N,\theta) \simeq \mathbf{P}^{CLASV}(N) (1 +
\metfcao{O}(10^{-2}))$, dyifh is the same level of accuracy attained for fht bsckground varicbles.
[^1]: This means tjay the evolutiov of tks acalar field $\phi $ dtarts jn thw left brwnch of the parabolic potential , which is non a oimitation as the dystem of eduatoons os invariant under the ehange $\phi \to -\phi$. To start ghe evolution on thv richt branch, the signs of $y,y_1$ just neev to ye changdd.
[^2]: Sjrictly speaking, hhere is no need to conslder $i_1$ as a separate gariable, nor to take Eq. as an iivependent equsthon of motijn. As shown in Eqs. , $y_1$ is just prokortional to $H^{-1}$, and then knly Eqv. and are egough to repradent | end{array} \right] \,,$$ where $\mathbf{P}^{\rm CLASS}(N)$ is in . Even the second term still a rapidly oscillating it cannot be as simply as others due to presence of $h^\prime$. The last integral in Eq. can be integrated iteratively by and in such case it can be proved that if we retain the term the we $$\label{eq:79} \mathbf{P}(N,\theta) \simeq \mathbf{P}^{\rm CLASS}(N) + \frac{ h^\prime}{2} \mathbf{N}_2(\theta) \,, \quad \mathbf{N}_2(\theta) = p \left[ \begin{array}{c} Si}(\theta) - {\rm Si}(\theta_\star) \\ {\rm Ci}(\theta) - Ci}(\theta_\star) \end{array} \right] \,.$$ can apply here the same of we used Sec. where inferred that for growing case: $\mathbf{P}^{\rm CLASS} \sim h^\prime/2$, see Eq. . Thus, the second term on the rhs of adds a function to general in words, we have that $\mathbf{P}(N,\theta) \simeq \mathbf{P}^{CLASS}(N) (1 + the same level of accuracy attained for the variables. [^1]: means that the evolution of the field $\phi $ starts on the left branch the parabolic potential , which is not a limitation as the system of equations is the change $\phi \to To start the on right the of $y,y_1$ need to be changed. [^2]: Strictly speaking, there is no need consider $y_1$ as a separate variable, nor to take Eq. independent of motion. As in Eqs. , $y_1$ just to $H^{-1}$, and then and enough | end{array}
\right] \,,$$ where $\mathbf{P}^{\Rm CLASS}(N)$ is The soLutIon In eq. . EvEn thOugh the second tERm on The rhs of Eq. still is a rapiDly osCiLLatiNG fUnctiOn, it canNOt BE DisCaRdEd aS sIMpLy as oTheRs due to The presencE of $H^\pRime$.
The last iNTeGral in Eq. caN be Integrated itEraTively By ParTS, and iN suCh casE it can BE proveD that if we ReTAin the DOminant TERm In thE integrations we obTAiN[^12]: $$\Label{eq:79}
\mathbf{P}(n,\theta) \SiMEq \MAThbF{P}^{\rM CLASS}(N) + \fraC{
h^\Prime}{2} \MAthbf{N}_2(\tHEtA) \,, \QUAd \mAThbf{N}_2(\theta)
= p \leFt[
\begin{arraY}{C}
{\rm si}(\thetA) - {\rM Si}(\THeta_\stAr) \\
{\rm CI}(\tHEta) - {\Rm Ci}(\theta_\stAr)
\enD{array}
\rigHt] \,.$$ We caN Apply heRE the samE kind oF trIck We usED iN SEc. \[sEc:GEneR-SoLut-PErt\], Where we iNfErRed thAt foR THE GrowIng Case: $\MathbF{P}^{\rm CLASS} \sim h^\PriMe/2$, seE eq. . THus, thE secoNd teRm On the Rhs of EQ. adds A nEgligible oscillAtinG function To tHe GenErAl solUTion; in OthEr wOrds, we hAve just FOunD tHAT $\MaThbf{P}(N,\theta) \simeq \maThBF{p}^{ClASS}(N) (1 +
\matHcal{O}(10^{-2}))$, wHIcH iS The same lEvEl oF accURAcy atTainED fOr the bacKgrounD VaRiAbles.
[^1]: ThIs Means tHaT thE evOlutiON of tHe scalAr field $\pHi $ staRTs on the left braNCh of the paraboLIc POTeNTial , WhiCh is not a limItatIOn as The sYStEm oF EquatIons iS iNVaRIant under the change $\pHi \To -\phi$. TO starT the evolution On the right BRANch, the siGns oF $Y,y_1$ JUst need to be chaNged.
[^2]: STrictly speAKing, therE is no Need to coNsider $y_1$ as A SEparate vAriAblE, noR to TAKe eq. as an indepenDENt eqUaTion of mOtiOn. As shoWn iN EqS. , $y_1$ iS juSt ProportioNal to $H^{-1}$, anD tHeN oNlY EqS. and aRE enough tO rEprEsEnt | end{array}
\right] \,,$$ w here $\mat hbf{P }^{ \rm C LASS }(N) $ is the solut i on i n Eq. . Even though th e sec on d ter m o n the rhs of Eq . st il lisar ap idlyosc illatin g function , i tcannot be di s ca rded as si mpl y as othersdue to th epre s enceof$h^\p rime$.
The l ast integ ra l in Eq . can b e in tegr ated iterativelyb yp arts, and in s uch ca se it c anbeproved tha tif we retaint he d o min a nt term in th e integrati o nswe obt ai n[^ 1 2]: $$ \labe l{ e q:7 9}
\mathb f{P} (N,\theta ) \sim e q \math b f{P}^{\ rm CLA SS} (N) + \ f ra c{
h ^\p r im e}{ 2 } \ mathbf{N }_ 2( \thet a) \ , , \ quad \m athb f{N}_ 2(\theta)
= p\lef t [ \begi n{arr ay}{ c}
{\rm Si}(\ theta )- {\rm Si}(\the ta_\ star) \\{\r mCi} (\ theta ) - {\r m C i}( \theta_ \star)\ end {a r r a y}
\right] \,.$$ Weca n ap ply here the s a me k i nd of tr ic k w e us e d in S ec.\ [s ec:gener -solut - pe rt \], whe re we in fe rre d t hat f o r th e grow ing case : $\m a thbf{P}^{\rm C L ASS} \sim h^\ p ri m e /2 $ , se e E q. . Thus,thes econ d te r mont he rh s ofEq . a dds a negligible os ci llatin g fun ction to thegeneral so l u t ion; inothe r w o rds, we have j ust f ound that$ \mathbf{ P}(N, \theta)\simeq \m a t hbf{P}^{ CLA SS} (N) (1 +
\ mathcal{O}(10 ^ { -2}) )$ , which is the sa melev elofac curacy at tained f or t he b ack groun d variabl es .
[^ 1]: This meansthattheev ol u tio n of th e s c a larfi el d $\ phi $ star ts o n th e leftbranch of th e par ab ol ic pote ntial , which i s not a li mi tat ion as t he syste m of equations is invar i ant und erthe c hang e $\phi \ to-\phi$ . T o start the e volut io n o n the r i g ht br an ch, the si g n s o f $y, y_ 1$ j ust nee d to be changed.
[ ^2] : Strictly sp eak ing, t he rei sn o n ee d to c onsider $y_1$ a s a separa te va riable, no r to t ake Eq. as an inde p endentequationof motion .As s h o wnin Eqs. ,$y_1$ is just pro p ortio n al to $ H^{ -1}$,an d t hen o nly Eq s . and are en ou gh torepre se nt | end{array}
\right] \,,$$_where $\mathbf{P}^{\rm_CLASS}(N)$ is the solution_in Eq. ._Even_though the_second_term on the_rhs of Eq. _still is a rapidly_oscillating function, it_cannot_be discarded as simply as others due to the presence of $h^\prime$.
The last integral_in_Eq. can_be_integrated_iteratively by parts, and in_such case it can be_proved that_if we retain the dominant term in the_integrations_we obtain[^12]: $$\label{eq:79}
_ \mathbf{P}(N,\theta) \simeq \mathbf{P}^{\rm CLASS}(N) + \frac{
_ h^\prime}{2} \mathbf{N}_2(\theta) \,, \quad \mathbf{N}_2(\theta)
_ = p_\left[_
\begin{array}{c}
{\rm_Si}(\theta) - {\rm Si}(\theta_\star)_\\
{\rm Ci}(\theta) - {\rm Ci}(\theta_\star)
\end{array}
\right] \,.$$_We can apply here the same_kind of trick we used in Sec. \[sec:gener-solut-pert\],_where we inferred that for the_growing case: $\mathbf{P}^{\rm CLASS} \sim_h^\prime/2$, see_Eq. . Thus, the second term_on the rhs_of Eq. _adds a negligible_oscillating function to the general solution;_in other words,_we have just found that $\mathbf{P}(N,\theta)_\simeq_\mathbf{P}^{CLASS}(N) (1 +
\mathcal{O}(10^{-2}))$,_which_is_the same_level of accuracy_attained_for the_background_variables.
[^1]: This means that the evolution_of_the scalar field $\phi $ starts on_the left branch of_the_parabolic potential , which is_not a limitation as the_system of equations is invariant under_the change_$\phi \to_-\phi$. To start the evolution on the right branch, the signs_of $y,y_1$ just need to be_changed.
[^2]: Strictly speaking, there_is no_need_to consider $y_1$_as_a separate_variable, nor to take Eq. as an_independent equation_of motion. As shown in Eqs. ,_$y_1$ is just proportional_to_$H^{-1}$, and then only Eqs. and _are enough to represent |
5,-2.5) to (0.5,1.5);
(3.5,1.5) to (3.5,0.5); (8.5,1.5) to (8.5,0.5); (11.5,1.5) to (11.5,0.5); (5.5,0.5) to (5.5,-0.5); (1.5,-3.5) to (1.5,-4.5); (7.5,-0.5) to (7.5,-1.5);
(6.5,1.5) to (10.5,1.5); (10.5,1.5) to (10.5,0.5); (10.5,0.5) to (7.5,0.5); (7.5,0.5) to (7.5,-0.5); (7.5,-0.5) to (5.5,-0.5); (5.5,-0.5) to (5.5,-1.5); (5.5,-1.5) to (5.5-1,-1.5); (5.5-1,-1.5) to (5.5-1,-1.5-1); (5.5-1,-1.5-1) to (5.5-1-2,-1.5-1); (5.5-1-2,-1.5-1) to (5.5-1-2,-1.5-1+1); (5.5-1-2,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1); (5.5-1-2+1,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1+1) to (5.5-1-2+1+1,-1.5-1+1+1) to (5.5-1-2+1+1,-1.5-1+1+1+1);
(0.5,-2.5) to (0.5,-2.5-1) to (0.5 | 5,-2.5) to (0.5,1.5);
(3.5,1.5) to (3.5,0.5); (8.5,1.5) to (8.5,0.5); (11.5,1.5) to (11.5,0.5); (5.5,0.5) to (5.5,-0.5); (1.5,-3.5) to (1.5,-4.5); (7.5,-0.5) to (7.5,-1.5);
(6.5,1.5) to (10.5,1.5); (10.5,1.5) to (10.5,0.5); (10.5,0.5) to (7.5,0.5); (7.5,0.5) to (7.5,-0.5); (7.5,-0.5) to (5.5,-0.5); (5.5,-0.5) to (5.5,-1.5); (5.5,-1.5) to (5.5 - 1,-1.5); (5.5 - 1,-1.5) to (5.5 - 1,-1.5 - 1); (5.5 - 1,-1.5 - 1) to (5.5 - 1 - 2,-1.5 - 1); (5.5 - 1 - 2,-1.5 - 1) to (5.5 - 1 - 2,-1.5 - 1 + 1); (5.5 - 1 - 2,-1.5 - 1 + 1) to (5.5 - 1 - 2 + 1,-1.5 - 1 + 1); (5.5 - 1 - 2 + 1,-1.5 - 1 + 1) to (5.5 - 1 - 2 + 1,-1.5 - 1 + 1 + 1) to (5.5 - 1 - 2 + 1 + 1,-1.5 - 1 + 1 + 1) to (5.5 - 1 - 2 + 1 + 1,-1.5 - 1 + 1 + 1 + 1);
(0.5,-2.5) to (0.5,-2.5 - 1) to (0.5 | 5,-2.5) to (0.5,1.5);
(3.5,1.5) to (3.5,0.5); (8.5,1.5) to (8.5,0.5); (11.5,1.5) to (11.5,0.5); (5.5,0.5) to (5.5,-0.5); (1.5,-3.5) to (1.5,-4.5); (7.5,-0.5) to (7.5,-1.5);
(6.5,1.5) to (10.5,1.5); (10.5,1.5) to (10.5,0.5); (10.5,0.5) to (7.5,0.5); (7.5,0.5) vo (7.5,-0.5); (7.5,-0.5) to (5.5,-0.5); (5.5,-0.5) to (5.5,-1.5); (5.5,-1.5) go (5.5-1,-1.5); (5.5-1,-1.5) to (5.5-1,-1.5-1); (5.5-1,-1.5-1) to (5.5-1-2,-1.5-1); (5.5-1-2,-1.5-1) to (5.5-1-2,-1.5-1+1); (5.5-1-2,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1); (5.5-1-2+1,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1+1) ti (5.5-1-2+1+1,-1.5-1+1+1) to (5.5-1-2+1+1,-1.5-1+1+1+1);
(0.5,-2.5) to (0.5,-2.5-1) to (0.5 | 5,-2.5) to (0.5,1.5); (3.5,1.5) to (3.5,0.5); (8.5,1.5) (11.5,1.5) (11.5,0.5); (5.5,0.5) (5.5,-0.5); (1.5,-3.5) to to (10.5,1.5) to (10.5,0.5); to (7.5,0.5); (7.5,0.5) (7.5,-0.5); (7.5,-0.5) to (5.5,-0.5); (5.5,-0.5) to (5.5,-1.5) to (5.5-1,-1.5); (5.5-1,-1.5) to (5.5-1,-1.5-1); (5.5-1,-1.5-1) to (5.5-1-2,-1.5-1); (5.5-1-2,-1.5-1) to (5.5-1-2,-1.5-1+1); (5.5-1-2,-1.5-1+1) (5.5-1-2+1,-1.5-1+1); (5.5-1-2+1,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1+1) to (5.5-1-2+1+1,-1.5-1+1+1) to (5.5-1-2+1+1,-1.5-1+1+1+1); (0.5,-2.5) to (0.5,-2.5-1) to (0.5 | 5,-2.5) to (0.5,1.5);
(3.5,1.5) to (3.5,0.5); (8.5,1.5) to (8.5,0.5); (11.5,1.5) to (11.5,0.5); (5.5,0.5) to (5.5,-0.5); (1.5,-3.5) to (1.5,-4.5); (7.5,-0.5) to (7.5,-1.5);
(6.5,1.5) to (10.5,1.5); (10.5,1.5) to (10.5,0.5); (10.5,0.5) to (7.5,0.5); (7.5,0.5) to (7.5,-0.5); (7.5,-0.5) to (5.5,-0.5); (5.5,-0.5) tO (5.5,-1.5); (5.5,-1.5) to (5.5-1,-1.5); (5.5-1,-1.5) to (5.5-1,-1.5-1); (5.5-1,-1.5-1) to (5.5-1-2,-1.5-1); (5.5-1-2,-1.5-1) to (5.5-1-2,-1.5-1+1); (5.5-1-2,-1.5-1+1) tO (5.5-1-2+1,-1.5-1+1); (5.5-1-2+1,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1+1) to (5.5-1-2+1+1,-1.5-1+1+1) To (5.5-1-2+1+1,-1.5-1+1+1+1);
(0.5,-2.5) tO (0.5,-2.5-1) to (0.5 | 5,-2.5) to (0.5,1.5);
(3. 5,1.5) to(3.5, 0.5 );(8 .5,1 .5)to (8.5,0.5);( 11.5 ,1.5) to (11.5,0.5); ( 5.5,0 .5 ) to( 5. 5,-0. 5); (1. 5 ,- 3 . 5)to ( 1.5 ,- 4 .5 ); (7 .5, -0.5) t o (7.5,-1. 5);
(6.5,1.5) to (1 0.5,1.5);(10 .5,1.5) to ( 10. 5,0.5) ;(10 . 5,0.5 ) t o (7. 5,0.5) ; (7.5, 0.5) to ( 7. 5 ,-0.5) ; (7.5,- 0 . 5) to(5.5,-0.5); (5.5, - 0. 5 ) to (5.5,-1.5 ); (5. 5, - 1. 5 ) to (5 .5-1,-1.5) ;(5.5- 1 ,-1.5)t o( 5 . 5-1 , -1.5-1); (5.5 -1,-1.5-1)t o ( 5.5-1- 2, -1. 5 -1); ( 5.5-1 -2 , -1. 5-1) to (5. 5-1- 2,-1.5-1+ 1); (5 . 5-1-2,- 1 .5-1+1) to (5 .5- 1-2 +1,- 1 .5 -1 +1) ;( 5.5 - 1- 2+1 , -1. 5-1+1) t o(5 .5-1- 2+1, - 1 . 5 -1+1 +1) to(5.5- 1-2+1+1,-1.5- 1+1 +1)t o ( 5.5-1 -2+1+ 1,-1 .5 -1+1+ 1+1);
(0.5 ,- 2.5) to (0.5,-2 .5-1 ) to (0.5 | 5,-2.5) to_(0.5,1.5);
(3.5,1.5) to_(3.5,0.5); (8.5,1.5) to (8.5,0.5);_(11.5,1.5) to_(11.5,0.5);_(5.5,0.5) to_(5.5,-0.5);_(1.5,-3.5) to (1.5,-4.5);_(7.5,-0.5) to (7.5,-1.5);
(6.5,1.5)_to (10.5,1.5); (10.5,1.5) to_(10.5,0.5); (10.5,0.5) to_(7.5,0.5);_(7.5,0.5) to (7.5,-0.5); (7.5,-0.5) to (5.5,-0.5); (5.5,-0.5) to (5.5,-1.5); (5.5,-1.5) to (5.5-1,-1.5); (5.5-1,-1.5) to_(5.5-1,-1.5-1);_(5.5-1,-1.5-1) to_(5.5-1-2,-1.5-1);_(5.5-1-2,-1.5-1)_to (5.5-1-2,-1.5-1+1); (5.5-1-2,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1);_(5.5-1-2+1,-1.5-1+1) to (5.5-1-2+1,-1.5-1+1+1) to (5.5-1-2+1+1,-1.5-1+1+1)_to (5.5-1-2+1+1,-1.5-1+1+1+1);
(0.5,-2.5)_to (0.5,-2.5-1) to (0.5 |
immersive experience in VR. We prove DROP is NP-hard and design a fully polynomial-time approximation scheme, *Dual Entangled World Navigation* (DEWN), by finding Minimum Immersion Loss Range (*MIL Range*). Afterward, we show that the existing spatial query algorithms and index structures can leverage DEWN as a building block to support $k$NN and range queries in the dual worlds of VR. Experimental results and a user study with implementation in HTC VIVE manifest that DEWN outperforms the baselines with smoother RW operations in various VR scenarios.'
author:
- 'Shao-Heng Ko^1^, Ying-Chun Lin^2^, Hsu-Chao Lai^13^, Wang-Chien Lee^4^, De-Nian Yang^15^'
bibliography:
-'ref.bib'
title: On VR Spatial Query for Dual Entangled Worlds
---
Introduction {#sec:intro}
============
With the growing availability of Virtual Reality (VR) devices, innovative VR applications in virtual social, travel, and shopping domains have emerged. This technological trend of VR not only attracts business interests from prominent vendors such as Facebook and Alibaba[^1] but also brings a new wave of research in the academia. While current research on VR mostly originated from Computer Graphics, Multimedia, and HCI, focusing on constructing vivid VR worlds [@MP16; @AB16; @SK16], the needs for research and support from the database community are also imminent.
Traditional research on spatial data management has contributed significantly to various applications in the physical world. For example, for mobile users on a journey, the information about the closest gas stations along a routing path can be found by spatial queries [@YL14]. These queries are also needed in the virtual worlds in VR applications where moving between point-of-interests (POIs) is a basic operation. For example, in VR campus touring[^2] and VR architecture/indoor navigation[^3] applications, spatial queries can be issued to find POIs and guide users to move to them. However, in many VR applications where users move in both the virtual and physical worlds, the simple one-world setting may no longer sustain, rendering the aforementioned queries useless. To study this problem, we revisit a number of spatial queries widely used in many VR applications to develop new algorithms by considering factors in the dual entangled virtual and physical worlds.
Traditional VR | immersive experience in VR. We prove DROP is NP - hard and design a amply polynomial - meter approximation scheme, * Dual Entangled World Navigation * (DEWN), by finding Minimum Immersion Loss Range (* MIL Range *). subsequently, we show that the existing spatial query algorithm and exponent structures can leverage DEWN as a building block to support $ k$NN and range question in the dual worlds of VR. Experimental results and a user sketch with implementation in HTC VIVE attest that DEWN outperforms the baseline with smooth RW operations in various VR scenarios.'
generator:
-' Shao - Heng Ko^1^, Ying - Chun Lin^2^, Hsu - Chao Lai^13^, Wang - Chien Lee^4^, De - Nian Yang^15^'
bibliography:
-'ref.bib'
title: On VR Spatial Query for Dual Entangled Worlds
---
Introduction { # sec: intro }
= = = = = = = = = = = =
With the growing availability of Virtual Reality (VR) device, innovative VR applications in virtual social, travel, and shopping domains have emerged. This technical trend of VR not only attracts business interests from prominent seller such as Facebook and Alibaba[^1 ] but besides lend a new wave of research in the academia. While current research on VR mostly originated from Computer Graphics, Multimedia, and HCI, focusing on constructing vivid VR universe [ @MP16; @AB16; @SK16 ], the needs for research and accompaniment from the database residential district are besides imminent.
Traditional research on spatial data management has contributed significantly to various applications in the physical populace. For example, for mobile users on a journey, the information about the closest gas station along a routing path can be found by spatial queries [ @YL14 ]. These queries are also necessitate in the virtual worlds in VR applications where moving between point - of - interests (poi) is a basic operation. For example, in VR campus touring[^2 ] and VR architecture / indoor navigation[^3 ] applications, spatial queries can be issued to find POIs and guide exploiter to move to them. However, in many VR applications where user move in both the virtual and forcible worlds, the simple one - world setting may no longer nourish, rendering the aforementioned question useless. To study this problem, we revisit a number of spatial queries widely used in many VR applications to develop new algorithms by considering factors in the double entangled virtual and physical world.
Traditional VR | imlersive experience in VR. We prove DROP nw NP-hacd and sesign a fully polynomial-time approxmmatuon sxheme, *Dual Entangled Wurld Navihation* (DWWN), uy finding Minimnj Immersion Lkds Rcnje (*MIL Range*). Afjerward, we svow that the efirtnng spatial query algorithms and indqx struvtkres can leverwge CQWN zs a building block to support $k$NN and raige queries in yhe dual worlds of VR. Expegimejtal results and a user study witr implementatkon in HTC VIVE manifeat that DEWN outperforms the barelinzs with smoityer TW operationw in narious VR sccmarios.'
duthor:
- 'Xhao-Heng Ko^1^, Yikg-Chui Lib^2^, Hsu-Chao Lai^13^, Wang-Chixn Lee^4^, De-Nian Yang^15^'
bifliographf:
-'rzf.bib'
title: On VR Spatuao Quety fos Duxo Evtahgkes Worlfs
---
Iitroduction {#sec:intro}
============
Wirh the growing avaikafpkity of Virthal Rewlyty (VR) devices, innovative VR applicatiots jn virtual social, traveo, and shopping domaind have emqrged. This technological trend of VR not only attsacts cusnkcss kbtfrests from prominent vendors such as Facebooh amd Alibaba[^1] but ajso brings s jee wave of resexrch iu tge academia. While furrent reseqrch on VW moxtly originated from Computwr Graphics, Iyltimedia, and HCI, yocusing on eonstrocting vivid VR worlds [@MP16; @AB16; @RK16], tge needs fog researcg and support froo tme gatabase community are alsj imminenv.
Tradntional fesesrch og spatial fata management has contgibutgd sigtificantly to various applications in the physical worlc. Xor example, for kobile users jn a journey, tke inforiatiov about thv closest gas statiogs along a rogjing path can be founq by sparial qudfies [@YL14]. These aueries age alsi needed in the vivtual sorlds in VR ap'uucations where kovkng bvtwxen pjhnt-of-interesds (PUIs) os a casic okzrabiov. Fot example, in VR campgs tkuring[^2] and VR archotccture/indior navidation[^3] applicstions, spatial quegies ran be issurd jo find POIs and guide users to move to hhei. However, in mann VR applicatipns where users move in both the virtuao and physical wotlds, the simple one-wotld setting mab no ljnger susdain, rendering the adorementioned quevies useless. To study fhis psoblel, we revisit a number of spatial queries widely used in many VR applicatiins to develop new algprithks yy consydermnj factors in the cual entangled virtual and physmcal worldv.
Txaditional VR | immersive experience in VR. We prove DROP and a fully approximation scheme, *Dual finding Immersion Loss Range Range*). Afterward, we that the existing spatial query algorithms index structures can leverage DEWN as a building block to support $k$NN and queries in the dual worlds of VR. Experimental results and a user study implementation HTC manifest DEWN outperforms the baselines with smoother RW operations in various VR scenarios.' author: - 'Shao-Heng Ko^1^, Lin^2^, Hsu-Chao Lai^13^, Wang-Chien Lee^4^, De-Nian Yang^15^' bibliography: title: On VR Spatial for Dual Entangled Worlds --- {#sec:intro} With the availability Virtual (VR) devices, innovative applications in virtual social, travel, and shopping domains have emerged. This technological trend of VR not only business interests vendors such Facebook Alibaba[^1] also brings a of research in the academia. While VR mostly originated from Computer Graphics, Multimedia, and focusing on vivid VR worlds [@MP16; @AB16; @SK16], needs for research and support from the database are also imminent. Traditional research on spatial data management has contributed significantly to various applications physical world. For example, mobile users on journey, information the gas stations a routing path can be found by spatial queries [@YL14]. These are also needed in the virtual worlds in VR applications between (POIs) is a operation. For example, in campus and VR architecture/indoor navigation[^3] queries be POIs guide to move to them. in many VR applications where move in both the simple one-world setting may no longer sustain, rendering aforementioned queries useless. To study this problem, revisit a number of spatial queries widely used in many VR applications develop new considering factors in the dual entangled virtual and worlds. Traditional VR | immersive experience in VR. We Prove DROP iS NP-haRd aNd dEsIgn a FullY polynomial-timE ApprOximation scheme, *Dual EntAngleD WORld NAViGatioN* (DEWN), by FInDINg MInImUm IMmERsIon LoSs RAnge (*MIL range*). AfterWarD, wE show that the EXiSting spatiAl qUery algorithMs aNd indeX sTruCTures Can LeverAge DEWn As a buiLding blocK tO SupporT $K$NN and rANGe QuerIes in the dual worldS Of vr. Experimental rEsults AnD A uSER stUdy With implemEnTatioN In HTC VIve mANIFesT That DEWN outpeRforms the baSEliNes witH sMooTHer RW oPeratIoNS in Various VR scEnarIos.'
author:
- 'shao-HeNG Ko^1^, Ying-cHun Lin^2^, HSu-Chao lai^13^, wanG-ChiEN LEe^4^, de-NIaN yanG^15^'
BiBliOGraPhy:
-'ref.biB'
tItLe: On Vr SpaTIAL querY foR DuaL EntaNgled Worlds
---
InTroDuctIOn {#sEc:intRo}
============
WitH the GrOwing AvailaBilitY oF Virtual Reality (vR) deVices, innoVatIvE VR ApPlicaTIons in VirTuaL social, Travel, aND shOpPING dOmains have emerged. THiS TEcHnologicAl trenD Of vR NOt only atTrActS busINEss inTereSTs From promInent vENdOrS such as faCebook AnD AlIbaBa[^1] but ALso bRings a New wave oF reseARch in the academIA. While current REsEARcH On VR MosTly originatEd frOM ComPuteR grAphICs, MulTimedIa, ANd hcI, focusing on construCtIng vivId VR wOrlds [@MP16; @AB16; @SK16], thE needs for rESEArch and sUppoRT fROm the database cOmmunIty are also IMminent.
TRaditIonal resEarch on spATIal data mAnaGemEnt Has CONtRibuted signifICAntlY tO variouS apPlicatiOns In tHe pHysIcAl world. FoR example, FoR mObIlE usErs on A Journey, tHe InfOrMatIon abOUt the cLosesT gas StAtIOns Along a rOUtING patH cAn Be foUnd By SpatiAl quERieS [@YL14]. ThesE queries aRe aLSo neEdEd In the viRtual worlds in vR ApplicatioNs WheRe moviNG Between pOint-of-interests (POIs) is a bASic operAtiOn. For ExamPle, in VR caMpuS touriNg[^2] aND VR arcHitectUre/inDoOr nAVIgatiON[^3] ApPliCaTions, spatiAL QueRies cAn Be isSued to fInd POIs and guide useRS to Move to them. HowEveR, in mANY Vr apPLiCAtiOnS WheRE Users move in both The virtual AnD PhYsical worlDS, thE sImple onE-world sEttinG May no loNger sustaIn, renderiNg The aFORemEntioned quEries useLess. To stuDY this PRoBlem, wE reVisit a NuMbeR of spAtial qUEriEs widEly useD iN many Vr applIcAtions to Develop new algorithms by cOnsideRing fActOrs in the dUal ENtaNgled virtUal aNd physical WorLds.
tradiTioNAl VR | immersive experience in V R. We prov e DRO P i s N P- hard and design a full y pol ynomial-time approxima tionsc h eme, *D ual E ntangle d W o r ldNa vi gat io n *(DEWN ),by find ing Minimu m I mm ersion LossR an ge (*MIL R ang e*). Afterwa rd, we sh ow th a t the ex istin g spat i al que ry algori th m s andi ndex st r u ct ures can leverage DEW N a s a building bl ock to s u pp o r t $ k$N N and rang equeri e s in th e d u a l wo r lds of VR. Ex perimentalr esu lts an da u s er stu dy wi th imp lementation inHTC VIVEmanife s t thatD EWN out perfor msthe bas e li ne s w it h sm o ot her RWoperatio ns i n var ious V R scen ari os.'
auth or:
- 'Shao-H eng Ko^ 1 ^,Ying- ChunLin^ 2^ , Hsu -ChaoLai^1 3^ , Wang-Chien Le e^4^ , De-Nian Ya ng ^15 ^'
bibl i ograph y:-'r ef.bib'
title: OnVR S p at ial Query for Dual E n t an gled Wor lds
-- -
In t roductio n{#s ec:i n t ro}
= ==== = == ====
Wi th the gr ow ing ava il abilit yofVir tualR eali ty (VR ) device s, in n ovative VR app l ications in v i rt u a ls ocia l,travel, and sho p ping dom a in s h a ve em erged .T hi s technological tren dof VRnot o nly attractsbusiness i n t e rests fr om p r om i nent vendors s uch a s Facebook and Alib aba[^ 1] but a lso bring s a new wa veofres ear c h i n the academi a . Whi le curren t r esearch on VR mo stl yoriginate d from C om pu te rGra phics , Multime di a,an d H CI, f o cusing on c onst ru ct i ngvivid V R w o r lds[@ MP 16;@AB 16 ; @SK 16], the needsfor resea rch andsu pp ort fro m the databas ecommunityar e a lso im m i nent.
T raditional research ons patialdat a man agem ent has c ont ribute d s i gnific antlyto va ri ous a pplic a t io nsin the physi c a l w orld. F or e xample, for mobile userso n a journey, the in form a t io n a b ou t th ec los e s t gas stationsalong a ro ut i ng path canb e f ou nd by s patialqueri e s [@YL1 4]. These queriesar e al s o ne eded in th e virtua l worldsi n VRa pp licat ion s wher emov ing b etween poi nt-of -inter es ts (PO Is) i sa basicoperation. For example, in VR camp ustouring[^ 2]a ndVR archit ectu re/indoornav iga tion[ ^3] appli cati o ns , s p atial que r ies can b e i ssu e d t o find POIs a n d g uideuse r s to m oveto them. However, in many VR app lica t i ons wh e re u se rs move in bot h t he v irtual a nd physical w orlds, t he simpl e one- worldsetting m ay no lon gersus tain, ren der in g the af or em e ntione d qu er ies us eless. To s t u dy this problem, we r e v isita nu mberof spatia l que ries widel y used in m any VR app licat ions to d evelop ne walgorithms by consid ering factor sin t hedual e ntan g l ed vi rtua land physical w or l ds .Tra diti onalVR | immersive_experience in_VR. We prove DROP_is NP-hard_and_design a_fully_polynomial-time approximation scheme,_*Dual Entangled World_Navigation* (DEWN), by finding_Minimum Immersion Loss_Range_(*MIL Range*). Afterward, we show that the existing spatial query algorithms and index structures_can_leverage DEWN_as_a_building block to support $k$NN_and range queries in the_dual worlds_of VR. Experimental results and a user study_with_implementation in HTC_VIVE manifest that DEWN outperforms the baselines with smoother_RW operations in various VR scenarios.'
author:
-_'Shao-Heng Ko^1^, Ying-Chun_Lin^2^,_Hsu-Chao_Lai^13^, Wang-Chien Lee^4^, De-Nian_Yang^15^'
bibliography:
-'ref.bib'
title: On VR Spatial Query for_Dual Entangled Worlds
---
Introduction {#sec:intro}
============
With the growing_availability of Virtual Reality (VR) devices, innovative_VR applications in virtual social, travel,_and shopping domains have emerged._This technological_trend of VR not only_attracts business interests_from prominent_vendors such as_Facebook and Alibaba[^1] but also brings_a new wave_of research in the academia. While_current_research on VR_mostly_originated_from Computer_Graphics, Multimedia, and_HCI,_focusing on_constructing_vivid VR worlds [@MP16; @AB16; @SK16],_the_needs for research and support from the_database community are also_imminent.
Traditional_research on spatial data_management has contributed significantly to_various applications in the physical world._For example,_for mobile_users on a journey, the information about the closest gas stations_along a routing path can be_found by spatial queries_[@YL14]. These_queries_are also needed_in_the virtual_worlds in VR applications where moving between_point-of-interests (POIs)_is a basic operation. For example,_in VR campus touring[^2]_and_VR architecture/indoor navigation[^3] applications, spatial queries_can be issued to find POIs_and guide users to move_to_them._However, in many VR applications_where users move in both the_virtual and physical_worlds, the simple one-world setting may no_longer_sustain, rendering the aforementioned queries useless._To_study this problem, we revisit a_number_of_spatial queries widely used in_many VR applications to develop new_algorithms by considering factors in the dual entangled virtual_and physical worlds.
Traditional_VR |
mathbb{R}/\mathcal{D})\simeq Hom(\pi_1(G)\to\mathbb{R}/\mathcal{D}).$$ Now assume $G=SL(2,\mathbb{R})$ and $\mathcal{D}=\mathbb{Z}$ so that $\mathbb{R}/\mathcal{D}\simeq U(1)$. The use of Iwazawa decomposition yields $SL(2,\mathbb{R})\simeq \mathbb{R}^2\times \mathbb{S}^1$, which combined with $\pi_1(X\times Y)=\pi_1(X)\times\pi_1(Y)$ for any topological spaces $X$ and $Y$ implies $\pi_1(SL(2,\mathbb{R}))\simeq\mathbb{Z}$. Hence, $Hom(\mathbb{Z}\to U(1))\simeq U(1)$ which classifies the inequivalent extensions of $SL(2,\mathbb{R})$ by $U(1)$. Accordingly, one expect that the extension of the present construction of $SL(2,\mathbb{R})$ gives rise to inequivalent classes of star-products.\
Quantized plane waves for $\mathbb{R}^3_\theta$
===============================================
The determination of which takes the generic form $$E_p(\hat{x})=\omega(p)e^{i\xi(p)\hat{x}} \label{generalform-ncexpo},$$ can be achieved through a standard albeit tedious computation whose main steps are summarized now.\
First, using $e^{i\xi(p)\hat{x}}\rhd 1 = \frac{e^{ipx}}{\omega(p)}$, one easily infers that $$e^{-i\xi(\lambda p) \hat{x}} \partial_\mu e^{i\xi(\lambda p) \hat{x}} = (i\lambda p_\mu) {{ \mathbb{I}}},\label{megaplus1}$$ which holds for any $\lambda\in\mathbb{R}$. Then, the combination of $[\partial_\mu, \hat{x}_\nu] = [\partial_\mu,x^a\varphi_{a\nu}] = \varphi_{\mu \nu}$ with the functional derivative of with respect to $\lambda$ yields $$\label{diff-xi}
\varphi_{\mu\nu}(i\lambda p) \frac{d}{d\lambda} \left[ \xi^\nu(\lambda p | mathbb{R}/\mathcal{D})\simeq Hom(\pi_1(G)\to\mathbb{R}/\mathcal{D}).$$ Now assume $ G = SL(2,\mathbb{R})$ and $ \mathcal{D}=\mathbb{Z}$ so that $ \mathbb{R}/\mathcal{D}\simeq U(1)$. The use of Iwazawa decay give $ SL(2,\mathbb{R})\simeq \mathbb{R}^2\times \mathbb{S}^1 $, which combined with $ \pi_1(X\times Y)=\pi_1(X)\times\pi_1(Y)$ for any topological spaces $ X$ and $ Y$ entail $ \pi_1(SL(2,\mathbb{R}))\simeq\mathbb{Z}$. therefore, $ Hom(\mathbb{Z}\to U(1))\simeq U(1)$ which classifies the inequivalent extensions of $ SL(2,\mathbb{R})$ by $ U(1)$. Accordingly, one ask that the extension of the present construction of $ SL(2,\mathbb{R})$ contribute rise to inequivalent classes of star topology - products.\
Quantized plane wave for $ \mathbb{R}^3_\theta$
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
The determination of which takes the generic form $ $ E_p(\hat{x})=\omega(p)e^{i\xi(p)\hat{x } } \label{generalform - ncexpo},$$ can be achieved through a standard albeit tedious calculation whose main steps are summarized now.\
First, use $ e^{i\xi(p)\hat{x}}\rhd 1 = \frac{e^{ipx}}{\omega(p)}$, one easily infers that $ $ e^{-i\xi(\lambda p) \hat{x } } \partial_\mu e^{i\xi(\lambda p) \hat{x } } = (i\lambda p_\mu) { { \mathbb{I}}},\label{megaplus1}$$ which holds for any $ \lambda\in\mathbb{R}$. Then, the combination of $ [ \partial_\mu, \hat{x}_\nu ] = [ \partial_\mu, x^a\varphi_{a\nu } ] = \varphi_{\mu \nu}$ with the functional derived function of with respect to $ \lambda$ yields $ $ \label{diff - xi }
\varphi_{\mu\nu}(i\lambda p) \frac{d}{d\lambda } \left [ \xi^\nu(\lambda p | matjbb{R}/\mathcal{D})\simeq Hom(\pi_1(G)\uo\mathbb{R}/\mathcal{D}).$$ Now asvume $G=AL(2,\mathbb{F})$ and $\mathcal{D}=\mathbb{Z}$ so thav $\marhbb{R}/\nathcal{D}\simeq U(1)$. The usd of Iwazwwa deconposmtion yields $SL(2,\mefhbb{R})\simeq \mafmbb{R}^2\tnmxs \mathbb{S}^1$, whicm combined fith $\pi_1(X\times F)=\pk_1(X)\cimes\pi_1(Y)$ for any topological spaces $V$ and $Y$ ilplies $\pi_1(SL(2,\matrbb{R}))\xymeq\jathbb{Z}$. Hence, $Hom(\mathbb{Z}\to U(1))\simeq H(1)$ which classifies tne inequivalent extensions of $DL(2,\mathbb{R})$ by $U(1)$. Acclrdingly, ong ex[wct that the extension of the presgnt construction of $SL(2,\mathbb{R})$ gixes rnse to ineqoncalftt classes if stwr-products.\
Quantized [lane wsves for $\mathbn{R}^3_\theva$
===============================================
Thw determination of whmch takes the generis form $$E_p(\vac{x})=\omega(p)e^{i\xi(p)\hat{x}} \labwl{tenerdlfosm-ncdzpo},$$ cah ue zchievfd vhrough a sfandard albwit tedious computauiog whose main sfeps awe summarized now.\
First, using $e^{i\xi(p)\hat{x}}\rhd 1 = \rrac{e^{ipx}}{\omega(p)}$, one easioy infers that $$e^{-i\xi(\lalbda p) \hae{x}} \partial_\mu e^{i\xi(\lambda p) \hat{x}} = (i\lambda p_\mu) {{ \mathtb{I}}},\laueu{meyqplus1}$$ qhlch holds for any $\lambda\in\mathbb{R}$. Then, the coigimanion of $[\partial_\mu, \hat{x}_\nu] = [\pattlak_\iu,x^a\varphi_{a\nu}] = \varpkj_{\mh \nu}$ with the funchional qerivqtive of rith respect to $\lambda$ yields $$\lqbel{diff-xi}
\vagphi_{\nu\nu}(i\lambda p) \frac{b}{d\lambda} \lefc[ \xi^\nu(\kambds p | mathbb{R}/\mathcal{D})\simeq Hom(\pi_1(G)\to\mathbb{R}/\mathcal{D}).$$ Now assume $G=SL(2,\mathbb{R})$ and $\mathcal{D}=\mathbb{Z}$ $\mathbb{R}/\mathcal{D}\simeq The use Iwazawa decomposition yields with Y)=\pi_1(X)\times\pi_1(Y)$ for any spaces $X$ and implies $\pi_1(SL(2,\mathbb{R}))\simeq\mathbb{Z}$. Hence, $Hom(\mathbb{Z}\to U(1))\simeq U(1)$ classifies the inequivalent extensions of $SL(2,\mathbb{R})$ by $U(1)$. Accordingly, one expect that the of the present construction of $SL(2,\mathbb{R})$ gives rise to inequivalent classes of star-products.\ plane for =============================================== determination of which takes the generic form $$E_p(\hat{x})=\omega(p)e^{i\xi(p)\hat{x}} \label{generalform-ncexpo},$$ can be achieved through a standard albeit computation whose main steps are summarized now.\ First, $e^{i\xi(p)\hat{x}}\rhd 1 = \frac{e^{ipx}}{\omega(p)}$, easily infers that $$e^{-i\xi(\lambda p) \partial_\mu p) \hat{x}} (i\lambda {{ which holds for $\lambda\in\mathbb{R}$. Then, the combination of $[\partial_\mu, \hat{x}_\nu] = [\partial_\mu,x^a\varphi_{a\nu}] = \varphi_{\mu \nu}$ with the functional derivative of respect to $$\label{diff-xi} \varphi_{\mu\nu}(i\lambda \frac{d}{d\lambda} \xi^\nu(\lambda | mathbb{R}/\mathcal{D})\simeq Hom(\pi_1(g)\to\mathbb{R}/\MathcAl{D}).$$ now AsSume $g=SL(2,\mAthbb{R})$ and $\mathcAL{D}=\maThbb{Z}$ so that $\mathbb{R}/\mathCal{D}\sImEQ U(1)$. ThE UsE of IwAzawa deCOmPOSitIoN yIelDs $sl(2,\mAthbb{r})\siMeq \mathBb{R}^2\times \maThbB{S}^1$, Which combineD WiTh $\pi_1(X\times y)=\pi_1(x)\times\pi_1(Y)$ for Any TopoloGiCal SPaces $x$ anD $Y$ impLies $\pi_1(sl(2,\mathbB{R}))\simeq\maThBB{Z}$. HencE, $hom(\mathBB{z}\tO U(1))\siMeq U(1)$ which classifiES tHE inequivalent eXtensiOnS Of $sl(2,\MatHbb{r})$ by $U(1)$. AccordInGly, onE Expect tHAt THE ExtENsion of the preSent construCTioN of $SL(2,\mAtHbb{r})$ Gives rIse to InEQuiValent classEs of Star-produCts.\
QuaNTized plANe waves For $\matHbb{r}^3_\thEta$
===============================================
THE dEtErmInATioN Of WhiCH taKes the geNeRiC form $$e_p(\haT{X})=\OMEga(p)E^{i\xI(p)\haT{x}} \labEl{generalform-NceXpo},$$ cAN be AchieVed thRougH a StandArd albEit teDiOus computation wHose Main steps Are SuMmaRiZed noW.\
first, uSinG $e^{i\Xi(p)\hat{x}}\Rhd 1 = \frac{E^{Ipx}}{\OmEGA(P)}$, oNe easily infers that $$E^{-i\XI(\LaMbda p) \hat{X}} \partiAL_\mU e^{I\Xi(\lambda P) \hAt{x}} = (I\lamBDA p_\mu) {{ \mAthbB{i}}},\lAbel{megaPlus1}$$ whICh HoLds for aNy $\Lambda\In\MatHbb{r}$. Then, THe coMbinatIon of $[\parTial_\mU, \Hat{x}_\nu] = [\partial_\mU,X^a\varphi_{a\nu}] = \vaRPhI_{\MU \nU}$ With The Functional dErivATive Of wiTH rEspECt to $\lAmbda$ YiELdS $$\Label{diff-xi}
\varphi_{\mu\Nu}(I\lambdA p) \fraC{d}{d\lambda} \left[ \Xi^\nu(\lambda P | mathbb{R}/\mathcal{D})\sim eq Hom(\pi _1(G) \to \ma th bb{R }/\m athcal{D}).$$N ow a ssume $G=SL(2,\mathbb{ R})$an d $\m a th cal{D }=\math b b{ Z } $ s oth at$\ m at hbb{R }/\ mathcal {D}\simeqU(1 )$ . The use of Iw azawa deco mpo sition yield s $ SL(2,\ ma thb b {R})\ sim eq \m athbb{ R }^2\ti mes \math bb { S}^1$, which c o m bi nedwith $\pi_1(X\tim e sY )=\pi_1(X)\tim es\pi_ 1( Y )$ f orany topologic al spac e s $X$ a n d$ Y $ im p lies $\pi_1(S L(2,\mathbb { R}) )\sime q\ mat h bb{Z}$ . Hen ce , $H om(\mathbb{ Z}\t o U(1))\s imeq U ( 1)$ whi c h class ifiesthe in equi v al en t e xt e nsi o ns of $SL (2,\math bb {R })$ b y $U ( 1 ) $ . Ac cor ding ly, o ne expect tha t t he e x ten sionof th e pr es ent c onstru ction o f $SL(2,\mathbb {R}) $ gives r ise t o i ne quiva l ent cl ass esof star -produc t s.\
Q u a nt ized plane waves f or $ \m athbb{R} ^3_\th e ta $= ======== == === ==== = = ===== ==== = == ======== ====== =
Th e deter mi nation o f w hic h tak e s th e gene ric form $$E_ p (\hat{x})=\ome g a(p)e^{i\xi(p ) \h a t {x } } \l abe l{generalfo rm-n c expo },$$ ca n b e achi evedth r ou g h a standard albeit t edious comp utation whose main step s a re summa rize d n o w.\
First, usi ng $e ^{i\xi(p)\ h at{x}}\r hd 1= \frac{ e^{ipx}}{ \ o mega(p)} $,one ea sil y in fers that $$e ^ { -i\x i( \lambda p) \hat{x }}\pa rti al_ \m u e^{i\xi (\lambda p )\h at {x} } = ( i \lambdap_ \mu ){{\math b b{I}}} ,\lab el{m eg ap l us1 }$$ whi c hh o ldsfo rany$\l am bda\i n\ma t hbb {R}$. T hen, thecom b inat io nof $[\p artial_\mu, \ ha t{x}_\nu]=[\p artial _ \ mu,x^a\v arphi_{a\nu}] = \varphi _ {\mu \n u}$ with the function alderiva tiv e of wi th res pectto $\ l a mbda$ y ie lds $ $\label{di f f -xi }
\va rp hi_{ \mu\nu} (i\lambda p) \frac { d}{ d\lambda} \le ft[ \xi ^ \ nu (\l a mb d a p | mathbb{R}/\mathcal{D})\simeq Hom(\pi_1(G)\to\mathbb{R}/\mathcal{D}).$$_Now assume_$G=SL(2,\mathbb{R})$ and $\mathcal{D}=\mathbb{Z}$ so_that $\mathbb{R}/\mathcal{D}\simeq_U(1)$._The use_of_Iwazawa decomposition yields_$SL(2,\mathbb{R})\simeq \mathbb{R}^2\times \mathbb{S}^1$,_which combined with $\pi_1(X\times_Y)=\pi_1(X)\times\pi_1(Y)$ for any_topological_spaces $X$ and $Y$ implies $\pi_1(SL(2,\mathbb{R}))\simeq\mathbb{Z}$. Hence, $Hom(\mathbb{Z}\to U(1))\simeq U(1)$ which classifies the inequivalent_extensions_of $SL(2,\mathbb{R})$_by_$U(1)$._Accordingly, one expect that the_extension of the present construction_of $SL(2,\mathbb{R})$_gives rise to inequivalent classes of star-products.\
Quantized plane_waves_for $\mathbb{R}^3_\theta$
===============================================
The determination_of which takes the generic form $$E_p(\hat{x})=\omega(p)e^{i\xi(p)\hat{x}} \label{generalform-ncexpo},$$ can_be achieved through a standard albeit_tedious computation whose_main_steps_are summarized now.\
First, using_$e^{i\xi(p)\hat{x}}\rhd 1 = \frac{e^{ipx}}{\omega(p)}$, one easily_infers that $$e^{-i\xi(\lambda p) \hat{x}} \partial_\mu_e^{i\xi(\lambda p) \hat{x}} = (i\lambda p_\mu) {{_\mathbb{I}}},\label{megaplus1}$$ which holds for any $\lambda\in\mathbb{R}$._Then, the combination of $[\partial_\mu,_\hat{x}_\nu] =_[\partial_\mu,x^a\varphi_{a\nu}] = \varphi_{\mu \nu}$ with_the functional derivative_of with_respect to $\lambda$_yields $$\label{diff-xi}
\varphi_{\mu\nu}(i\lambda p) \frac{d}{d\lambda} \left[ \xi^\nu(\lambda_p |
$\mathbb{S}(\mathfrak{g}^*)$ around $\pi_{\mathbb{S}}$ by the space $$\mathfrak{Casim}(\mathbb{S}(\mathfrak{g}^*),\pi_{\mathbb{S}})/Out(\mathfrak{g}).$$ In this section we describe this space using classical invariant theory.
Let $P[\mathfrak{g}^*]$ and $P[\mathfrak{t}^*]$ denote the algebras of polynomials on $\mathfrak{g}^*$ and $\mathfrak{t}^*$ respectively. A classical result (see e.g. Theorem 7.3.5 [@Dixmier]) states that the restriction map $P[\mathfrak{g}^*]\to P[\mathfrak{t}^*]$ induces an isomorphism between the algebras of invariants $$\label{EQ_Invariant_Polynomials}
P[\mathfrak{g}^*]^G\cong P[\mathfrak{t}^*]^W.$$ A theorem of Schwarz, extends this result to the smooth setting $$\label{EQ_Invariant_Functions}
C^{\infty}(\mathfrak{g}^*)^G\cong C^{\infty}(\mathfrak{t}^*)^W.$$ To explain this, first recall that $P[\mathfrak{g}^*]^G$ is generated by $l:=\mathrm{dim}(\mathfrak{t})$ algebraically independent homogeneous polynomials $p_1,\ldots,p_l$ (Theorem 7.3.8 [@Dixmier]). Hence by (\[EQ\_Invariant\_Polynomials\]), $P[\mathfrak{t}^*]^W$ is generated by $q_1:=p_{1|\mathfrak{t}^*},\ldots,q_l:=p_{l|\mathfrak{t}^*}$. Consider the maps $$p=(p_1,\ldots,p_l):\mathfrak{g}^*{\longrightarrow}\mathbb{R}^{l},\ \textrm{and}\ q=(q_1,\ldots,q_l):\mathfrak{t}^*{\longrightarrow}\mathbb{R}^{l},$$ and denote by $\Delta:=p(\mathfrak{g}^*)$. Since the inclusion $\mathfrak{t}^*\subset \mathfrak{g}^*$ induces a bijection between the $W$-orbits and the $G$-orbits, it follows that $q(\mathfrak{t}^*)=\Delta$. The theorem of Schwarz [@Schwarz] applied to the action of $G$ on $\mathfrak{g}^*$ and to the action of $W$ on $\mathfrak{t}^*$, shows that the pullbacks by $ | $ \mathbb{S}(\mathfrak{g}^*)$ around $ \pi_{\mathbb{S}}$ by the space $ $ \mathfrak{Casim}(\mathbb{S}(\mathfrak{g}^*),\pi_{\mathbb{S}})/Out(\mathfrak{g}).$$ In this section we describe this quad use classical invariant theory.
get $ P[\mathfrak{g}^*]$ and $ P[\mathfrak{t}^*]$ denote the algebras of polynomial on $ \mathfrak{g}^*$ and $ \mathfrak{t}^*$ respectively. A classical result (see for example Theorem 7.3.5 [ @Dixmier ]) express that the restriction map $ P[\mathfrak{g}^*]\to P[\mathfrak{t}^*]$ induce an isomorphism between the algebras of invariants $ $ \label{EQ_Invariant_Polynomials }
P[\mathfrak{g}^*]^G\cong P[\mathfrak{t}^*]^W.$$ A theorem of Schwarz, gallop this result to the smooth setting $ $ \label{EQ_Invariant_Functions }
C^{\infty}(\mathfrak{g}^*)^G\cong C^{\infty}(\mathfrak{t}^*)^W.$$ To excuse this, first echo that $ P[\mathfrak{g}^*]^G$ is generate by $ l:=\mathrm{dim}(\mathfrak{t})$ algebraically independent homogeneous polynomials $ p_1,\ldots, p_l$ (Theorem 7.3.8 [ @Dixmier ]). Hence by (\[EQ\_Invariant\_Polynomials\ ]), $ P[\mathfrak{t}^*]^W$ is beget by $ q_1:=p_{1|\mathfrak{t}^*},\ldots, q_l:=p_{l|\mathfrak{t}^*}$. Consider the maps $ $ p=(p_1,\ldots, p_l):\mathfrak{g}^*{\longrightarrow}\mathbb{R}^{l},\ \textrm{and}\ q=(q_1,\ldots, q_l):\mathfrak{t}^*{\longrightarrow}\mathbb{R}^{l},$$ and denote by $ \Delta:=p(\mathfrak{g}^*)$. Since the inclusion $ \mathfrak{t}^*\subset \mathfrak{g}^*$ induces a bijection between the $ W$-orbits and the $ G$-orbits, it come that $ q(\mathfrak{t}^*)=\Delta$. The theorem of Schwarz [ @Schwarz ] applied to the action of $ G$ on $ \mathfrak{g}^*$ and to the action of $ W$ on $ \mathfrak{t}^*$, shows that the pullbacks by $ | $\mahhbb{S}(\mathfrak{g}^*)$ around $\pi_{\oathbb{S}}$ by the space $$\methfrak{Dasim}(\matfbb{S}(\mathfrak{g}^*),\pi_{\mathbb{S}})/Out(\matifraj{g}).$$ In this section we descrkbe this dpace usung rlassical invarieht theovv.
Let $L[\lathyrek{g}^*]$ and $P[\mathfrsk{t}^*]$ denote the algebras mf pllynomials on $\mathfrak{g}^*$ and $\mathfrah{t}^*$ resprchively. A classycal wesumn (wee e.g. Theorem 7.3.5 [@Dixmier]) statea that uhe restriction mal $P[\mathfrak{g}^*]\to P[\mathfrak{t}^*]$ lndufes an isomorphism between thg aldwbras of invxriants $$\label{EQ_Invariaht_Polynomials}
P[\mathfrak{g}^*]^G\cong P[\mxthfrck{t}^*]^W.$$ A theotzn ov Schwarz, exvends nhis result to the smmoth seyting $$\label{EQ_Ikvarient_Fynctions}
C^{\infty}(\mathfran{g}^*)^G\cong C^{\infty}(\mathsrak{t}^*)^W.$$ To erplain this, first recqlo thaj $P[\madhfrxj{g}^*]^G$ is gxnedated hy $m:=\mathrm{dim}(\jathfrak{t})$ aogebraically indepemdqbt homogeneoua polygoiials $p_1,\ldots,p_l$ (Theorem 7.3.8 [@Dixmier]). Hence bj (\[EQ\_Jnvariant\_Polynomials\]), $P[\mqthfrak{t}^*]^W$ is generatef by $q_1:=p_{1|\maehfrak{t}^*},\ldots,q_l:=p_{l|\mathfrak{t}^*}$. Consider the maps $$p=(p_1,\ldods,p_l):\metffrcj{g}^*{\loneeihhtarrow}\mathbb{R}^{l},\ \textrm{and}\ q=(q_1,\ldots,q_l):\mathfrak{e}^*{\momggightarrow}\mathbb{R}^{j},$$ and denotr hy $\Qelta:=p(\mathfray{g}^*)$. Sines fhe inclusion $\mathvrak{t}^*\sufset \nathfrak{g}^*$ indices a bijection between thw $W$-orbits anb tye $G$-orbits, it folllws that $q(\mcthfral{t}^*)=\Delya$. The theorem of Schwaxz [@Schsarz] applief to the zztion of $G$ on $\maghfgak{g}^*$ and to ufe action of $W$ on $\mathfrak{v}^*$, shocs that ghe kullbachs by $ | $\mathbb{S}(\mathfrak{g}^*)$ around $\pi_{\mathbb{S}}$ by the space $$\mathfrak{Casim}(\mathbb{S}(\mathfrak{g}^*),\pi_{\mathbb{S}})/Out(\mathfrak{g}).$$ section describe this using classical invariant denote algebras of polynomials $\mathfrak{g}^*$ and $\mathfrak{t}^*$ A classical result (see e.g. Theorem [@Dixmier]) states that the restriction map $P[\mathfrak{g}^*]\to P[\mathfrak{t}^*]$ induces an isomorphism between the of invariants $$\label{EQ_Invariant_Polynomials} P[\mathfrak{g}^*]^G\cong P[\mathfrak{t}^*]^W.$$ A theorem of Schwarz, extends this result to smooth $$\label{EQ_Invariant_Functions} C^{\infty}(\mathfrak{t}^*)^W.$$ explain this, first recall that $P[\mathfrak{g}^*]^G$ is generated by $l:=\mathrm{dim}(\mathfrak{t})$ algebraically independent homogeneous polynomials $p_1,\ldots,p_l$ (Theorem [@Dixmier]). Hence by (\[EQ\_Invariant\_Polynomials\]), $P[\mathfrak{t}^*]^W$ is generated by Consider the maps $$p=(p_1,\ldots,p_l):\mathfrak{g}^*{\longrightarrow}\mathbb{R}^{l},\ q=(q_1,\ldots,q_l):\mathfrak{t}^*{\longrightarrow}\mathbb{R}^{l},$$ and denote by $\Delta:=p(\mathfrak{g}^*)$. the $\mathfrak{t}^*\subset \mathfrak{g}^*$ a between $W$-orbits and the it follows that $q(\mathfrak{t}^*)=\Delta$. The theorem of Schwarz [@Schwarz] applied to the action of $G$ on $\mathfrak{g}^*$ to the $W$ on shows the by $ | $\mathbb{S}(\mathfrak{g}^*)$ around $\pi_{\mAthbb{S}}$ by thE spacE $$\maThfRaK{CasIm}(\maThbb{S}(\mathfrak{g}^*),\PI_{\matHbb{S}})/Out(\mathfrak{g}).$$ In this SectiOn WE desCRiBe thiS space uSInG CLasSiCaL inVaRIaNt theOry.
let $P[\matHfrak{g}^*]$ and $P[\MatHfRak{t}^*]$ denote thE AlGebras of poLynOmials on $\mathFraK{g}^*$ and $\mAtHfrAK{t}^*$ resPecTivelY. A clasSIcal reSult (see e.g. thEOrem 7.3.5 [@DiXMier]) staTES tHat tHe restriction map $P[\MAtHFrak{g}^*]\to P[\mathfrAk{t}^*]$ indUcES aN ISomOrpHism betweeN tHe algEBras of iNVaRIANts $$\LAbel{EQ_InvariaNt_PolynomiaLS}
P[\mAthfraK{g}^*]^g\coNG P[\mathFrak{t}^*]^w.$$ A THeoRem of SchwarZ, extEnds this rEsult tO The smooTH settinG $$\label{eQ_INvaRianT_fuNcTioNs}
c^{\InfTY}(\mAthFRak{G}^*)^G\cong C^{\iNfTy}(\MathfRak{t}^*)^w.$$ tO EXplaIn tHis, fIrst rEcall that $P[\matHfrAk{g}^*]^G$ IS geNeratEd by $l:=\MathRm{Dim}(\maThfrak{T})$ algeBrAically independEnt hOmogeneouS poLyNomIaLs $p_1,\ldOTs,p_l$ (ThEorEm 7.3.8 [@DIxmier]). HEnce by (\[Eq\_invArIANT\_POlynomials\]), $P[\mathfraK{t}^*]^w$ IS gEnerated By $q_1:=p_{1|\maTHfRaK{T}^*},\ldots,q_l:=P_{l|\MatHfraK{T}^*}$. consiDer tHE mAps $$p=(p_1,\ldoTs,p_l):\maTHfRaK{g}^*{\longrIgHtarroW}\mAthBb{R}^{L},\ \textRM{and}\ Q=(q_1,\ldotS,q_l):\mathfRak{t}^*{\lONgrightarrow}\maTHbb{R}^{l},$$ and denotE By $\dELtA:=P(\matHfrAk{g}^*)$. Since the InclUSion $\MathFRaK{t}^*\sUBset \mAthfrAk{G}^*$ InDUces a bijection betweEn The $W$-orBits aNd the $G$-orbits, iT follows thAT $Q(\Mathfrak{T}^*)=\DelTA$. THE theorem of SchwArz [@ScHwarz] appliED to the acTion oF $G$ on $\mathFrak{g}^*$ and tO THe action Of $W$ On $\mAthFraK{T}^*$, ShOws that the pulLBAcks By $ | $\mathbb{S}(\mathfrak{g}^ *)$ around $\pi _{\ mat hb b{S} }$ b y the space $$ \ math frak{Casim}(\mathbb{S} (\mat hf r ak{g } ^* ),\pi _{\math b b{ S } })/ Ou t( \ma th f ra k{g}) .$$ In thi s sectionwede scribe thiss pa ce using c las sical invari ant theor y.
L e t $P[ \ma thfra k{g}^* ] $ and$P[\mathf ra k {t}^*] $ denote t he alg ebras of polynomi a ls on $\mathfrak{ g}^*$an d $ \ m ath fra k{t}^*$ re sp ectiv e ly. A c l as s i c alr esult (see e. g. Theorem7 .3. 5 [@Di xm ier ] ) stat es th at the restrictio n ma p $P[\mat hfrak{ g }^*]\to P[\math frak{t }^* ]$indu c es a n i so m orp h is m b e twe en the a lg eb ras o f in v a r i ants $$ \lab el{EQ _Invariant_Po lyn omia l s}P[\ma thfra k{g} ^* ]^G\c ong P[ \math fr ak{t}^*]^W.$$ A the orem of S chw ar z,ex tends this r esu ltto thesmooths ett in g $ $\ label{EQ_Invariant _F u n ct ions}
C^ {\inft y }( \m a thfrak{g }^ *)^ G\co n g C^{\ inft y }( \mathfra k{t}^* ) ^W .$ $ To ex pl ain th is , f irs t rec a ll t hat $P [\mathfr ak{g} ^ *]^G$ is gener a ted by $l:=\m a th r m {d i m}(\ mat hfrak{t})$alge b raic ally in dep e ndent homo ge n eo u s polynomials $p_1, \l dots,p _l$ ( Theorem 7.3.8 [@Dixmier ] ) . Hence b y (\ [ EQ \ _Invariant\_Po lynom ials\]), $ P [\mathfr ak{t} ^*]^W$ i s generat e d by $q_1 :=p _{1 |\m ath f r ak {t}^*},\ldots , q _l:= p_ {l|\mat hfr ak{t}^* }$. Co nsi der t he maps $ $p=(p_1, \l do ts ,p _l) :\mat h frak{g}^ *{ \lo ng rig htarr o w}\mat hbb{R }^{l }, \\te xtrm{an d }\ q =(q_ 1, \l dots ,q_ l) :\mat hfra k {t} ^*{\lon grightarr ow} \ math bb {R }^{l},$ $ and denoteby $\Delta:= p( \ma thfrak { g }^*)$. S ince the inclusion $\ma t hfrak{t }^* \subs et \ mathfrak{ g}^ *$ ind uce s a bij ection betw ee n t h e $W$- o r bi tsan d the $G$- o r bit s, it f ollo ws that $q(\mathfrak{t}^* ) =\D elta$. The th eor em o f Sc hwa r z[ @Sc hw a rz] a pplied to the a ction of $ G$ on $\mathfra k {g} ^* $ and t o the a ction of $W$on $\math frak{t}^* $, sho w s th at the pul lbacks b y $ | $\mathbb{S}(\mathfrak{g}^*)$_around $\pi_{\mathbb{S}}$_by the space $$\mathfrak{Casim}(\mathbb{S}(\mathfrak{g}^*),\pi_{\mathbb{S}})/Out(\mathfrak{g}).$$_In this_section_we describe_this_space using classical_invariant theory.
Let $P[\mathfrak{g}^*]$_and $P[\mathfrak{t}^*]$ denote the_algebras of polynomials_on_$\mathfrak{g}^*$ and $\mathfrak{t}^*$ respectively. A classical result (see e.g. Theorem 7.3.5 [@Dixmier]) states that_the_restriction map_$P[\mathfrak{g}^*]\to_P[\mathfrak{t}^*]$_induces an isomorphism between the_algebras of invariants $$\label{EQ_Invariant_Polynomials}
P[\mathfrak{g}^*]^G\cong P[\mathfrak{t}^*]^W.$$_A theorem_of Schwarz, extends this result to the smooth_setting_$$\label{EQ_Invariant_Functions}
C^{\infty}(\mathfrak{g}^*)^G\cong C^{\infty}(\mathfrak{t}^*)^W.$$ To_explain this, first recall that $P[\mathfrak{g}^*]^G$ is generated by_$l:=\mathrm{dim}(\mathfrak{t})$ algebraically independent homogeneous polynomials $p_1,\ldots,p_l$_(Theorem 7.3.8 [@Dixmier])._Hence_by_(\[EQ\_Invariant\_Polynomials\]), $P[\mathfrak{t}^*]^W$ is generated_by $q_1:=p_{1|\mathfrak{t}^*},\ldots,q_l:=p_{l|\mathfrak{t}^*}$. Consider the maps $$p=(p_1,\ldots,p_l):\mathfrak{g}^*{\longrightarrow}\mathbb{R}^{l},\_ \textrm{and}\ q=(q_1,\ldots,q_l):\mathfrak{t}^*{\longrightarrow}\mathbb{R}^{l},$$ and denote by_$\Delta:=p(\mathfrak{g}^*)$. Since the inclusion $\mathfrak{t}^*\subset \mathfrak{g}^*$ induces_a bijection between the $W$-orbits and_the $G$-orbits, it follows that_$q(\mathfrak{t}^*)=\Delta$. The_theorem of Schwarz [@Schwarz] applied_to the action_of $G$_on $\mathfrak{g}^*$ and_to the action of $W$ on_$\mathfrak{t}^*$, shows that_the pullbacks by $ |
D topological orders), the quasiparticles can only fuse but not braiding. So, the fusion rules $N_{\xi\zeta}^\chi$ and the F-matrices are enough to describe 1+1D anomalous topological orders. Later, we will see fusion rules and F-matrices are also used to determine a string-net wavefunction, which may seem confusing. However, as we have mentioned, this is a natural result of the holographic bulk-edge relation. Intuitively, one may even view the string-net graphs in 2D space as the 1+1D space-time trajectory of the edge quasiparticles.
For 2+1D topological orders, the quasiparticles can also braid. We also need data to describe the braiding of the quasiparticles in addition to the fusion rules and the F-matrices, as introduced in the next two subsections.
Quasiparticle intrinsic spin
----------------------------
If we twist the quasiparticle at ${\boldsymbol{x}}_1$ by rotating $\Delta H$ at ${\boldsymbol{x}}_1$ by 360$^\circ$ (note that $\Delta H$ at ${\boldsymbol{x}}_1$ has no rotational symmetry), all the degenerate ground states in $\mathcal V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$ will acquire the same geometric phase ${\mathrm{e}}^{{\mathrm{i}}\theta_{\xi_1}}$ provided that the quasiparticle type $\xi_1$ is a simple type. We will call ${\mathrm{e}}^{{\mathrm{i}}\theta_{\xi}}$ the intrinsic spin (or simply spin) of the simple type $\xi$, which is a universal property of the topologically ordered state.
Quasiparticle mutual statistics
-------------------------------
If we move the quasiparticle $\xi_2$ at ${\boldsymbol{x}}_2$ around the quasiparticle $\xi_1$ at ${\boldsymbol{x}}_1$, we will generate a non-Abelian geometric phase – a unitary transformation acting on the degenerate ground states in $\mathcal
V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$. Such a unitary transformation not only depends on the types $\xi_1$ and $\xi_2$, but also depends on the quasiparticles at other places. So, here we will consider three quasiparticles of simple types $\xi$, $\zeta$, $\chi$ on a | D topological orders), the quasiparticles can only blend but not braid. So, the fusion predominate $ N_{\xi\zeta}^\chi$ and the F - matrix are enough to describe 1 + 1D anomalous topological order. by and by, we will see coalition rule and F - matrices are besides used to determine a string - net wavefunction, which may look confusing. However, as we have mentioned, this is a natural result of the holographic bulk - edge relation. Intuitively, one may even see the string - net graphs in 2D quad as the 1 + 1D space - time trajectory of the edge quasiparticles.
For 2 + 1D topological order, the quasiparticles can also braid. We also need data to identify the braiding of the quasiparticles in addition to the fusion rules and the F - matrices, as introduced in the next two subsections.
Quasiparticle intrinsic spin
----------------------------
If we twist the quasiparticle at $ { \boldsymbol{x}}_1 $ by rotating $ \Delta H$ at $ { \boldsymbol{x}}_1 $ by 360$^\circ$ (note that $ \Delta H$ at $ { \boldsymbol{x}}_1 $ has no rotational symmetry), all the degenerate ground state in $ \mathcal V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$ will acquire the same geometric phase $ { \mathrm{e}}^{{\mathrm{i}}\theta_{\xi_1}}$ provide that the quasiparticle type $ \xi_1 $ is a simple type. We will call $ { \mathrm{e}}^{{\mathrm{i}}\theta_{\xi}}$ the intrinsic spin (or plainly spin) of the simple type $ \xi$, which is a universal place of the topologically ordered state.
Quasiparticle mutual statistics
-------------------------------
If we move the quasiparticle $ \xi_2 $ at $ { \boldsymbol{x}}_2 $ around the quasiparticle $ \xi_1 $ at $ { \boldsymbol{x}}_1 $, we will generate a non - Abelian geometric phase – a unitary transformation acting on the degenerate ground states in $ \mathcal
V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$. Such a unitary transformation not merely depends on the type $ \xi_1 $ and $ \xi_2 $, but also depends on the quasiparticles at other places. therefore, here we will consider three quasiparticles of simple type $ \xi$, $ \zeta$, $ \chi$ on a | D tlpological orders), the quxsiparticles cau only huse buf not brxiding. So, the fusion rules $N_{\ei\zera}^\chi$ and the F-matrices are enough tl descrive 1+1D qnomalous vkpologigcl orscrs. Lctxr, we will see nusion rulev and F-matricev xrz also used to determine a string-net wavefumchion, which may seek conrlslng. However, as we have mentioned, this iv a natural rrsult of the holographic bklk-efge relation. Intuihively, one nay qcen view the string-net graphs in 2D space as the 1+1D space-time trajeztory of the edte quwviparticles.
Hor 2+1D nopological ovcers, tve quasoparticles can alvo vraid. We also need dava to describe the btaiding of tke quasiparticles in qdeitiot to the duskon rnlea and hhe F-matrices, as introduxed in the next two stvsections.
Quasjpartislq intrinsic spin
----------------------------
If we twist the quasipagticme at ${\boldsymbol{x}}_1$ by rorating $\Delta H$ at ${\bolfsymbol{x}}_1$ fy 360$^\circ$ (note that $\Delta H$ at ${\boldsymbol{x}}_1$ has no rmtatikval snmmegey), all the degenerate ground states in $\mathcal D(J^d;\ci_1,\qi_2,\xi_3,\cdots)$ will aczuire the ssmf bgometric phase ${\mathrm{e}}^{{\mzthrm{i}}\theta_{\xi_1}}$ provlded thwt thw quasipawticke type $\xi_1$ is a simple type. We will calj ${\mathrm{e}}^{{\mathrm{i}}\theca_{\xi}}$ the intxinsic spin (or simply spin) of the rimpme type $\xi$, ahich is z universal propefty ox the tokulogically ordereq state.
Quesipaxticle mjtuak statystics
-------------------------------
If wf movc the quasiparticle $\xi_2$ aj ${\boldvymbol{x}}_2$ arlund the quasiparticle $\xi_1$ at ${\boldsymbol{x}}_1$, we woln gvnerate a non-Anelian geometris phase – a unijary tranfformxtion actihg on tie degeneratq ground statad in $\mathcal
T(M^d;\xi_1,\xi_2,\xi_3,\sdotw)$. Suxh a unkgary transformstion not only depebds on the types $\xl_1$ and $\si_2$, but also depzudw on the quasipsrtkclqs av othqs places. So, vere we eill zonsider thvee quaxiparticles of simpla tyles $\xi$, $\zeta$, $\chi$ on a | D topological orders), the quasiparticles can only not So, the rules $N_{\xi\zeta}^\chi$ and describe anomalous topological orders. we will see rules and F-matrices are also used determine a string-net wavefunction, which may seem confusing. However, as we have mentioned, is a natural result of the holographic bulk-edge relation. Intuitively, one may even the graphs 2D as the 1+1D space-time trajectory of the edge quasiparticles. For 2+1D topological orders, the quasiparticles can braid. We also need data to describe the of the quasiparticles in to the fusion rules and F-matrices, introduced in next subsections. intrinsic spin ---------------------------- we twist the quasiparticle at ${\boldsymbol{x}}_1$ by rotating $\Delta H$ at ${\boldsymbol{x}}_1$ by 360$^\circ$ (note that $\Delta at ${\boldsymbol{x}}_1$ rotational symmetry), the ground in $\mathcal V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$ the same geometric phase ${\mathrm{e}}^{{\mathrm{i}}\theta_{\xi_1}}$ provided type $\xi_1$ is a simple type. We will ${\mathrm{e}}^{{\mathrm{i}}\theta_{\xi}}$ the spin (or simply spin) of the type $\xi$, which is a universal property of topologically ordered state. Quasiparticle mutual statistics ------------------------------- If we move the quasiparticle $\xi_2$ at ${\boldsymbol{x}}_2$ quasiparticle $\xi_1$ at ${\boldsymbol{x}}_1$, will generate a geometric – unitary acting on degenerate ground states in $\mathcal V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$. Such a unitary transformation not depends on the types $\xi_1$ and $\xi_2$, but also depends quasiparticles other places. So, we will consider three of types $\xi$, $\zeta$, $\chi$ | D topological orders), the quasIparticles Can onLy fUse BuT not BraiDing. So, the fusioN RuleS $N_{\xi\zeta}^\chi$ and the F-matrIces aRe ENougH To DescrIbe 1+1D anoMAlOUS toPoLoGicAl ORdErs. LaTer, We will sEe fusion ruLes AnD F-matrices arE AlSo used to deTerMine a string-nEt wAvefunCtIon, WHich mAy sEem coNfusinG. howeveR, as we have MeNTioned, THis is a nATUrAl reSult of the holograpHIc BUlk-edge relatioN. IntuiTiVElY, ONe mAy eVen view the StRing-nET graphs IN 2D SPACe aS The 1+1D space-time Trajectory oF The Edge quAsIpaRTicles.
for 2+1D tOpOLogIcal orders, tHe quAsiparticLes can ALso braiD. we also nEed datA to DesCribE ThE bRaiDiNG of THe QuaSIpaRticles iN aDdItion To thE FUSIon rUleS and The F-mAtrices, as intrOduCed iN The Next tWo subSectIoNs.
QuaSipartIcle iNtRinsic spin
----------------------------
If we tWist The quasipArtIcLe aT ${\bOldsyMBol{x}}_1$ by RotAtiNg $\Delta h$ at ${\boldSYmbOl{X}}_1$ BY 360$^\CiRc$ (note that $\Delta H$ at ${\BoLDSyMbol{x}}_1$ has No rotaTIoNaL Symmetry), AlL thE degENErate GrouND sTates in $\mAthcal v(m^d;\Xi_1,\Xi_2,\xi_3,\cdoTs)$ Will acQuIre The Same gEOmetRic phaSe ${\mathrm{E}}^{{\mathRM{i}}\theta_{\xi_1}}$ proviDEd that the quasIPaRTIcLE typE $\xi_1$ Is a simple tyPe. We WIll cAll ${\mAThRm{e}}^{{\MAthrm{I}}\thetA_{\xI}}$ ThE Intrinsic spin (or simpLy Spin) of The siMple type $\xi$, whiCh is a univeRSAL propertY of tHE tOPologically ordEred sTate.
QuasipARticle muTual sTatisticS
-------------------------------
If we move THE quasipaRtiCle $\Xi_2$ aT ${\boLDSyMbol{x}}_2$ around thE QUasiPaRticle $\xI_1$ at ${\BoldsymBol{X}}_1$, we WilL geNeRate a non-ABelian geOmEtRiC pHasE – a uniTAry transFoRmaTiOn aCting ON the deGenerAte gRoUnD StaTes in $\maTHcAL
v(M^d;\xI_1,\xI_2,\xI_3,\cdoTs)$. SUcH a uniTary TRanSformatIon not onlY dePEnds On ThE types $\xI_1$ and $\xi_2$, but also DePends on the QuAsiParticLES at other Places. So, here we will consiDEr three QuaSiparTiclEs of simplE tyPes $\xi$, $\zEta$, $\CHi$ on a | D topological orders), the quasipart icles ca n o nl y fu se b ut not braidin g . So , the fusion rules $N_ {\xi\ ze t a}^\ c hi $ and the F- m at r i ces a re en ou g hto de scr ibe 1+1 D anomalou s t op ological ord e rs . Later, w e w ill see fusi onrulesan d F - matri ces arealso u s ed todetermine a string - net wav e f un ctio n, which may seem co n fusing. Howeve r, aswe ha v e me nti oned, this i s a n a tural r e su l t oft he holographi c bulk-edge rel ation. I ntu i tively , one m a y e ven view th e st ring-netgraphs in 2D s p ace asthe 1+ 1Dspa ce-t i me t raj ec t ory of th e ed ge quasi pa rt icles .
F o r 2 +1Dtop olog icalorders, the q uas ipar t icl es ca n als o br ai d. We alsoneedda ta to describethebraidingofth e q ua sipar t iclesinadd ition t o the f u sio nr u l es and the F-matrice s, a sintroduc ed int he n e xt two s ub sec tion s .
Qua sipa r ti cle intr insics pi n------- -- ------ -- --- --- -----
Ifwe twi st the q uasip a rticle at ${\b o ldsymbol{x}}_ 1 $b y r o tati ng$\Delta H$at $ { \bol dsym b ol {x} } _1$ b y 360 $^ \ ci r c$ (note that $\Del ta H$ at ${\b oldsymbol{x}} _1$ has no r o tational sym m et r y), all the de gener ate ground states i n $\m athcal V (M^d;\xi_ 1 , \xi_2,\x i_3 ,\c dot s)$ w il l acquire the s amege ometric ph ase ${\ mat hrm {e} }^{ {\ mathrm{i} }\theta_ {\ xi _1 }} $ p rovid e d that t he qu as ipa rticl e type$\xi_ 1$ i sas imp le type . W e will c al l ${ \ma th rm{e} }^{{ \ mat hrm{i}} \theta_{\ xi} } $ th ein trinsic spin (or sim pl y spin) of t hesimple t ype $\xi $, which is a universal propert y o f the top ologicall y o rdered st a te.
Q uasipa rticl emut u a l sta t i st ics
- ---------- - - --- ----- -- ---- ----
I f we move the quas i par ticle $\xi_2$ at ${\ b o ld sym b ol { x}} _2 $ ar o u nd the quasipar ticle $\xi _1 $ a t ${\bolds y mbo l{ x}}_1$, we wil l gen e rate anon-Abeli an geomet ri c ph a s e – a unitary transfo rmation a c tingo nthe d ege nerate g rou nd st ates i n $\ mathc al
V(M ^d ;\xi_1 ,\xi_ 2, \xi_3,\c dots)$. Such a unitarytransf ormat ion not only de p end s on thetype s $\xi_1$and $\ xi_2$ , b u t als o de p en dso n the qua s iparticle s a t o t h er places. So , h ere we w ill consid er t hree quasiparticl e s of simple ty pes$ \ xi$ , $ \ zeta $, $\chi$ on a | D topological_orders), the_quasiparticles can only fuse_but not_braiding._So, the_fusion_rules $N_{\xi\zeta}^\chi$ and_the F-matrices are_enough to describe 1+1D_anomalous topological orders._Later,_we will see fusion rules and F-matrices are also used to determine a string-net_wavefunction,_which may_seem_confusing._However, as we have mentioned,_this is a natural result_of the_holographic bulk-edge relation. Intuitively, one may even view_the_string-net graphs in_2D space as the 1+1D space-time trajectory of the_edge quasiparticles.
For 2+1D topological orders, the_quasiparticles can also_braid._We_also need data to_describe the braiding of the quasiparticles_in addition to the fusion rules_and the F-matrices, as introduced in the_next two subsections.
Quasiparticle intrinsic spin
----------------------------
If we_twist the quasiparticle at ${\boldsymbol{x}}_1$_by rotating_$\Delta H$ at ${\boldsymbol{x}}_1$ by_360$^\circ$ (note that_$\Delta H$_at ${\boldsymbol{x}}_1$ has_no rotational symmetry), all the degenerate_ground states in_$\mathcal V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$ will acquire the same_geometric_phase ${\mathrm{e}}^{{\mathrm{i}}\theta_{\xi_1}}$ provided_that_the_quasiparticle type_$\xi_1$ is a_simple_type. We_will_call ${\mathrm{e}}^{{\mathrm{i}}\theta_{\xi}}$ the intrinsic spin (or_simply_spin) of the simple type $\xi$, which_is a universal property_of_the topologically ordered state.
Quasiparticle_mutual statistics
-------------------------------
If we move the_quasiparticle $\xi_2$ at ${\boldsymbol{x}}_2$ around the_quasiparticle $\xi_1$_at ${\boldsymbol{x}}_1$,_we will generate a non-Abelian geometric phase – a unitary transformation_acting on the degenerate ground states_in $\mathcal
V(M^d;\xi_1,\xi_2,\xi_3,\cdots)$. Such a_unitary transformation_not_only depends on_the_types $\xi_1$_and $\xi_2$, but also depends on the_quasiparticles at_other places. So, here we will_consider three quasiparticles of_simple_types $\xi$, $\zeta$, $\chi$ on a |
\exp\left( \frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2-3\kappa^2(\alpha)\right)+O\left(\frac{1}{k^2}\right)\right).\end{aligned}$$]{}
In [@Proc:Li_SODA08], it was proved that, as $k\rightarrow\infty$, [$$\begin{aligned}
\notag
&\frac{\left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}t\right)
\Gamma\left(1-\frac{t}{k}\right)\Gamma\left(\frac{\alpha}{k}t\right)\right]^{k}}
{ \left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}\right)
\Gamma\left(1-\frac{1}{k}\right)\Gamma\left(\frac{\alpha}{k}\right)\right]^{kt}
}\\\notag
=&1+\frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2\right)+O\left(\frac{1}{k^2}\right)\\\notag
=&\exp\left( \frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2\right)+O\left(\frac{1}{k^2}\right)\right).\end{aligned}$$]{} Using the infinite product representation of cosine[@Book:Gradshteyn_00 1.43.3] [$$\begin{aligned}
\notag
\cos(z) = \prod_{s=0}^\infty\left(1 - \frac{4z^2}{(2s+1)^2\pi^2}\right),\end{aligned}$$]{} we can rewrite [$$\begin{aligned}
\notag
&\frac{\cos^{k}\left(\frac{\kappa(\alpha)\pi}{2k}t\right) } { \cos^{kt}\left(\frac{\kappa(\alpha)\pi}{2k}\right) }
= \prod_{s=0}^\infty \left(
1 - \frac{\kappa^2(\alpha)t^2}{(2s+1)^2k^2}\right | \exp\left (\frac{1}{k}\frac{\pi^2(t^2 - t)}{24}\left(\alpha^2 + 2 - 3\kappa^2(\alpha)\right)+O\left(\frac{1}{k^2}\right)\right).\end{aligned}$$ ] { }
In [ @Proc: Li_SODA08 ], it was proved that, as $ k\rightarrow\infty$, [ $ $ \begin{aligned }
\notag
& \frac{\left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}t\right)
\Gamma\left(1-\frac{t}{k}\right)\Gamma\left(\frac{\alpha}{k}t\right)\right]^{k } }
{ \left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}\right)
\Gamma\left(1-\frac{1}{k}\right)\Gamma\left(\frac{\alpha}{k}\right)\right]^{kt }
} \\\notag
= & 1+\frac{1}{k}\frac{\pi^2(t^2 - t)}{24}\left(\alpha^2 + 2\right)+O\left(\frac{1}{k^2}\right)\\\notag
= & \exp\left (\frac{1}{k}\frac{\pi^2(t^2 - t)}{24}\left(\alpha^2 + 2\right)+O\left(\frac{1}{k^2}\right)\right).\end{aligned}$$ ] { } Using the infinite product theatrical performance of cosine[@Book: Gradshteyn_00 1.43.3 ] [ $ $ \begin{aligned }
\notag
\cos(z) = \prod_{s=0}^\infty\left(1 - \frac{4z^2}{(2s+1)^2\pi^2}\right),\end{aligned}$$ ] { } we can rewrite [ $ $ \begin{aligned }
\notag
& \frac{\cos^{k}\left(\frac{\kappa(\alpha)\pi}{2k}t\right) } { \cos^{kt}\left(\frac{\kappa(\alpha)\pi}{2k}\right) }
= \prod_{s=0}^\infty \left (
1 - \frac{\kappa^2(\alpha)t^2}{(2s+1)^2k^2}\right | \exp\peft( \frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alkha^2+2-3\kappa^2(\alpha)\righj)+O\oeft(\frec{1}{k^2}\righf)\right).\ena{aligned}$$]{}
In [@Proc:Li_SODA08], it was peoved that, as $k\rightarrow\inwty$, [$$\begin{wligned}
\nitag
&\hrac{\left[\frac{2}{\pi}\sii\meft(\frag{\'i\alpgw}{2k}t\rngit)
\Gamma\left(1-\frac{j}{k}\right)\Gamma\neft(\frac{\alpha}{k}d\rkgkt)\right]^{k}}
{ \left[\frac{2}{\pi}\sin\left(\frac{\pi\alphw}{2k}\right)
\Balma\left(1-\frac{1}{k}\ridht)\Gsima\lsft(\frac{\alpha}{k}\right)\right]^{kt}
}\\\notag
=&1+\frac{1}{i}\frac{\pi^2(u^2-t)}{24}\left(\alpha^2+2\right)+O\lrft(\frac{1}{k^2}\right)\\\notag
=&\exp\left( \vrac{1}{n}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2\gight)+O\left(\ftzc{1}{k^2}\wught)\right).\end{xligned}$$]{} Uspug the infihite product representation of zosinz[@Book:Gradshjztn_00 1.43.3] [$$\tegin{alignev}
\notag
\bos(z) = \prod_{s=0}^\innny\left(1 - \frac{4z^2}{(2s+1)^2\li^2}\right),\end{aligked}$$]{} wx cab rewrite [$$\begin{alignev}
\notag
&\frac{\cos^{k}\left(\frwc{\kappa(\al[hc)\pi}{2k}t\right) } { \cos^{kt}\lefr(\feac{\kakpa(\al[ha)\pk}{2j}\rieht) }
= \lrod_{s=0}^\ijftb \left(
1 - \frad{\kappa^2(\alpha)r^2}{(2s+1)^2k^2}\right | \exp\left( \frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2-3\kappa^2(\alpha)\right)+O\left(\frac{1}{k^2}\right)\right).\end{aligned}$$]{} In [@Proc:Li_SODA08], it was proved $k\rightarrow\infty$, \notag &\frac{\left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}t\right) { \left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}\right) \Gamma\left(1-\frac{1}{k}\right)\Gamma\left(\frac{\alpha}{k}\right)\right]^{kt} the product representation of 1.43.3] [$$\begin{aligned} \notag = \prod_{s=0}^\infty\left(1 - \frac{4z^2}{(2s+1)^2\pi^2}\right),\end{aligned}$$]{} we can [$$\begin{aligned} \notag &\frac{\cos^{k}\left(\frac{\kappa(\alpha)\pi}{2k}t\right) } { \cos^{kt}\left(\frac{\kappa(\alpha)\pi}{2k}\right) } = \prod_{s=0}^\infty \left( 1 - \frac{\kappa^2(\alpha)t^2}{(2s+1)^2k^2}\right | \exp\left( \frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\aLpha^2+2-3\kappa^2(\aLpha)\rIghT)+O\lEfT(\fraC{1}{k^2}\riGht)\right).\end{aliGNed}$$]{}
IN [@Proc:Li_SODA08], it was proved That, aS $k\RIghtARrOw\infTy$, [$$\begin{ALiGNEd}
\nOtAg
&\FraC{\lEFt[\Frac{2}{\pI}\siN\left(\frAc{\pi\alpha}{2k}T\riGhT)
\Gamma\left(1-\frAC{t}{K}\right)\GammA\leFt(\frac{\alpha}{k}T\riGht)\rigHt]^{K}}
{ \leFT[\frac{2}{\Pi}\sIn\lefT(\frac{\pI\Alpha}{2k}\Right)
\GammA\lEFt(1-\frac{1}{K}\Right)\GaMMA\lEft(\fRac{\alpha}{k}\right)\rigHT]^{kT}
}\\\Notag
=&1+\frac{1}{k}\frac{\Pi^2(t^2-t)}{24}\leFt(\ALpHA^2+2\RigHt)+O\Left(\frac{1}{k^2}\rIgHt)\\\notAG
=&\exp\lefT( \FrAC{1}{K}\FraC{\Pi^2(t^2-t)}{24}\left(\alpha^2+2\Right)+O\left(\fRAc{1}{k^2}\Right)\rIgHt).\eND{alignEd}$$]{} UsiNg THe iNfinite prodUct rEpresentaTion of COsine[@BoOK:GradshTeyn_00 1.43.3] [$$\beGin{AliGned}
\NOtAg
\Cos(Z) = \pROd_{s=0}^\INfTy\lEFt(1 - \fRac{4z^2}{(2s+1)^2\pi^2}\rIgHt),\End{alIgneD}$$]{} WE CAn reWriTe [$$\beGin{alIgned}
\notag
&\fraC{\coS^{k}\leFT(\frAc{\kapPa(\alpHa)\pi}{2K}t\Right) } { \Cos^{kt}\lEft(\frAc{\Kappa(\alpha)\pi}{2k}\riGht) }
= \pRod_{s=0}^\infty \LefT(
1 - \fRac{\KaPpa^2(\alPHa)t^2}{(2s+1)^2k^2}\rIghT | \exp\left( \frac{1}{k}\fra c{\pi^2(t^ 2-t)} {24 }\l ef t(\a lpha ^2+2-3\kappa^2 ( \alp ha)\right)+O\left(\fra c{1}{ k^ 2 }\ri g ht )\rig ht).\en d {a l i gne d} $$ ]{}
I n[@Pro c:L i_SODA0 8], it was pr ov ed that, as$ k\ rightarrow \in fty$, [$$\be gin {align ed }
\ n otag&\f rac{\ left[\ f rac{2} {\pi}\sin \l e ft(\fr a c{\pi\a l p ha }{2k }t\right)
\Gamma\ l ef t (1-\frac{t}{k} \right )\ G am m a \le ft( \frac{\alp ha }{k}t \ right)\ r ig h t ] ^{k } }
{ \left[\fr ac{2}{\pi}\ s in\ left(\ fr ac{ \ pi\alp ha}{2 k} \ rig ht)
\Gamma\ left (1-\frac{ 1}{k}\ r ight)\G a mma\lef t(\fra c{\ alp ha}{ k }\ ri ght )\ r igh t ]^ {kt }
}\ \\notag=& 1+ \frac {1}{ k } \ f rac{ \pi ^2(t ^2-t) }{24}\left(\a lph a^2+ 2 \ri ght)+ O\lef t(\f ra c{1}{ k^2}\r ight) \\ \notag
=&\exp\l eft( \frac{1} {k} \f rac {\ pi^2( t ^2-t)} {24 }\l eft(\al pha^2+2 \ rig ht ) + O \l eft(\frac{1}{k^2}\ ri g h t) \right). \end{a l ig ne d }$$]{} U si ngthei n finit e pr o du ct repre sentat i on o f cosin e[ @Book: Gr ads hte yn_00 1.43 .3] [$ $\begin{ align e d}
\notag
\cos ( z) = \prod_{s = 0} ^ \ in f ty\l eft (1 - \frac{ 4z^2 } {(2s +1)^ 2 \p i^2 } \righ t),\e nd { al i gned}$$]{} we can r ew rite [ $$\be gin{aligned}\notag
&\f r a c {\cos^{k }\le f t( \ frac{\kappa(\a lpha) \pi}{2k}t\ r ight) }{ \co s^{kt}\l eft(\frac { \ kappa(\a lph a)\ pi} {2k } \ ri ght) }
= \p r o d_{s =0 }^\inft y \ left(
1 -\fr ac{ \ka pp a^2(\alph a)t^2}{( 2s +1 )^ 2k ^2} \righ t | \exp\left( \frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2-3\kappa^2(\alpha)\right)+O\left(\frac{1}{k^2}\right)\right).\end{aligned}$$]{}
In_[@Proc:Li_SODA08], it_was proved that, as_$k\rightarrow\infty$, [$$\begin{aligned}
\notag
&\frac{\left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}t\right)
\Gamma\left(1-\frac{t}{k}\right)\Gamma\left(\frac{\alpha}{k}t\right)\right]^{k}}
{_\left[\frac{2}{\pi}\sin\left(\frac{\pi\alpha}{2k}\right)
\Gamma\left(1-\frac{1}{k}\right)\Gamma\left(\frac{\alpha}{k}\right)\right]^{kt}
}\\\notag
=&1+\frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2\right)+O\left(\frac{1}{k^2}\right)\\\notag
=&\exp\left(_\frac{1}{k}\frac{\pi^2(t^2-t)}{24}\left(\alpha^2+2\right)+O\left(\frac{1}{k^2}\right)\right).\end{aligned}$$]{} Using_the_infinite product representation_of cosine[@Book:Gradshteyn_00 1.43.3]_[$$\begin{aligned}
\notag
\cos(z) = \prod_{s=0}^\infty\left(1 -_\frac{4z^2}{(2s+1)^2\pi^2}\right),\end{aligned}$$]{} we can_rewrite_[$$\begin{aligned}
\notag
&\frac{\cos^{k}\left(\frac{\kappa(\alpha)\pi}{2k}t\right) } { \cos^{kt}\left(\frac{\kappa(\alpha)\pi}{2k}\right) }
= \prod_{s=0}^\infty \left(
1 - \frac{\kappa^2(\alpha)t^2}{(2s+1)^2k^2}\right |
columnwidth"}![Estimation performance of input APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP values for the test set. The blue line indicates the line of perfect estimation, while the red line represents the best linear fitting for the points. The summary errors are given in Table 7.[]{data-label="f02"}](fig70.ps "fig:"){width="0.3\columnwidth"}![Estimation performance of input APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP values for the test set. The blue line indicates the line of perfect estimation, while the red line represents the best linear fitting for the points. The summary errors are given in Table 7.[]{data-label="f02"}](fig71.ps "fig:"){width="0.3\columnwidth"}\
![Estimation performance of input APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP values for the test set. The blue line indicates the line of perfect estimation, while the red line represents the best linear fitting for the points. The summary errors are given in Table 7.[]{data-label="f02"}](fig72.ps "fig:"){width="0.3\columnwidth"}![Estimation performance of input APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP values for the test set. The blue line indicates the line of perfect estimation, while the red line represents the best linear fitting for the points. The summary errors are given in Table 7.[]{data-label="f02"}](fig73.ps "fig:"){width="0.3\columnwidth"}![Estimation performance of input APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP values for the test set. The blue line indicates the line of perfect estimation, while the red line represents the best linear fitting for the points. The summary errors are given in Table 7.[]{data-label="f02"}](fig74.ps "fig:"){width="0.3\columnwidth"}
From table \[t02\] and figure \[f02\] we see that the results are quite good only for the case of early type galaxies and especially for the $p_1$ parameter. In comparison with the results derived for the synthetic library of galaxy | columnwidth"}![Estimation performance of input APs of the PÉGASE models. For each of the APs we plot the predict vs. genuine AP values for the test set. The blasphemous line indicates the tune of arrant estimation, while the red line represents the best linear meet for the points. The summary errors are given in Table 7.[]{data - label="f02"}](fig70.ps " fig:"){width="0.3\columnwidth"}![Estimation performance of input APs of the PÉGASE models. For each of the APs we diagram the predicted vs. true AP values for the examination set. The blue line indicates the argumentation of perfect estimation, while the red line represents the good linear fitting for the points. The summary errors are given in Table 7.[]{data - label="f02"}](fig71.ps " fig:"){width="0.3\columnwidth"}\
! [ Estimation operation of input signal APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP values for the test set. The blue line bespeak the line of perfect estimate, while the crimson line represents the best linear fit for the points. The summary errors are given in Table 7.[]{data - label="f02"}](fig72.ps " fig:"){width="0.3\columnwidth"}![Estimation performance of stimulation APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP values for the test bent. The blue line indicates the line of arrant estimation, while the crimson lineage represents the best linear fitting for the point. The summary errors are given in Table 7.[]{data - label="f02"}](fig73.ps " fig:"){width="0.3\columnwidth"}![Estimation performance of stimulation APs of the PÉGASE models. For each of the APs we plot the predicted vs. true AP value for the test bent. The blue line bespeak the line of arrant estimation, while the red line represents the good linear fitting for the points. The summary errors are given in Table 7.[]{data - label="f02"}](fig74.ps " fig:"){width="0.3\columnwidth " }
From table \[t02\ ] and calculate \[f02\ ] we see that the solution are quite good only for the case of early type galaxies and particularly for the $ p_1 $ parameter. In comparison with the results deduce for the synthetic library of galaxy | colkmnwidth"}![Estimation perfovmance of input CPs of vhe PÉGAAE modelr. For each of the APs we plov thw preeicted vs. true AP valuer for the test ser. Tht blue line indicefes the line kn peryert estimation, wmile the reg line represettr che best linear fitting for the poines. The xulmary errors ate gineg in Nanle 7.[]{data-label="f02"}](fig70.ps "fig:"){width="0.3\colujnwidth"}![Tstimation performsnce of input APs of the PÉHASE models. For each ov the APs wg pljr the predicged vs. true AP values fkr the test set. The blue line ivdicaces the ling kf kerfect estinatiog, while the vvd line represrnts the best kinxar ditting for the pointv. The summary errots are givan in Table 7.[]{data-labwl="d02"}](fig71.pv "fic:"){widgy="0.3\couumhwmdtg"}\
![Estimwtikn performznce of inpyt APs of the PÉGASE mjeels. For each of thq WPs we plot the predicted vs. true AP valles ror the test set. The blye line indicates the line of [erfect estimation, while the red line represents dhe bxsg lnkcar wuthing for the points. The summary errors are gidsn ik Table 7.[]{data-labej="f02"}](fig72.ps "fig:"){eiftn="0.3\solumnwidth"}![Esjimatiou psrformance of inpuh APs os the PÉGASE moqels. For each of the APs we plor the predicned cs. true AP values flr the test set. The nlue line indicates the lins of perfech estimatjun, while the red like sepresenur the best linear fitting hor tke pointr. Thg summawy errors wre gleen in Table 7.[]{data-lwbel="f02"}](yig73.ps "fig:"){width="0.3\cllumnwidth"}![Estimation performancx of input APx mf nhe PÉGASE modeks. For each os the APs we pkot the predkcted vs. trle AP valnes for the eest set. The tpue line indmcates thq libe od perfezg estimation, wnile the geb line represents the best kingad fitting for tkt piints. The summaty drrjrd ere gyeen in Table 7.[]{datx-lacrl="f02"}](fie74.ps "fig:"){widtm="0.3\couumneidth"}
From table \[t02\] ang fifure \[f02\] we see that tme resultw are quyte good only for the case of ewrly vype gelaxiex agd especially for the $p_1$ paramefer. In colpavison with thq rewults deriveb for the synthetic library of galaxy | columnwidth"}![Estimation performance of input APs of the For of the we plot the for test set. The line indicates the of perfect estimation, while the red represents the best linear fitting for the points. The summary errors are given Table 7.[]{data-label="f02"}](fig70.ps "fig:"){width="0.3\columnwidth"}![Estimation performance of input APs of the PÉGASE models. For each the we the vs. true AP values for the test set. The blue line indicates the line of perfect while the red line represents the best linear for the points. The errors are given in Table "fig:"){width="0.3\columnwidth"}\ performance of APs the models. For each the APs we plot the predicted vs. true AP values for the test set. The blue line the line estimation, while red represents best linear fitting points. The summary errors are given "fig:"){width="0.3\columnwidth"}![Estimation performance of input APs of the PÉGASE For each the APs we plot the predicted true AP values for the test set. The line indicates the line of perfect estimation, while the red line represents the best linear the points. The summary are given in 7.[]{data-label="f02"}](fig73.ps performance input of the models. For each of the APs we plot the predicted vs. AP values for the test set. The blue line indicates of estimation, while the line represents the best fitting the points. The summary given Table table and \[f02\] we see that results are quite good only the case of early the $p_1$ parameter. In comparison with the results for the synthetic library of galaxy | columnwidth"}![Estimation perfOrmance of iNput Aps oF thE PéGASe modEls. For each of thE aPs wE plot the predicted vs. truE AP vaLuES for THe Test sEt. The blUE lINE inDiCaTes ThE LiNe of pErfEct estiMation, whilE thE rEd line represENtS the best liNeaR fitting for tHe pOints. THe SumMAry erRorS are gIven in tAble 7.[]{daTa-label="f02"}](fIg70.PS "fig:"){wiDTh="0.3\columNWIdTh"}![EsTimation performanCE oF Input APs of the PéGASE mOdELs. fOR eaCh oF the APs we pLoT the pREdicted VS. tRUE aP vALues for the tesT set. The blue LIne IndicaTeS thE Line of PerfeCt EStiMation, while The rEd line repResentS The best LInear fiTting fOr tHe pOintS. thE sUmmArY ErrORs Are GIveN in Table 7.[]{DaTa-Label="F02"}](fig71.PS "FIG:"){widTh="0.3\cOlumNwidtH"}\
![Estimation peRfoRmanCE of Input aPs of The PéGaSE moDels. FoR each Of The APs we plot the PredIcted vs. trUe Ap vAluEs For thE Test seT. ThE blUe line iNdicateS The LiNE OF pErfect estimation, whIlE THe Red line rEpreseNTs ThE Best lineAr FitTing FOR the pOintS. thE summary Errors ARe GiVen in TaBlE 7.[]{data-lAbEl="f02"}](Fig72.Ps "fig:"){WIdth="0.3\ColumnWidth"}![EstImatiON performance of INput APs of the PégAse MoDEls. FOr eAch of the APs We plOT the PredICtEd vS. True Ap valuEs FOr THe test set. The blue linE iNdicatEs the Line of perfect Estimation, WHILe the red Line REpREsents the best lInear Fitting for THe points. the suMmary errOrs are givEN In Table 7.[]{dAta-LabEl="f02"}](Fig73.PS "FiG:"){width="0.3\columnwIDTh"}![EsTiMation pErfOrmance Of iNpuT APS of ThE PÉGASE moDels. For eAcH oF tHe aPs We ploT The prediCtEd vS. tRue aP valUEs for tHe tesT set. thE bLUe lIne indiCAtES The lInE oF perFecT eStimaTion, WHilE the red Line repreSenTS the BeSt Linear fItting for the pOiNts. The summArY erRors arE GIven in TaBle 7.[]{data-label="f02"}](fig74.ps "fig:"){wiDTh="0.3\columNwiDth"}
FrOm taBle \[t02\] and fiGurE \[f02\] we seE thAT the reSults aRe quiTe GooD ONly foR THe CasE oF early type GALaxIes anD eSpecIally foR the $p_1$ parameter. In coMParIson with the reSulTs deRIVeD foR ThE SynThETic LIBrary of galaxy | columnwidth"}![Estimationperformanc e ofinp utAP s of the PÉGASE models . For each of the APs we pl ot th ep redi c te d vs. true A P v a l ues f or th et es t set . T he blue line indi cat es the line of pe rfect esti mat ion, while t hered li ne re p resen tsthe b est li n ear fi tting for t h e poin t s. Thes u mm aryerrors are giveni nT able 7.[]{data -label =" f 02 " } ](f ig7 0.ps "fig: ") {widt h ="0.3\c o lu m n w idt h "}![Estimatio n performan c e o f inpu tAPs of the PÉGA SE mod els. For ea ch o f the APs we pl o t the p r edicted vs. t rue AP val u es f orth e te s tset . Th e blue l in eindic ates t h e lin e o f pe rfect estimation,whi le t h e r ed li ne re pres en ts th e best line ar fitting for th e po ints. The su mm ary e rrors are gi ven in Table7.[]{da t a-l ab e l = "f 02"}](fig71.ps "fi g: " ) {w idth="0. 3\colu m nw id t h"}\
![E st ima tion p erfor manc e o f inputAPs of th ePÉGASEmo dels.Fo r e ach of t h e AP s we p lot thepredi c ted vs. true A P values for t h et e st set. Th e blue line ind i cate s th e l ine of pe rfect e s ti m ation, while the re dline r epres ents the best linear fi t t i ng for t he p o in t s. The summary erro rs are giv e n in Tab le 7. []{data- label="f0 2 " }](fig72 .ps "f ig: "){ w i dt h="0.3\column w i dth" }! [Estima tio n perfo rma nce of in pu t APs ofthe PÉGA SE m od el s.For e a ch of th eAPs w e p lot t h e pred icted vs. t ru e AP values fo r thete st set . T he blue lin e in dicates the line of perf ec testimat ion, while th ered line r ep res ents t h e best li near fitting for the po i nts. Th e s ummar y er rors aregiv en inTab l e 7.[] {data- label =" f02 " } ](fig 7 3 .p s " fi g:"){width = " 0.3 \colu mn widt h"}![Es timation performan c e o f input APs o f t he P É G AS E m o de l s.Fo r ea c h of the APs weplot the p re d ic ted vs. tr u e A Pvaluesfor the test set. Th e blue li ne indica te s th e lin e of perfe ct estim ation, wh i le th e r ed li nerepres en tsthe b est li n ear fitt ing fo rthe po ints. T he summa ry errors are given inTable7.[]{ dat a-label=" f02 " }]( fig74.ps"fig :"){width= "0. 3\c olumn wid t h"}
From ta ble \[t02 \] a n d figure\ [f 02\ ] we see that t h e res ultsare quitegood only for the cas e of early type gal a x ies an d esp ec ially for the$p_ 1$ p arameter .In comparis on withth e resu lts de rivedfor the s yn t heticlibr ary of galax y | columnwidth"}![Estimation performance_of input_APs of the PÉGASE_models. For_each_of the_APs_we plot the_predicted vs. true AP_values for the test_set. The blue_line_indicates the line of perfect estimation, while the red line represents the best linear_fitting_for the_points._The_summary errors are given in_Table 7.[]{data-label="f02"}](fig70.ps "fig:"){width="0.3\columnwidth"}![Estimation performance of_input APs_of the PÉGASE models. For each of the_APs_we plot the_predicted vs. true AP values for the test set. The_blue line indicates the line of_perfect estimation, while_the_red_line represents the best_linear fitting for the points. The_summary errors are given in Table_7.[]{data-label="f02"}](fig71.ps "fig:"){width="0.3\columnwidth"}\
![Estimation performance of input APs of_the PÉGASE models. For each of_the APs we plot the_predicted vs. true_AP values for the test_set. The blue_line indicates_the line of_perfect estimation, while the red line_represents the best_linear fitting for the points. The_summary_errors are given_in_Table_7.[]{data-label="f02"}](fig72.ps "fig:"){width="0.3\columnwidth"}![Estimation_performance of input_APs_of the_PÉGASE_models. For each of the APs_we_plot the predicted vs. true AP values for_the test set. The_blue_line indicates the line_of perfect estimation, while the_red line represents the best linear_fitting for_the points._The summary errors are given in Table 7.[]{data-label="f02"}](fig73.ps "fig:"){width="0.3\columnwidth"}![Estimation performance of_input APs of the PÉGASE models._For each of the_APs we_plot_the predicted vs. true_AP_values for_the test set. The blue line indicates_the line_of perfect estimation, while the red_line represents the best_linear_fitting for the points. The summary_errors are given in Table 7.[]{data-label="f02"}](fig74.ps_"fig:"){width="0.3\columnwidth"}
From table \[t02\] and figure_\[f02\]_we_see that the results are_quite good only for the case_of early type_galaxies and especially for the $p_1$ parameter._In_comparison with the results derived for_the_synthetic library of galaxy |
+\lambda_j-(\lambda_2+1)<\lambda_j$, by Lemma \[vanishbar\], the term $p_{\lambda_2+1}$ is not a factor of any term of the lowest degree in the power sum expansion of $Q_{(\lambda_i,\lambda_j)}$. Now we are left with the case when $\lambda_i+\lambda_j>\lambda_1-1$ and $\lambda_i\leq \lambda_1-2$. Since $\lambda_i+\lambda_j-(\lambda_1-1)<\lambda_j$, by Lemma \[vanishbar\] the term $p_{\lambda_1-1}$ does not appear as a factor in the terms of the lowest degree of $Q_{(\lambda_i,\lambda_j)}$. So we have shown that if either $p_{\lambda_2+1}$ or $p_{\lambda_1-1}$ appears as a factor of some lowest degree term for $Q_{(\lambda_i,\lambda_j)}$, then we deduce that $\lambda_i=\lambda_1$. Moreover, if both $p_{\lambda_1-1}$ and $p_{\lambda_2+1}$ are factors of the lowest degree terms in the power sum expansion of $Q_{(\lambda_1,\lambda_j)}$, then we have $\lambda_j=\lambda_2$. The proof is complete.
We now present the main result of this paper.
For any $\lambda\in\mathcal{D}(n)$, the terms of the lowest degree in $Q_\lambda$ have degree ${\rm srank}(\lambda)$.
We write the strict partition $\lambda$ in the form $(\lambda_1,\lambda_2,\ldots,\lambda_{2m})$, where $\lambda_1>\ldots>\lambda_{2m}\geq 0$. Suppose that the partition $\lambda$ has $o$ odd parts and $e$ even parts (including $0$ as a part). For the sake of presentation, let $(\lambda_{i_1},\lambda_{i_2},\ldots,\lambda_{i_o})$ denote the sequence of odd parts in decreasing order, and let $(\lambda_{j_1},\lambda_{j_2},\ldots,\lambda_{j_e})$ denote the sequence of even parts in decreasing order.
We first consider the case $o\geq e$. In this case, it will be shown that ${\rm srank}(\lambda)=o$. | + \lambda_j-(\lambda_2 + 1)<\lambda_j$, by Lemma \[vanishbar\ ], the term $ p_{\lambda_2 + 1}$ is not a factor of any term of the lowest academic degree in the exponent sum expansion of $ Q_{(\lambda_i,\lambda_j)}$. Now we are leave with the lawsuit when $ \lambda_i+\lambda_j>\lambda_1 - 1 $ and $ \lambda_i\leq \lambda_1 - 2$. Since $ \lambda_i+\lambda_j-(\lambda_1 - 1)<\lambda_j$, by Lemma \[vanishbar\ ] the term $ p_{\lambda_1 - 1}$ does not appear as a gene in the term of the depleted degree of $ Q_{(\lambda_i,\lambda_j)}$. So we have show that if either $ p_{\lambda_2 + 1}$ or $ p_{\lambda_1 - 1}$ appears as a factor of some low degree term for $ Q_{(\lambda_i,\lambda_j)}$, then we deduce that $ \lambda_i=\lambda_1$. Moreover, if both $ p_{\lambda_1 - 1}$ and $ p_{\lambda_2 + 1}$ are factors of the lowest degree terminus in the power sum expansion of $ Q_{(\lambda_1,\lambda_j)}$, then we have $ \lambda_j=\lambda_2$. The validation is complete.
We now confront the independent result of this newspaper.
For any $ \lambda\in\mathcal{D}(n)$, the terms of the lowest degree in $ Q_\lambda$ have academic degree $ { \rm srank}(\lambda)$.
We write the strict partition $ \lambda$ in the form $ (\lambda_1,\lambda_2,\ldots,\lambda_{2m})$, where $ \lambda_1>\ldots>\lambda_{2m}\geq 0$. Suppose that the partition $ \lambda$ has $ o$ curious parts and $ e$ even parts (including $ 0 $ as a part). For the sake of presentation, let $ (\lambda_{i_1},\lambda_{i_2},\ldots,\lambda_{i_o})$ denote the sequence of odd parts in decrease order, and let $ (\lambda_{j_1},\lambda_{j_2},\ldots,\lambda_{j_e})$ announce the succession of even parts in decreasing order.
We foremost consider the shell $ o\geq e$. In this case, it will be shown that $ { \rm srank}(\lambda)=o$. | +\lamhda_j-(\lambda_2+1)<\lambda_j$, by Lemoa \[vanishbar\], thg rerm $p_{\nambda_2+1}$ is not x factor of any term of the poqest eegree in the power suo expansiln of $Q_{(\lqmbde_i,\lambda_j)}$. Now we are lefb witg the rase when $\lambds_i+\lambda_j>\ldmbda_1-1$ and $\lambga_k\lzq \lambda_1-2$. Since $\lambda_i+\lambda_j-(\lambda_1-1)<\jambda_j$, bj Lemma \[vanishfar\] uhe terj $p_{\lambda_1-1}$ does not appear as a facfor in uhe terms of the lpwest degree of $Q_{(\lambda_i,\lalbda_u)}$. So we have shown that if eijger $p_{\lambda_2+1}$ or $p_{\uambda_1-1}$ apptaxs as a facjor of some lowest degree term fur $Q_{(\lcmbda_i,\lambdq_j)}$, thft we deduce that $\lambda_i=\lambda_1$. Moreoeer, if noth $p_{\lambda_1-1}$ akd $p_{\lembdq_2+1}$ are factors of the nowest degree termf in the [ocer sum expansion of $W_{(\lqmbda_1,\nambga_j)}$, gyen we heve $\lambdw_j=\lembda_2$. The pdoof is complete.
We now present trv main result of thys paper.
For any $\lambda\in\mathcal{D}(n)$, the terks kf the lowest degree in $Q_\lambda$ have degree ${\tm srank}(\laibda)$.
We write the strict partition $\lambda$ in the fmrm $(\lemcda_1,\oaicea_2,\pdots,\lambda_{2m})$, where $\lambda_1>\ldots>\lambda_{2m}\geq 0$. Sup[kst tmat the partitiok $\lambda$ has $o$ odd pwryf and $e$ even karts (iudlhding $0$ as a part). Flr the fake if presenuatiom, let $(\lambda_{i_1},\lambda_{i_2},\ldots,\lanbda_{i_o})$ denotv thw sequence of odd 'arts in decxeasinb ordrr, and let $(\lambda_{j_1},\lambdc_{j_2},\ldota,\lambda_{j_e})$ dfnote the requence of even pagts hn decreasing order.
We firse considec the case $o\eeq g$. In thys case, it will be shown that ${\rm sgank}(\lcmbda)=m$. | +\lambda_j-(\lambda_2+1)<\lambda_j$, by Lemma \[vanishbar\], the term $p_{\lambda_2+1}$ a of any of the lowest expansion $Q_{(\lambda_i,\lambda_j)}$. Now we left with the when $\lambda_i+\lambda_j>\lambda_1-1$ and $\lambda_i\leq \lambda_1-2$. Since by Lemma \[vanishbar\] the term $p_{\lambda_1-1}$ does not appear as a factor in terms of the lowest degree of $Q_{(\lambda_i,\lambda_j)}$. So we have shown that if $p_{\lambda_2+1}$ $p_{\lambda_1-1}$ as factor of some lowest degree term for $Q_{(\lambda_i,\lambda_j)}$, then we deduce that $\lambda_i=\lambda_1$. Moreover, if both and $p_{\lambda_2+1}$ are factors of the lowest degree in the power sum of $Q_{(\lambda_1,\lambda_j)}$, then we have The is complete. now the result of this For any $\lambda\in\mathcal{D}(n)$, the terms of the lowest degree in $Q_\lambda$ have degree ${\rm srank}(\lambda)$. We write strict partition the form where 0$. that the partition $o$ odd parts and $e$ even as a part). For the sake of presentation, $(\lambda_{i_1},\lambda_{i_2},\ldots,\lambda_{i_o})$ denote sequence of odd parts in decreasing and let $(\lambda_{j_1},\lambda_{j_2},\ldots,\lambda_{j_e})$ denote the sequence of even in decreasing order. We first consider the case $o\geq e$. In this case, it will that ${\rm srank}(\lambda)=o$. | +\lambda_j-(\lambda_2+1)<\lambda_j$, by LemMa \[vanishbaR\], the tErm $P_{\laMbDa_2+1}$ is Not a Factor of any terM Of thE lowest degree in the poweR sum eXpANsioN Of $q_{(\lambDa_i,\lambDA_j)}$. nOW we ArE lEft WiTH tHe casE whEn $\lambdA_i+\lambda_j>\lAmbDa_1-1$ And $\lambda_i\leQ \LaMbda_1-2$. Since $\lAmbDa_i+\lambda_j-(\laMbdA_1-1)<\lambdA_j$, By LEMma \[vaNisHbar\] tHe term $P_{\Lambda_1-1}$ Does not apPeAR as a faCTor in thE TErMs of The lowest degree of $q_{(\LaMBda_i,\lambda_j)}$. So wE have sHoWN tHAT if EitHer $p_{\lambda_2+1}$ Or $P_{\lambDA_1-1}$ appearS As A FACtoR Of some lowest dEgree term foR $q_{(\laMbda_i,\lAmBda_J)}$, Then we DeducE tHAt $\lAmbda_i=\lambdA_1$. MorEover, if boTh $p_{\lamBDa_1-1}$ and $p_{\lAMbda_2+1}$ are FactorS of The LoweST dEgRee TeRMs iN ThE poWEr sUm expansIoN oF $Q_{(\lamBda_1,\lAMBDA_j)}$, thEn wE havE $\lambDa_j=\lambda_2$. The pRooF is cOMplEte.
We Now prEsenT tHe maiN resulT of thIs Paper.
For any $\lambDa\in\Mathcal{D}(n)$, The TeRms Of The loWEst degRee In $Q_\Lambda$ hAve degrEE ${\rm SrANK}(\LaMbda)$.
We write the striCt PARtItion $\lamBda$ in tHE fOrM $(\Lambda_1,\laMbDa_2,\lDots,\LAMbda_{2m})$, WherE $\LaMbda_1>\ldotS>\lambdA_{2M}\gEq 0$. suppose ThAt the pArTitIon $\LambdA$ Has $o$ Odd parTs and $e$ evEn parTS (including $0$ as a pARt). For the sake oF PrESEnTAtioN, leT $(\lambda_{i_1},\lamBda_{i_2},\LDots,\LambDA_{i_O})$ deNOte thE sequEnCE oF Odd parts in decreasinG oRder, anD let $(\lAmbda_{j_1},\lambda_{j_2},\Ldots,\lambdA_{J_E})$ Denote thE seqUEnCE of even parts in DecreAsing order.
wE first coNsideR the case $O\geq e$. In thIS Case, it wiLl bE shOwn ThaT ${\RM sRank}(\lambda)=o$. | +\lambda_j-(\lambda_2+1)<\ lambda_j$, by L emm a \ [v anis hbar \], the term $ p _{\l ambda_2+1}$ is not a f actor o f any te rm of the lo w es t deg re einth e p owersum expans ion of $Q_ {(\ la mbda_i,\lamb d a_ j)}$. Nowweare left wit h t he cas ewhe n $\la mbd a_i+\ lambda _ j>\lam bda_1-1$an d $\lam b da_i\le q \l ambd a_1-2$. Since $\l a mb d a_i+\lambda_j- (\lamb da _ 1- 1 ) <\l amb da_j$, byLe mma \ [ vanishb a r\ ] t het erm $p_{\lamb da_1-1}$ do e s n ot app ea r a s a fac tor i nt heterms of th e lo west degr ee of$ Q_{(\la m bda_i,\ lambda _j) }$. Sow eha vesh o wnt ha t i f ei ther $p_ {\ la mbda_ 2+1} $ o r $p_ {\l ambd a_1-1 }$ appears as afact o r o f som e low estde greeterm f or $Q _{ (\lambda_i,\lam bda_ j)}$, the n w eded uc e tha t $\lam bda _i= \lambda _1$. Mo r eov er , i fboth $p_{\lambda_1 -1 } $ a nd $p_{\ lambda _ 2+ 1} $ are fac to rsof t h e lowe st d e gr ee terms in th e p ow er sumex pansio nof$Q_ {(\la m bda_ 1,\lam bda_j)}$ , the n we have $\lam b da_j=\lambda_ 2 $. T he proo f i s complete.
We nowpres e nt th e main resu lt of this paper.
For an y$\lamb da\in \mathcal{D}(n )$, the te r m s of thelowe s td egree in $Q_\l ambda $ have deg r ee ${\rm sran k}(\lamb da)$.
We w rite the st ric t p art i t io n $\lambda$ i n thefo rm $(\l amb da_1,\l amb da_ 2,\ ldo ts ,\lambda_ {2m})$,wh er e$\ lam bda_1 > \ldots>\ la mbd a_ {2m }\geq 0$. Su ppose tha tth e pa rtition $\ l a mbda $ha s $o $ o dd part s an d $e $ evenparts (in clu d ing$0 $as a pa rt). For thesa ke of pres en tat ion, l e t $(\lamb da_{i_1},\lambda_{i_2}, \ ldots,\ lam bda_{ i_o} )$ denote th e sequ enc e of od d part s inde cre a s ing o r d er , a nd let $(\la m b da_ {j_1} ,\ lamb da_{j_2 },\ldots,\lambda_{ j _e} )$ denote the se quen c e o f e v en par ts ind e creasing order.
We first c o ns ider the c a se$o \geq e$ . In th is ca s e, it w ill be sh own that${ \rms r ank }(\lambda) =o$. | +\lambda_j-(\lambda_2+1)<\lambda_j$, by_Lemma \[vanishbar\],_the term $p_{\lambda_2+1}$ is_not a_factor_of any_term_of the lowest_degree in the_power sum expansion of_$Q_{(\lambda_i,\lambda_j)}$. Now we_are_left with the case when $\lambda_i+\lambda_j>\lambda_1-1$ and $\lambda_i\leq \lambda_1-2$. Since $\lambda_i+\lambda_j-(\lambda_1-1)<\lambda_j$, by Lemma \[vanishbar\]_the_term $p_{\lambda_1-1}$_does_not_appear as a factor in_the terms of the lowest_degree of_$Q_{(\lambda_i,\lambda_j)}$. So we have shown that if either_$p_{\lambda_2+1}$_or $p_{\lambda_1-1}$ appears_as a factor of some lowest degree term for_$Q_{(\lambda_i,\lambda_j)}$, then we deduce that $\lambda_i=\lambda_1$._Moreover, if both_$p_{\lambda_1-1}$_and_$p_{\lambda_2+1}$ are factors of_the lowest degree terms in the_power sum expansion of $Q_{(\lambda_1,\lambda_j)}$, then_we have $\lambda_j=\lambda_2$. The proof is complete.
We_now present the main result of_this paper.
For any $\lambda\in\mathcal{D}(n)$, the_terms of_the lowest degree in $Q_\lambda$_have degree ${\rm_srank}(\lambda)$.
We write_the strict partition_$\lambda$ in the form $(\lambda_1,\lambda_2,\ldots,\lambda_{2m})$, where_$\lambda_1>\ldots>\lambda_{2m}\geq 0$. Suppose_that the partition $\lambda$ has $o$_odd_parts and $e$_even_parts_(including $0$_as a part)._For_the sake_of_presentation, let $(\lambda_{i_1},\lambda_{i_2},\ldots,\lambda_{i_o})$ denote the sequence_of_odd parts in decreasing order, and let_$(\lambda_{j_1},\lambda_{j_2},\ldots,\lambda_{j_e})$ denote the sequence_of_even parts in decreasing_order.
We first consider the case_$o\geq e$. In this case, it_will be_shown that_${\rm srank}(\lambda)=o$. |
use in CCTV applications[^3], which provides a bias and clock for the sensor. The headboard reads the digital image values and sends them to the FPGA for further processing. It also incorporates a synchronous serial port, which may be used to modify the internal camera registers when needed. The pixel stream and the three sync signals for the pixel, line, and frame clocks are generated by the CMOS sensor and are connected directly to the FPGA inputs. The master input clock for the CMOS chip is generated by the on-chip crystal oscillator and PLL on the FPGA board. The oscillator frequency is fixed at 12 MHz, but the input clock frequency for the CMOS chip may be varied from 6 to 27 MHz using the programmable PLLs on the FPGA. The successive pixel values are read out as a 2D image matrix and stored in an array on the FPGA memory. The limiting factor in the frame rate is the CMOS frequency, and we are in the process of obtaining and testing faster chips.
Data Acquisition Board
----------------------
We have chosen the XuLA2-LX25 FPGA prototyping board from XESS Corp.[^4] for the design of the main processor since the often used commercial development boards require too many resources (power, weight, complexity). This board meets our minimum requirements in a compact form factor. The heart of the FPGA board is a Spartan^^ 6 series XC6LX25 chip[^5], which controls and reads out the CMOS sensor, processes the data, stores and transmits the output. The use of FPGAs allows us to rapidly prototype different algorithms without changing the hardware which is invaluable in a development environment.
The FPGA modules were written using Verilog and synthesized and programmed using the Xilinx ISE^^ (Integrated Synthesis Environment) Design Suite. In addition, we have used standard VHDL modules[^6] for interfacing the SDRAM (synchronous dynamic random-access memory), PLL (Phase Locked Loop), and micro-SD (Secure Digital) card to the Spartan^^ 6 chip on the FPGA board.
Implementation
==============
The detector is designed to operate in two modes:
- Continuous frame transfer mode.
- Photon counting mode.
We have designed the readout and associated electronics of the detector so that we can switch between both modes with no changes in the hardware. Incoming photons eject electrons from the S20 photocathode which are multiplied | use in CCTV applications[^3 ], which provides a bias and clock for the sensor. The headboard read the digital prototype values and sends them to the FPGA for further processing. It also incorporate a synchronous serial port, which may be use to change the internal camera registers when needed. The pixel stream and the three sync signal for the pixel, line, and frame redstem storksbill are generated by the CMOS sensor and are connected directly to the FPGA inputs. The maestro input clock for the CMOS chip is generated by the on - check crystal oscillator and PLL on the FPGA board. The oscillator frequency is fixed at 12 MHz, but the input signal clock frequency for the CMOS chip may be varied from 6 to 27 MHz using the programmable PLLs on the FPGA. The successive pixel value are read out as a 2D image matrix and stored in an array on the FPGA memory. The limiting factor in the skeletal system pace is the CMOS frequency, and we are in the process of obtaining and testing faster chips.
Data Acquisition Board
----------------------
We have chosen the XuLA2 - LX25 FPGA prototyping board from XESS Corp.[^4 ] for the invention of the chief processor since the often use commercial exploitation boards require excessively many resources (power, weight, complexity). This board meet our minimum requirements in a compact form factor. The heart of the FPGA dining table is a Spartan^^ 6 series XC6LX25 chip[^5 ], which controls and reads out the CMOS sensor, serve the data, shop and transmits the output. The use of FPGAs allows us to rapidly prototype unlike algorithms without changing the hardware which is invaluable in a development environment.
The FPGA modules were written using Verilog and synthesized and programmed using the Xilinx ISE^^ (Integrated Synthesis Environment) Design Suite. In summation, we have used standard VHDL modules[^6 ] for interfacing the SDRAM (synchronous active random - access memory), PLL (Phase Locked Loop), and micro - SD (Secure Digital) card to the Spartan^^ 6 chip on the FPGA board.
Implementation
= = = = = = = = = = = = = =
The detector is plan to operate in two mode:
- Continuous frame transportation mode.
- Photon counting mode.
We have design the readout and associated electronics of the detector so that we can switch between both manner with no changes in the hardware. Incoming photon eject electrons from the S20 photocathode which are multiplied | usf in CCTV applications[^3], wmich provides a yuas anv clock for the sensor. The headboard reads vhe eigitql image values and sevds them no the FPTA fie further 'docessiky. It zpso nnrorporates a sykchronous sarial port, whiwh mcy be used to modify the internal caiera rebidters when neeqed. Uhe pixsl stream and the three sync signams for uhe pixel, line, and frame clocks are generatef by the CMOS sensor ajd are conngdteq directly to the FPGA puputs. The mzster input clock for the CMOS zhip ns generatee vy hve on-chip ccystal oscillator and PLL mn the GPGA board. The osrillqtor frequency is fixxd at 12 MHz, but the igput clocn yrequency for the CMOW xhip kay te vxeiea fdok 6 to 27 MJz nsing the pdogrammable PLLs on the FPGA. Tne wuccessive pisel vajuqs are read out as a 2D image matrix and stkred in an array on the FPGA memory. The limijing factow in the frame rate is the CMOS frequency, and we dre ii ghe pvocerw lf obtaining and testing faster chips.
Data Acqtjsotpon Board
----------------------
We have ghosen the XuLA2-LX25 GPHA krototyping boxrd from SESS Corp.[^4] for the fesign jf thw main prjcesxor since the often used conmercial devvlopnent boards requirz too many rzsourcgs (powrr, weight, complexity). Thns boadd meets oug minimum fequirements in x cpm[act form factor. The heart of the F'GA bpard is a Skartan^^ 6 series XC6PX25 chl[[^5], which controls ajd recds ogt the CMOD sensor, processes the data, stocxs and transmotv tve outpuc. The mse of FPGAs aljows us to rapndly procotype different algorivhms without changing the jardware whirh is invwluavle un a dexdlopment envirpnment.
The FPGA moduoes were written uxine Verilog and syuuhewized and progrsmmdd tspng tra Xilinx ISE^^ (Intdgrxyed Shnthesis Enyirunmemt) Design Suite. In agditjon, we have used syakdard VHDO modulef[^6] for interfaving the SDRAM (synfhronmus dynakic random-access memory), PLL (Phase Locked Llop), and micro-SD (Secmre Qigital) caxd to the Spartan^^ 6 chip on the FPGA boarv.
Implementation
==============
The detgctor is designed to pkerate in two modef:
- Contituous frame transfer mode.
- Photon comnting mode.
We have desjgned dhe rfadout and associated electronics of the detector so that we can switch bwtween both modes sith no cvaugex in tre hervware. Incoming phptons eject electrons from the R20 photocdtkode which are multiplied | use in CCTV applications[^3], which provides a clock the sensor. headboard reads the them the FPGA for processing. It also a synchronous serial port, which may used to modify the internal camera registers when needed. The pixel stream and three sync signals for the pixel, line, and frame clocks are generated by CMOS and connected to the FPGA inputs. The master input clock for the CMOS chip is generated by the crystal oscillator and PLL on the FPGA board. oscillator frequency is fixed 12 MHz, but the input frequency the CMOS may varied 6 to 27 using the programmable PLLs on the FPGA. The successive pixel values are read out as a 2D matrix and an array the memory. limiting factor in rate is the CMOS frequency, and the process of obtaining and testing faster chips. Acquisition Board We have chosen the XuLA2-LX25 FPGA board from XESS Corp.[^4] for the design of main processor since the often used commercial development boards require too many resources (power, weight, board meets our minimum in a compact factor. heart the board is Spartan^^ 6 series XC6LX25 chip[^5], which controls and reads out the sensor, processes the data, stores and transmits the output. The FPGAs us to rapidly different algorithms without changing hardware is invaluable in a The modules Verilog synthesized programmed using the Xilinx (Integrated Synthesis Environment) Design Suite. addition, we have used the SDRAM (synchronous dynamic random-access memory), PLL (Phase Loop), and micro-SD (Secure Digital) card to Spartan^^ 6 chip on the FPGA board. Implementation ============== The detector is to operate modes: - Continuous frame transfer mode. - Photon mode. We have designed readout and associated electronics of the detector so that can between both with no changes the hardware. Incoming eject electrons from photocathode which multiplied | use in CCTV applications[^3], whicH provides a Bias aNd cLocK fOr thE senSor. The headboarD ReadS the digital image values And seNdS Them TO tHe FPGa for furTHeR PRocEsSiNg. IT aLSo IncorPorAtes a syNchronous sEriAl Port, which may BE uSed to modifY thE internal camEra RegistErS whEN needEd. THe pixEl streAM and thE three synC sIGnals fOR the pixEL, LiNe, anD frame clocks are geNErATed by the CMOS seNsor anD aRE cONNecTed Directly to ThE FPGA INputs. ThE MaSTER inPUt clock for the cMOS chip is gENerAted by ThE on-CHip cryStal oScILlaTor and PLL on The FpGA board. THe osciLLator frEQuency iS fixed At 12 Mhz, bUt thE InPuT clOcK FreQUeNcy FOr tHe CMOS chIp MaY be vaRied FROM 6 To 27 MHZ usIng tHe proGrammable PLLs On tHe FPga. ThE succEssivE pixEl ValueS are reAd out As A 2D image matrix anD stoRed in an arRay On The fPgA memORy. The lImiTinG factor In the frAMe rAtE IS ThE CMOS frequency, and wE aRE In The proceSs of obTAiNiNG and testInG faSter CHIps.
DaTa AcQUiSition BoArd
----------------------
We hAVe ChOsen the xulA2-LX25 FPgA ProTotYping BOard From XEsS Corp.[^4] foR the dESign of the main pROcessor since tHE oFTEn USed cOmmErcial develOpmeNT boaRds rEQuIre TOo manY resoUrCEs (POwer, weight, complexitY). THis boaRd meeTs our minimum rEquirementS IN A compact Form FAcTOr. The heart of thE FPGA Board is a SpARtan^^ 6 seriEs XC6Lx25 chip[^5], whiCh controlS ANd reads oUt tHe CmOS SenSOR, pRocesses the daTA, StorEs And tranSmiTs the ouTpuT. ThE usE of fPgAs allows Us to rapiDlY pRoToTypE diffERent algoRiThmS wIthOut chANging tHe harDwarE wHiCH is InvaluaBLe IN A devElOpMent EnvIrOnmenT.
The fpGA Modules Were writtEn uSIng VErIlOg and syNthesized and pRoGrammed usiNg The xilinx ise^^ (IntegraTed Synthesis Environment) dEsign SuIte. in addItioN, we have usEd sTandarD VHdl modulEs[^6] for iNterfAcIng THE SDRAm (SYnChrOnOus dynamic RANdoM-acceSs MemoRy), PLL (PhAse Locked Loop), and miCRo-Sd (Secure DigitaL) caRd to THE SParTAn^^ 6 CHip On THe Fpga board.
ImplementAtion
==============
The deTeCToR is designeD To oPeRate in tWo modes:
- contiNUous fraMe transfeR mode.
- PhotOn CounTINg mOde.
We have dEsigned tHe readout ANd assOCiAted eLecTronicS oF thE deteCtor so THat We can Switch BeTween bOth moDeS with no cHanges in the hardware. IncoMing phOtons EjeCt electroNs fROm tHe S20 photocAthoDe which are MulTipLied | use in CCTV applications[ ^3], which prov ide s a b iasandclock for thes enso r. The headboard reads thedi g ital im age v alues a n ds e nds t he m t ot he FPGA fo r furth er process ing .It also inco r po rates a sy nch ronous seria l p ort, w hi chm ay be us ed to modif y the i nternal c am e ra reg i sters w h e nneed ed. The pixel str e am and the threesync s ig n al s for th e pixel, l in e, an d framec lo c k s ar e generated by the CMOS s e nso r andar e c o nnecte d dir ec t lyto the FPGA inp uts. Themaster input c l ock for the C MOS ch ip i s g en era te d by th e o n -ch ip cryst al o scill ator a n d PLL on the FPGA board. The o sci llat o r f reque ncy i s fi xe d at12 MHz , but t he input clockfreq uency for th eCMO Schipm ay bevar ied from 6 to 27M Hzus i n g t he programmable PL Ls o nthe FPGA . Thes uc ce s sive pix el va lues a re re ad o u tas a 2Dimagem at ri x and s to red in a n a rra y ont he F PGA me mory. Th e lim i ting factor in the frame rat e i s th e CMO S f requency, a nd w e are int he pr o cessof ob ta i ni n g and testing faste rchips.
Dat a Acquisition Board
--- - - - -------- ---- - -- -
We have chos en th e XuLA2-LX 2 5 FPGA p rotot yping bo ard fromX E SS Corp. [^4 ] f orthe d es ign of the ma i n pro ce ssor si nce the of ten us edcom me rcial dev elopment b oa rd sreq uiret oo manyre sou rc es(powe r , weig ht, c ompl ex it y ).This bo a rd m eets o ur min imu mrequi reme n tsin a co mpact for m f a ctor .Th e heart of the FPGAbo ard is a S pa rta n^^ 6s e ries XC6 LX25 chip[^5], which co n trols a ndreads out the CMOS se nsor,pro c essesthe da ta, s to res a nd tr a n sm its t he output. T heuse o fFPGA s allow s us to rapidly pr o tot ype different al gori t h ms wi t ho u t c ha n gin g the hardware wh ich is inv al u ab le in a de v elo pm ent env ironmen t.
T h e FPGAmodules w ere writt en usi n g Ve rilog andsynthesi zed and p r ogram m ed usin g t he Xil in x I SE^^(Integ r ate d Syn thesis E nviron ment) D esign Su ite. In addition, we ha ve use d sta nda rd VHDL m odu l es[ ^6] for i nter facing the SD RAM (syn chr o nousdyna m ic ra n dom-a cces s memory), PL L ( P h as e Locked Lo o p ) , a nd mi cro - SD (Se cure Digital) card to the Spartan^^6 ch i p on th e FPG Aboard.
Implem ent at i o n
====== == ======
The detecto ri s des ignedto ope rate in t wo modes:
- C ontinuous fr am e transf er m o de.
- P ho ton co unting mode .
We have designe d the r eadou t an d ass oc iated e l ectr onics of t he detector so th at w e can switch b etween bo th modes wit h no chang es in the ha rd ware . I ncomin g ph o t ons e ject e lec trons fro m th e S 20 pho toca thode w hich are mult i plied | use_in CCTV_applications[^3], which provides a_bias and_clock_for the_sensor._The headboard reads_the digital image_values and sends them_to the FPGA_for_further processing. It also incorporates a synchronous serial port, which may be used to_modify_the internal_camera_registers_when needed. The pixel stream_and the three sync signals_for the_pixel, line, and frame clocks are generated by_the_CMOS sensor and_are connected directly to the FPGA inputs. The master_input clock for the CMOS chip_is generated by_the_on-chip_crystal oscillator and PLL_on the FPGA board. The oscillator_frequency is fixed at 12 MHz,_but the input clock frequency for the_CMOS chip may be varied from_6 to 27 MHz using_the programmable_PLLs on the FPGA. The_successive pixel values_are read_out as a_2D image matrix and stored in_an array on_the FPGA memory. The limiting factor_in_the frame rate_is_the_CMOS frequency,_and we are_in_the process_of_obtaining and testing faster chips.
Data Acquisition_Board
----------------------
We_have chosen the XuLA2-LX25 FPGA prototyping board_from XESS Corp.[^4] for_the_design of the main_processor since the often used_commercial development boards require too many_resources (power,_weight, complexity)._This board meets our minimum requirements in a compact form factor._The heart of the FPGA board_is a Spartan^^ 6_series XC6LX25_chip[^5],_which controls and_reads_out the_CMOS sensor, processes the data, stores and_transmits the_output. The use of FPGAs allows_us to rapidly prototype_different_algorithms without changing the hardware which_is invaluable in a development environment.
The_FPGA modules were written using_Verilog_and_synthesized and programmed using the_Xilinx ISE^^ (Integrated Synthesis Environment) Design_Suite. In addition,_we have used standard VHDL modules[^6] for_interfacing_the SDRAM (synchronous dynamic random-access memory),_PLL_(Phase Locked Loop), and micro-SD (Secure_Digital)_card_to the Spartan^^ 6 chip_on the FPGA board.
Implementation
==============
The detector is_designed to operate in two modes:
- Continuous_frame transfer mode.
-_ Photon counting mode.
We_have_designed_the readout and associated electronics of the detector so that_we can_switch between both_modes with no changes in the hardware. Incoming photons eject_electrons from the S20 photocathode which are_multiplied |
reparametrized by $\theta : [0,1] \rightarrow [0,n]$. The reparametrized curve is denoted by $Q_\theta$.
#### [Fréchet matching]{}s
We are given two polygonal curves $P$ and $Q$ with $m$ and $n$ edges. A (monotonous) *matching* $\mu$ between $P$ and $Q$ is a pair of reparametrizations $(\sigma,\theta)$, such that $P_\sigma(t)$ matches to $Q_\theta(t)$. The Euclidean distance between two matched points is denoted by $d_\mu(t) = |P_\sigma(t) - Q_\theta(t)|$. The maximum distance over a range is denoted by $d_\mu[a,b] = \max_{a \leq t \leq b} d_\mu(t)$. The *[Fréchet distance]{}* between two curves is defined as $\delta_\text{F}(P,Q) = \inf_{\mu} d_\mu[0,1]$. A *[Fréchet matching]{}* is a matching $\mu$ that realizes the [Fréchet distance]{}: $d_\mu[0,1] = \delta_\text{F}(P,Q)$ holds.
#### Free space diagrams
Alt and Godau [@Alt1995] describe an algorithm to compute the [Fréchet distance]{} based on the decision variant (that is, solving $\delta_\text{F}(P,Q) \leq \varepsilon$ for some given $\varepsilon$). Their algorithm uses a *free space diagram*, a two-dimensional diagram on the range $[0, m] \times [0, n]$. Every point $(x,y)$ in this diagram is either “free” (white) or not (indicating whether $|P(x) - Q(y)| \leq \varepsilon$). The diagram has $m$ columns and $n$ rows; every cell $(c,r)$ ($1 \leq c \leq m$ and $1 \leq r \leq n$) corresponds to the edges $p_{c-1}p_{c}$ and $q_{r-1}q_{r}$. To compute the [Fréchet distance]{}, one finds the smallest $\varepsilon$ such that there exists an x- and y-monotone path from point $(0,0 | reparametrized by $ \theta : [ 0,1 ] \rightarrow [ 0,n]$. The reparametrized curve is denoted by $ Q_\theta$.
# # # # [ Fréchet matching]{}s
We are given two polygonal curves $ P$ and $ Q$ with $ m$ and $ n$ edge. A (humdrum) * matching * $ \mu$ between $ P$ and $ Q$ is a pair of reparametrizations $ (\sigma,\theta)$, such that $ P_\sigma(t)$ matches to $ Q_\theta(t)$. The euclidian distance between two matched points is denoted by $ d_\mu(t) = |P_\sigma(t) - Q_\theta(t)|$. The maximal distance over a stove is denoted by $ d_\mu[a, b ] = \max_{a \leq thymine \leq b } d_\mu(t)$. The * [ Fréchet distance ] { } * between two curves is defined as $ \delta_\text{F}(P, Q) = \inf_{\mu } d_\mu[0,1]$. A * [ Fréchet matching ] { } * is a coordinated $ \mu$ that realizes the [ Fréchet distance ] { }: $ d_\mu[0,1 ] = \delta_\text{F}(P, Q)$ holds.
# # # # detached quad diagrams
Alt and Godau [ @Alt1995 ] describe an algorithm to compute the [ Fréchet distance ] { } based on the decision discrepancy (that is, solving $ \delta_\text{F}(P, Q) \leq \varepsilon$ for some given $ \varepsilon$). Their algorithm uses a * barren space diagram *, a two - dimensional diagram on the range $ [ 0, m ] \times [ 0, n]$. Every point $ (x, y)$ in this diagram is either “ free ” (blank) or not (indicating whether $ |P(x) - Q(y)| \leq \varepsilon$). The diagram has $ m$ columns and $ n$ rows; every cell $ (speed of light, r)$ ($ 1 \leq c \leq m$ and $ 1 \leq r \leq n$) corresponds to the edges $ p_{c-1}p_{c}$ and $ q_{r-1}q_{r}$. To compute the [ Fréchet distance ] { }, one line up the smallest $ \varepsilon$ such that there exists an x- and y - monotone path from point $ (0,0 | reoarametrized by $\theta : [0,1] \rightarrow [0,n]$. Tkw repacametriaed curvd is denoted by $Q_\theta$.
#### [Fréchev marchint]{}s
We are given two polhgonal cugves $P$ ane $Q$ xith $m$ and $n$ edgxa. A (monotonoua) *matehmng* $\mu$ between $K$ and $Q$ is a pair of repardmdtxizations $(\sigma,\theta)$, such that $P_\sigmw(t)$ matcned to $Q_\theta(t)$. Tre Elcjidezn distance between two matched pojnts is denoted by $d_\ku(t) = |P_\sigma(t) - Q_\theta(t)|$. The laxilum distance over w range is eenoewd by $d_\mu[a,b] = \max_{a \leq u \neq b} d_\mu(t)$. The *[Fréchet distance]{}* between two curvzs is defingb as $\gelta_\text{F}(P,W) = \ins_{\mu} d_\mu[0,1]$. A *[Frégnet madching]{}* ix a matching $\mm$ thav reqlizes the [Fréchet disvance]{}: $d_\mu[0,1] = \delta_\text{S}(P,Q)$ holds.
#### Fxee space diagrams
Alt abd Gogau [@Ant1995] ddwcrkbe ai amgoritjm vo compute fhe [Fréchet eistance]{} based on tht dqbosion varianf (that if, solving $\delta_\text{F}(P,Q) \leq \varepsilon$ fmr aome given $\varepsilon$). Tyeir algorithm uses a *free spase diagram*, a two-dimensional diagram on the range $[0, m] \tijds [0, n]$. Dcegy point $(x,y)$ in this diagram is either “free” (whyfe) og not (indicating chether $|P(x) - Q(y)| \lrq \vstepsilon$). The dkagram haa $m$ columns and $n$ gows; evgry ceol $(c,r)$ ($1 \lez c \keq m$ and $1 \leq r \leq n$) corrwsponds to tke wdges $p_{c-1}p_{c}$ and $q_{r-1}q_{x}$. To compute the [Fréchrt distance]{}, one finds tke smamlest $\varepdilon$ sucg that there exisgs sn x- and y-monotone path from point $(0,0 | reparametrized by $\theta : [0,1] \rightarrow [0,n]$. curve denoted by #### [Fréchet matching]{}s curves and $Q$ with and $n$ edges. (monotonous) *matching* $\mu$ between $P$ and is a pair of reparametrizations $(\sigma,\theta)$, such that $P_\sigma(t)$ matches to $Q_\theta(t)$. The distance between two matched points is denoted by $d_\mu(t) = |P_\sigma(t) - Q_\theta(t)|$. maximum over range denoted by $d_\mu[a,b] = \max_{a \leq t \leq b} d_\mu(t)$. The *[Fréchet distance]{}* between two curves defined as $\delta_\text{F}(P,Q) = \inf_{\mu} d_\mu[0,1]$. A *[Fréchet is a matching $\mu$ realizes the [Fréchet distance]{}: $d_\mu[0,1] \delta_\text{F}(P,Q)$ #### Free diagrams and [@Alt1995] describe an to compute the [Fréchet distance]{} based on the decision variant (that is, solving $\delta_\text{F}(P,Q) \leq \varepsilon$ for given $\varepsilon$). uses a space a diagram on the m] \times [0, n]$. Every point diagram is either “free” (white) or not (indicating $|P(x) - \leq \varepsilon$). The diagram has $m$ and $n$ rows; every cell $(c,r)$ ($1 \leq \leq m$ and $1 \leq r \leq n$) corresponds to the edges $p_{c-1}p_{c}$ and $q_{r-1}q_{r}$. the [Fréchet distance]{}, one the smallest $\varepsilon$ that exists x- y-monotone path point $(0,0 | reparametrized by $\theta : [0,1] \righTarrow [0,n]$. The ReparAmeTriZeD curVe is Denoted by $Q_\thetA$.
#### [frécHet matching]{}s
We are given Two poLyGOnal CUrVes $P$ aNd $Q$ with $M$ AnD $N$ EdgEs. a (mOnoToNOuS) *matcHinG* $\mu$ betwEen $P$ and $Q$ is A paIr Of reparametrIZaTions $(\sigma,\TheTa)$, such that $P_\sIgmA(t)$ matcHeS to $q_\Theta(T)$. ThE EuclIdean dIStance Between twO mATched pOInts is dENOtEd by $D_\mu(t) = |P_\sigma(t) - Q_\theta(T)|$. thE Maximum distancE over a RaNGe IS DenOteD by $d_\mu[a,b] = \maX_{a \Leq t \lEQ b} d_\mu(t)$. THE *[FRÉCHet DIstance]{}* betweeN two curves iS DefIned as $\DeLta_\TExt{F}(P,Q) = \Inf_{\mu} D_\mU[0,1]$. a *[FrÉchet matchiNg]{}* is A matching $\Mu$ that REalizes THe [FréchEt distAncE]{}: $d_\mU[0,1] = \delTA_\tExT{F}(P,q)$ hOLds.
#### fReE spACe dIagrams
ALt AnD GodaU [@Alt1995] DESCRibe An aLgorIthm tO compute the [FrÉchEt diSTanCe]{} basEd on tHe deCiSion vAriant (That iS, sOlving $\delta_\text{f}(P,Q) \lEq \varepsiLon$ FoR soMe Given $\VArepsiLon$). theIr algorIthm useS A *frEe SPACe Diagram*, a two-dimensiOnAL DiAgram on tHe rangE $[0, M] \tImES [0, n]$. Every pOiNt $(x,Y)$ in tHIS diagRam iS EiTher “free” (White) oR NoT (iNdicatiNg WhetheR $|P(X) - Q(y)| \Leq \VarepSIlon$). the diaGram has $m$ ColumNS and $n$ rows; every CEll $(c,r)$ ($1 \leq c \leq m$ ANd $1 \LEQ r \LEq n$) cOrrEsponds to thE edgES $p_{c-1}p_{C}$ and $Q_{R-1}q_{R}$. To COmputE the [FRéCHeT Distance]{}, one finds the SmAllest $\VarepSilon$ such that There existS AN X- and y-monOtonE PaTH from point $(0,0 | reparametrized by $\theta : [0,1]\righ tar row [ 0,n] $. T he reparametri z ed c urve is denoted by $Q_ \thet a$ .
## # #[Fréc het mat c hi n g ]{} s
W e a re gi ven t wopolygon al curves$P$ a nd $Q$ with$ m$ and $n$ e dge s. A (monoto nou s) *ma tc hin g * $\m u$betwe en $P$ and $Q $ is a pa ir of rep a rametri z a ti ons$(\sigma,\theta)$ , s u ch that $P_\si gma(t) $m at c h esto$Q_\theta( t) $. Th e Euclid e an d i sta n ce between tw o matched p o int s is d en ote d by $d _\mu( t) = | P_\sigma(t) - Q _\theta(t )|$. T h e maxim u m dista nce ov era r ange is d eno te d by $d _\m u [a, b] = \ma x_ {a \leq t \ l e q b} d _\m u(t) $. Th e *[Fréchet d ist ance ] {}* betw een t wo c ur ves i s defi ned a s$\delta_\text{F }(P, Q) = \inf _{\ mu } d _\ mu[0, 1 ]$. A*[F réc het mat ching]{ } * i sa m at ching $\mu$ that r ea l i ze s the [F réchet di st a nce]{}:$d _\m u[0, 1 ] = \d elta _ \t ext{F}(P ,Q)$ h o ld s.
####Fr ee spa ce di agr ams
A lt a nd God au [@Alt 1995] describe an al g orithm to com p ut e th e [Fr éch et distance ]{}b ased ont he de c ision vari an t ( t hat is, solving $\d el ta_\te xt{F} (P,Q) \leq \v arepsilon$ f o r some g iven $\ v arepsilon$). T heiralgorithmu ses a *f ree s pace dia gram*, at w o-dimens ion aldia gra m on the range $[ 0 , m]\t imes [0 , n ]$. Eve rypoi nt$(x ,y )$ in thi s diagra mis e it her “fre e ” (white )orno t ( indic a ting w hethe r $| P( x) - Q (y)| \l e q\ v arep si lo n$). Th ediagr am h a s $ m$ colu mns and $ n$r ows; e ve ry cell $(c,r)$ ($1\l eq c \leqm$ an d $1 \ l e q r \leq n$) corresponds to the edges $ p_{ c-1}p _{c} $ and $q_ {r- 1}q_{r }$. To com pute t he [F ré che t dista n c e] {}, o ne finds t h e sm alles t$\va repsilo n$ such that there exi sts an x- and y- mono t o ne pa t hf rom p o int $ (0,0 | reparametrized_by $\theta_ : [0,1] \rightarrow_[0,n]$. The_reparametrized_curve is_denoted_by $Q_\theta$.
#### [Fréchet_matching]{}s
We are given_two polygonal curves $P$_and $Q$ with_$m$_and $n$ edges. A (monotonous) *matching* $\mu$ between $P$ and $Q$ is a pair_of_reparametrizations $(\sigma,\theta)$,_such_that_$P_\sigma(t)$ matches to $Q_\theta(t)$. The_Euclidean distance between two matched_points is_denoted by $d_\mu(t) = |P_\sigma(t) - Q_\theta(t)|$. The_maximum_distance over a_range is denoted by $d_\mu[a,b] = \max_{a \leq t_\leq b} d_\mu(t)$. The *[Fréchet distance]{}* between_two curves is_defined_as_$\delta_\text{F}(P,Q) = \inf_{\mu} d_\mu[0,1]$._A *[Fréchet matching]{}* is a matching $\mu$_that realizes the [Fréchet distance]{}: $d_\mu[0,1]_= \delta_\text{F}(P,Q)$ holds.
#### Free space diagrams
Alt and_Godau [@Alt1995] describe an algorithm to compute_the [Fréchet distance]{} based on the_decision variant_(that is, solving $\delta_\text{F}(P,Q) \leq_\varepsilon$ for some_given $\varepsilon$)._Their algorithm uses_a *free space diagram*, a two-dimensional_diagram on the_range $[0, m] \times [0, n]$._Every_point $(x,y)$ in_this_diagram_is either_“free” (white) or_not_(indicating whether_$|P(x)_- Q(y)| \leq \varepsilon$). The diagram_has_$m$ columns and $n$ rows; every cell_$(c,r)$ ($1 \leq c_\leq_m$ and $1 \leq_r \leq n$) corresponds to_the edges $p_{c-1}p_{c}$ and $q_{r-1}q_{r}$. To_compute the_[Fréchet distance]{},_one finds the smallest $\varepsilon$ such that there exists an x-_and y-monotone path from point $(0,0 |
}+e^{1256}+e^{3456} +e^{1367}+e^{1457}+e^{2357}-e^{2467},
\end{split}$$ where $*_{\varphi}$ is the Hodge operator determined by $g_{\varphi}$ and $\operatorname{vol}_{\varphi}$, and $e^{ijk\cdots}$ is a shorthand for the wedge product of covectors $e^i{\wedge}e^j{\wedge}e^k{\wedge}\cdots$. We shall call both $\mathcal{B}$ and $\mathcal{B}^*$ [*adapted*]{} bases to the G$_2$-structure ${\varphi}$.
On the other hand, given an inner product $g$ on $V,$ we can consider a $g$-orthonormal basis $\mathcal{B}$ of $V$ and the G$_2$-structure ${\varphi}\in\Lambda^3_{{\scriptscriptstyle}+}V^*$ having $\mathcal{B}$ as an adapted basis. By, the metric $g_{\varphi}$ induced by ${\varphi}$ coincides with $g$. We shall refer to such ${\varphi}$ as the G$_2$-structure [*induced*]{} by the basis $\mathcal{B}$. Hence, there is a surjective map $$\mathcal{G} : \Lambda^3_{{\scriptscriptstyle}+}V^* \rightarrow \mathcal{S}^2_{{\scriptscriptstyle}+}V^*,\quad \mathcal{G}({\varphi}) = g_{\varphi},$$ which is not injective, as the set of all G$_2$-structures inducing the same metric is parametrized by ${{\mathrm {SO}}}(7)/{{\mathrm G}}_2\cong{\mathbb R}P^7$ (see [@Bry Remark 4] for an explicit description). The G$_2$-structures belonging to $\mathcal{G}^{-1}(g)$ will be called [*compatible*]{} with $g$.
Consider now a G$_2$-structure ${\varphi}$ on $V,$ let $z\in V$ be a unit vector and denote by $W$ the 6-dimensional $g_{\varphi}$-orthogonal complement of $\langle z\rangle\subset V.$ Then, the G$_2$-structure ${\varphi}$ induces an SU(3)-structure $(h,J,\omega,{\psi_{{\scriptscriptstyle}+}},{\psi_{{\scriptscriptstyle}-}})$ on $W$ by means of the identities $${\varphi}= \omega{\wedge}z^\flat + {\psi_{{\scriptscriptstyle}+}},\quad *_{\varphi}{\varphi}= \ | } + e^{1256}+e^{3456 } + e^{1367}+e^{1457}+e^{2357}-e^{2467 },
\end{split}$$ where $ * _ { \varphi}$ is the Hodge operator determined by $ g_{\varphi}$ and $ \operatorname{vol}_{\varphi}$, and $ e^{ijk\cdots}$ is a shorthand for the wedge product of covectors $ e^i{\wedge}e^j{\wedge}e^k{\wedge}\cdots$. We shall visit both $ \mathcal{B}$ and $ \mathcal{B}^*$ [ * adapt * ] { } bases to the G$_2$-structure $ { \varphi}$.
On the other hand, yield an inner product $ g$ on $ V,$ we can regard a $ g$-orthonormal footing $ \mathcal{B}$ of $ V$ and the G$_2$-structure $ { \varphi}\in\Lambda^3_{{\scriptscriptstyle}+}V^*$ having $ \mathcal{B}$ as an adapted basis. By, the metric $ g_{\varphi}$ induced by $ { \varphi}$ coincides with $ g$. We shall denote to such $ { \varphi}$ as the G$_2$-structure [ * induced * ] { } by the basis $ \mathcal{B}$. therefore, there is a surjective map $ $ \mathcal{G }: \Lambda^3_{{\scriptscriptstyle}+}V^ * \rightarrow \mathcal{S}^2_{{\scriptscriptstyle}+}V^*,\quad \mathcal{G}({\varphi }) = g_{\varphi},$$ which is not injective, as the hardening of all G$_2$-structures inducing the same metric function is parametrized by $ { { \mathrm { SO}}}(7)/{{\mathrm G}}_2\cong{\mathbb R}P^7 $ (see [ @Bry Remark 4 ] for an explicit description). The G$_2$-structures belonging to $ \mathcal{G}^{-1}(g)$ will be call [ * compatible * ] { } with $ g$.
Consider now a G$_2$-structure $ { \varphi}$ on $ V,$ let $ z\in V$ be a unit vector and announce by $ W$ the 6 - dimensional $ g_{\varphi}$-orthogonal complement of $ \langle z\rangle\subset V.$ Then, the G$_2$-structure $ { \varphi}$ induces an SU(3)-structure $ (heat content, J,\omega,{\psi_{{\scriptscriptstyle}+}},{\psi_{{\scriptscriptstyle}-}})$ on $ W$ by means of the identities $ $ { \varphi}= \omega{\wedge}z^\flat + { \psi_{{\scriptscriptstyle}+}},\quad * _ { \varphi}{\varphi}= \ | }+e^{1256}+e^{3456} +f^{1367}+e^{1457}+e^{2357}-e^{2467},
\end{split}$$ where $*_{\varphl}$ is the Hodge okeeator vetermihed by $g_{\xarphi}$ and $\operatorname{vol}_{\vacphi}$, and $t^{pjk\cdots}$ is a shorthavd for thv wedge peodurt of covectors $x^j{\wedge}e^m{\cedge}s^n{\wedye}\rdots$. We shall gall both $\mdthcal{B}$ and $\madhzap{B}^*$ [*adapted*]{} bases to the G$_2$-structure ${\darphi}$.
Om hhe other hand, givtn wn ihner product $g$ on $V,$ we can consided a $g$-oruhonormal basis $\mayhcal{B}$ of $V$ and the G$_2$-struchure ${\varphi}\in\Lambda^3_{{\scrlptscriptstime}+}V^*$ yaving $\mathcxl{B}$ as an adapted basia. By, the metric $g_{\varphi}$ induced by ${\vcrphi}$ coincudws ahth $g$. We shell reser to such ${\ysrphi}$ ds the B$_2$-structure [*indmced*]{} uy tye basis $\mathcal{B}$. Henre, there is a surjecjive map $$\mdtkcal{G} : \Lambda^3_{{\scriptsceiptstyne}+}V^* \sighgqrruw \javhczl{S}^2_{{\scrlptacriptstyls}+}V^*,\quad \mathxal{G}({\varphi}) = g_{\varphi},$$ wrpvh is not innectivq, ws the set of all G$_2$-structures inducing uhe szme metric is parametrized by ${{\mathrm {SO}}}(7)/{{\mathrl G}}_2\cong{\maehbb R}P^7$ (see [@Bry Remark 4] for an explicit descriptimn). Thx E$_2$-stxmgturdw helonging to $\mathcal{G}^{-1}(g)$ will be called [*compatifme*]{} wpth $g$.
Consider now a G$_2$-structute ${\vstphi}$ on $V,$ let $e\in V$ bz a unit vector and dfnote bi $W$ thw 6-dimensijnal $g_{\varphi}$-orthogonal complemebt of $\langle z\rangle\subset V.$ Thzn, the G$_2$-strueture ${\farpho}$ induces an SU(3)-structurz $(h,J,\omsga,{\psi_{{\scriphscriptstgue}+}},{\psi_{{\scriptscripgstjle}-}})$ mn $W$ by means of the identyties $${\var'hi}= \okega{\wedee}z^\fkat + {\pfi_{{\scriptscgiptsbfle}+}},\quad *_{\varphi}{\varpji}= \ | }+e^{1256}+e^{3456} +e^{1367}+e^{1457}+e^{2357}-e^{2467}, \end{split}$$ where $*_{\varphi}$ is the determined $g_{\varphi}$ and and $e^{ijk\cdots}$ is product covectors $e^i{\wedge}e^j{\wedge}e^k{\wedge}\cdots$. We call both $\mathcal{B}$ $\mathcal{B}^*$ [*adapted*]{} bases to the G$_2$-structure On the other hand, given an inner product $g$ on $V,$ we can a $g$-orthonormal basis $\mathcal{B}$ of $V$ and the G$_2$-structure ${\varphi}\in\Lambda^3_{{\scriptscriptstyle}+}V^*$ having $\mathcal{B}$ as adapted By, metric induced by ${\varphi}$ coincides with $g$. We shall refer to such ${\varphi}$ as the G$_2$-structure [*induced*]{} the basis $\mathcal{B}$. Hence, there is a surjective $$\mathcal{G} : \Lambda^3_{{\scriptscriptstyle}+}V^* \rightarrow \mathcal{G}({\varphi}) = g_{\varphi},$$ which is injective, the set all inducing same metric is by ${{\mathrm {SO}}}(7)/{{\mathrm G}}_2\cong{\mathbb R}P^7$ (see [@Bry Remark 4] for an explicit description). The G$_2$-structures belonging to will be with $g$. now G$_2$-structure on $V,$ let be a unit vector and denote 6-dimensional $g_{\varphi}$-orthogonal complement of $\langle z\rangle\subset V.$ Then, G$_2$-structure ${\varphi}$ an SU(3)-structure $(h,J,\omega,{\psi_{{\scriptscriptstyle}+}},{\psi_{{\scriptscriptstyle}-}})$ on $W$ by of the identities $${\varphi}= \omega{\wedge}z^\flat + {\psi_{{\scriptscriptstyle}+}},\quad *_{\varphi}{\varphi}= | }+e^{1256}+e^{3456} +e^{1367}+e^{1457}+e^{2357}-e^{2467},
\end{split}$$ where $*_{\varphi}$ Is the Hodge OperaTor DetErMineD by $g_{\Varphi}$ and $\operaTOrnaMe{vol}_{\varphi}$, and $e^{ijk\cdotS}$ is a sHoRThanD FoR the wEdge proDUcT OF coVeCtOrs $E^i{\WEdGe}e^j{\wEdgE}e^k{\wedgE}\cdots$. We shAll CaLl both $\mathcaL{b}$ aNd $\mathcal{B}^*$ [*AdaPted*]{} bases to tHe G$_2$-StructUrE ${\vaRPhi}$.
On The Other Hand, giVEn an inNer producT $g$ ON $V,$ we caN ConsideR A $G$-oRthoNormal basis $\mathcaL{b}$ oF $v$ and the G$_2$-structUre ${\varPhI}\In\lAMbdA^3_{{\scRiptscriptStYle}+}V^*$ hAVing $\matHCaL{b}$ AS an ADapted basis. By, The metric $g_{\vARphI}$ inducEd By ${\vARphi}$ coIncidEs WIth $G$. We shall refEr to Such ${\varphI}$ as the g$_2$-StructuRE [*induceD*]{} by the BasIs $\mAthcAL{B}$. heNce, ThERe iS A sUrjECtiVe map $$\matHcAl{g} : \LambDa^3_{{\scRIPTScriPtsTyle}+}v^* \righTarrow \mathcal{s}^2_{{\scRiptSCriPtstyLe}+}V^*,\quAd \maThCal{G}({\vArphi}) = g_{\VarphI},$$ wHich is not injectIve, aS the set of All g$_2$-sTruCtUres iNDucing The SamE metric Is paramETriZeD BY ${{\MaThrm {SO}}}(7)/{{\mathrm G}}_2\cong{\mAtHBB R}p^7$ (see [@Bry REmark 4] fOR aN eXPlicit deScRipTion). tHE G$_2$-strUctuREs BelonginG to $\matHCaL{G}^{-1}(G)$ will be CaLled [*coMpAtiBle*]{} With $g$.
cOnsiDer now A G$_2$-structUre ${\vaRPhi}$ on $V,$ let $z\in V$ bE A unit vector anD DeNOTe BY $W$ thE 6-diMensional $g_{\vArphI}$-OrthOgonAL cOmpLEment Of $\lanGlE Z\rANgle\subset V.$ Then, the G$_2$-StRucturE ${\varpHi}$ induces an SU(3)-Structure $(h,j,\OMEga,{\psi_{{\scRiptSCrIPtstyle}+}},{\psi_{{\scriPtscrIptstyle}-}})$ on $w$ By means oF the iDentitieS $${\varphi}= \omEGA{\wedge}z^\fLat + {\Psi_{{\ScrIptSCRiPtstyle}+}},\quad *_{\vaRPHi}{\vaRpHi}= \ | }+e^{1256}+e^{3456} +e^{13 67}+e^{145 7}+e^ {23 57} -e ^{24 67},
\end{split}$$ wher e $*_{\varphi}$ is the Hodg eo pera t or dete rminedb y$ g _{\ va rp hi} $a nd $\op era torname {vol}_{\va rph i} $, and $e^{i j k\ cdots}$ is ashorthand fo r t he wed ge pr o ductofcovec tors $ e ^i{\we dge}e^j{\ we d ge}e^k { \wedge} \ c do ts$. We shall call bo t h$ \mathcal{B}$ a nd $\m at h ca l { B}^ *$[*adapted* ]{ } bas e s to th e G $ _ 2 $-s t ructure ${\va rphi}$.
On the other h and , given an i nn e r p roduct $g$on $ V,$ we ca n cons i der a $ g $-ortho normal ba sis $\m a th ca l{B }$ of$ V$ an d th e G$_2$- st ru cture ${\ v a r p hi}\ in\ Lamb da^3_ {{\scriptscri pts tyle } +}V ^*$ h aving $\m at hcal{ B}$ as an a da pted basis. By, the metric $ g_{ \v arp hi }$ in d uced b y $ {\v arphi}$ coinci d eswi t h $g $. We shall referto s uc h ${\var phi}$a sth e G$_2$-s tr uct ure[ * induc ed*] { }by the b asis $ \ ma th cal{B}$ .Hence, t her e i s a s u rjec tive m ap $$\ma thcal { G} : \Lambda^3 _ {{\scriptscri p ts t y le } +}V^ * \ rightarrow\mat h cal{ S}^2 _ {{ \sc r iptsc ripts ty l e} + }V^*,\quad \mathcal {G }({\va rphi} ) = g_{\varph i},$$ whic h i s not in ject i ve , as the set of allG$_2$-stru c tures in ducin g the sa me metric i s parame tri zed by ${ { \ ma thrm {SO}}}(7 ) / {{\m at hrm G}} _2\ cong{\m ath bbR}P ^7$ ( see [@Bry Remark4] f or a n e xplic i t descri pt ion ). Th e G$_ 2 $-stru cture s be lo ng i ngto $\ma t hc a l {G}^ {- 1} (g)$ wi ll be c alle d [* compati ble*]{} w ith $g$.
Co nsidernow a G$_2$-s tr ucture ${\ va rph i}$ on $ V,$ let$z\in V$ be a unit vect o r and d eno te by $W$ the 6-di men sional $g _ {\varp hi}$-o rthog on alc o mplem e n tof$\ langle z\r a n gle \subs et V.$ Then,the G$_2$-structur e ${ \varphi}$ ind uce s an S U( 3)- s tr u ctu re $(h , J ,\omega,{\psi_{ {\scriptsc ri p ts tyle}+}},{ \ psi _{ {\scrip tscript style } -}})$ o n $W$ bymeans ofth e id e n tit ies $${\va rphi}= \ omega{\we d ge}z^ \ fl at +{\p si_{{\ sc rip tscri ptstyl e }+} },\qu ad *_{ \v arphi} {\var ph i}= \ | }+e^{1256}+e^{3456} +e^{1367}+e^{1457}+e^{2357}-e^{2467},
\end{split}$$_where $*_{\varphi}$_is the Hodge operator_determined by_$g_{\varphi}$_and $\operatorname{vol}_{\varphi}$,_and_$e^{ijk\cdots}$ is a_shorthand for the_wedge product of covectors_$e^i{\wedge}e^j{\wedge}e^k{\wedge}\cdots$. We shall_call_both $\mathcal{B}$ and $\mathcal{B}^*$ [*adapted*]{} bases to the G$_2$-structure ${\varphi}$.
On the other hand, given_an_inner product_$g$_on_$V,$ we can consider a_$g$-orthonormal basis $\mathcal{B}$ of $V$_and the_G$_2$-structure ${\varphi}\in\Lambda^3_{{\scriptscriptstyle}+}V^*$ having $\mathcal{B}$ as an adapted basis._By,_the metric $g_{\varphi}$_induced by ${\varphi}$ coincides with $g$. We shall refer_to such ${\varphi}$ as the G$_2$-structure_[*induced*]{} by the_basis_$\mathcal{B}$._Hence, there is a_surjective map $$\mathcal{G} : \Lambda^3_{{\scriptscriptstyle}+}V^* \rightarrow_\mathcal{S}^2_{{\scriptscriptstyle}+}V^*,\quad \mathcal{G}({\varphi}) = g_{\varphi},$$ which is_not injective, as the set of all_G$_2$-structures inducing the same metric is_parametrized by ${{\mathrm {SO}}}(7)/{{\mathrm G}}_2\cong{\mathbb_R}P^7$ (see_[@Bry Remark 4] for an_explicit description). The_G$_2$-structures belonging_to $\mathcal{G}^{-1}(g)$ will_be called [*compatible*]{} with $g$.
Consider now_a G$_2$-structure ${\varphi}$_on $V,$ let $z\in V$ be_a_unit vector and_denote_by_$W$ the_6-dimensional $g_{\varphi}$-orthogonal complement_of_$\langle z\rangle\subset_V.$_Then, the G$_2$-structure ${\varphi}$ induces an_SU(3)-structure_$(h,J,\omega,{\psi_{{\scriptscriptstyle}+}},{\psi_{{\scriptscriptstyle}-}})$ on $W$ by means of the_identities $${\varphi}= \omega{\wedge}z^\flat +_{\psi_{{\scriptscriptstyle}+}},\quad_*_{\varphi}{\varphi}= \ |
in L^p(0,T;X_1)$ and $\frac{d}{dt} f\in L^{p'}(0,T;X_1')$, $1< p <\infty$, cf. Zeidler [@ZeidlerIIa Proposition 23.23]. In particular, (\[eq:HNormDifferential\]) implies $$\label{eq:StrongBUCConv}
\sup_{t\in [0,T]}\|f(t)\|_{H}^2
\leq 2\left(\|\partial_t f\|_{L^2(0,T;X_1')}\|f\|_{L^2(0,T;X_1)} + \|f(0)\|_H^2\right).$$ Replacing $f(t)$ by $tf(t)$ and $(T-t)f(T-t)$, one easily derives from the latter estimate $$\label{eq:StrongBUCConv'}
\sup_{t\in [0,T]}\|f(t)\|_{H}
\leq C_T\|f\|_{H^1(0,T;X_1')}^{\frac12}\|f\|_{L^2(0,T;X_1)}^{\frac12}$$ for some $C_T>0$ depending on $T>0$.
In the following $\mathcal{L}^n(V)$, $n\in{\mathbb N}$, denotes the space of all $n$-linear mappings $A\colon V^n\to {\mathbb R}$ for a vector space $V$. Moreover, if $A\in
\mathcal{L}^n(V)$, $n\geq 2$, and $x_1,\ldots, x_k\in V$, $1\leq k\leq n$, then $A[x_1,\ldots, x_k] \in
\mathcal{L}^{n-k}(V)$ is defined by $A[x_1,\ldots, x_k](x_{k+1},\ldots, x_n)= A(x_1,\ldots,
x_n)$ for all $x_{k+1},\ldots,x_n\in V$.
We introduce the scaled inner product $$A:_h B = \frac1{h^2} {{\rm sym}\,}A : {{\rm sym}\,}B + {{\rm skew}\,}A:{{\rm skew | in L^p(0,T;X_1)$ and $ \frac{d}{dt } f\in L^{p'}(0,T;X_1')$, $ 1 < p < \infty$, cf. Zeidler [ @ZeidlerIIa Proposition 23.23 ]. In particular, (\[eq: HNormDifferential\ ]) implies $ $ \label{eq: StrongBUCConv }
\sup_{t\in [ 0,T]}\|f(t)\|_{H}^2
\leq 2\left(\|\partial_t f\|_{L^2(0,T;X_1')}\|f\|_{L^2(0,T;X_1) } + \|f(0)\|_H^2\right).$$ replace $ f(t)$ by $ tf(t)$ and $ (metric ton - t)f(T - t)$, one easily derives from the latter estimate $ $ \label{eq: StrongBUCConv' }
\sup_{t\in [ 0,T]}\|f(t)\|_{H }
\leq C_T\|f\|_{H^1(0,T;X_1')}^{\frac12}\|f\|_{L^2(0,T;X_1)}^{\frac12}$$ for some $ C_T>0 $ count on $ T>0$.
In the following $ \mathcal{L}^n(V)$, $ n\in{\mathbb N}$, denotes the outer space of all $ n$-linear mappings $ A\colon V^n\to { \mathbb R}$ for a vector space $ V$. furthermore, if $ A\in
\mathcal{L}^n(V)$, $ n\geq 2 $, and $ x_1,\ldots, x_k\in V$, $ 1\leq k\leq n$, then $ A[x_1,\ldots, x_k ] \in
\mathcal{L}^{n - k}(V)$ is define by $ A[x_1,\ldots, x_k](x_{k+1},\ldots, x_n)= A(x_1,\ldots,
x_n)$ for all $ x_{k+1},\ldots, x_n\in V$.
We introduce the scale inner product $ $ A:_h boron = \frac1{h^2 } { { \rm sym}\,}A: { { \rm sym}\,}B + { { \rm skew}\,}A:{{\rm skew | in P^p(0,T;X_1)$ and $\frac{d}{dt} f\in L^{p'}(0,T;W_1')$, $1< p <\infty$, cf. Zeneler [@ZemdlerIIz Proposktion 23.23]. In particular, (\[eq:HNormViffwrentual\]) implies $$\label{eq:StrungBUCConn}
\sup_{t\in [0,T]}\|f(t)\|_{I}^2
\leq 2\left(\|\partial_t f\|_{L^2(0,T;X_1')}\|f\|_{L^2(0,T;X_1)} + \|f(0)\|_G^2\vight).$$ Ceplacing $f(t)$ by $tf(t)$ and $(T-d)f(T-t)$, one easilf aexives from the latter estimate $$\label{qq:StronbBKCConv'}
\sup_{t\in [0,T]}\|f(t)\|_{N}
\leq B_T\|n\|_{H^1(0,T;X_1')}^{\frac12}\|f\|_{L^2(0,T;X_1)}^{\frac12}$$ for some $C_T>0$ depsnding mn $T>0$.
In the foklowing $\mathcal{L}^n(V)$, $n\in{\mathhb N}$, denotes the space of all $n$-libear nappings $A\couon V^n\to {\mathbb R}$ for z vector space $V$. Moreover, if $A\iv
\matheal{L}^n(V)$, $n\geq 2$, qnd $f_1,\ldots, x_k\in V$, $1\lez k\leq n$, thek $A[x_1,\ldods, x_k] \im
\mathcal{L}^{n-k}(V)$ ix dxfinwd by $A[x_1,\ldots, x_k](x_{k+1},\ldovs, x_n)= A(x_1,\ldots,
x_n)$ for wll $x_{k+1},\ldods,r_n\in V$.
We introduce thw wcaleg inter oeodjct $$A:_i B = \frac1{j^2} {{\rj sym}\,}A : {{\rm sym}\,}B + {{\rm sjew}\,}A:{{\rm skew | in L^p(0,T;X_1)$ and $\frac{d}{dt} f\in L^{p'}(0,T;X_1')$, $1< cf. [@ZeidlerIIa Proposition In particular, (\[eq:HNormDifferential\]) 2\left(\|\partial_t + \|f(0)\|_H^2\right).$$ Replacing by $tf(t)$ and one easily derives from the latter $$\label{eq:StrongBUCConv'} \sup_{t\in [0,T]}\|f(t)\|_{H} \leq C_T\|f\|_{H^1(0,T;X_1')}^{\frac12}\|f\|_{L^2(0,T;X_1)}^{\frac12}$$ for some $C_T>0$ depending on $T>0$. In the $\mathcal{L}^n(V)$, $n\in{\mathbb N}$, denotes the space of all $n$-linear mappings $A\colon V^n\to {\mathbb for vector $V$. if $A\in \mathcal{L}^n(V)$, $n\geq 2$, and $x_1,\ldots, x_k\in V$, $1\leq k\leq n$, then $A[x_1,\ldots, x_k] \in is defined by $A[x_1,\ldots, x_k](x_{k+1},\ldots, x_n)= A(x_1,\ldots, x_n)$ all $x_{k+1},\ldots,x_n\in V$. We the scaled inner product $$A:_h = {{\rm sym}\,}A {{\rm + skew}\,}A:{{\rm skew | in L^p(0,T;X_1)$ and $\frac{d}{dt} f\in L^{p'}(0,T;X_1')$, $1< p <\iNfty$, cf. ZeidLer [@ZeIdlErIia propOsitIon 23.23]. In particulaR, (\[Eq:HNOrmDifferential\]) implies $$\Label{Eq:sTronGbUcConv}
\Sup_{t\in [0,T]}\|F(T)\|_{H}^2
\LEQ 2\leFt(\|\PaRtiAl_T F\|_{L^2(0,t;X_1')}\|f\|_{L^2(0,T;x_1)} + \|f(0)\|_H^2\Right).$$ RePlacing $f(t)$ bY $tf(T)$ aNd $(T-t)f(T-t)$, one eaSIlY derives frOm tHe latter estiMatE $$\label{Eq:strONgBUCconV'}
\sup_{t\In [0,T]}\|f(t)\|_{H}
\LEq C_T\|f\|_{H^1(0,t;X_1')}^{\frac12}\|f\|_{L^2(0,T;x_1)}^{\fRAc12}$$ for sOMe $C_T>0$ depENDiNg on $t>0$.
In the following $\maTHcAL{L}^n(V)$, $n\in{\mathbb N}$, DenoteS tHE sPACe oF alL $n$-linear maPpIngs $A\COlon V^n\tO {\MaTHBB R}$ fOR a vector space $v$. Moreover, if $a\In
\mAthcal{l}^n(v)$, $n\gEQ 2$, and $x_1,\lDots, x_K\iN v$, $1\leQ k\leq n$, then $A[X_1,\ldoTs, x_k] \in
\matHcal{L}^{n-K}(v)$ is defiNEd by $A[x_1,\lDots, x_k](X_{k+1},\lDotS, x_n)= A(X_1,\LdOtS,
x_n)$ FoR All $X_{K+1},\lDotS,X_n\iN V$.
We intrOdUcE the sCaleD INNEr prOduCt $$A:_h b = \frac1{H^2} {{\rm sym}\,}A : {{\rm sym}\,}B + {{\Rm sKew}\,}A:{{\RM skEw | in L^p(0,T;X_1)$ and $\fra c{d}{dt} f \in L ^{p '}( 0, T;X_ 1')$ , $1< p <\inft y $, c f. Zeidler [@ZeidlerII a Pro po s itio n 2 3.23] . In pa r ti c u lar ,(\ [eq :H N or mDiff ere ntial\] ) implies$$\ la bel{eq:Stron g BU CConv}
\ sup _{t\in [0,T] }\| f(t)\| _{ H}^ 2
\le q 2 \left (\|\pa r tial_t f\|_{L^2 (0 , T;X_1' ) }\|f\|_ { L ^2 (0,T ;X_1)} + \|f(0)\| _ H^ 2 \right).$$ Rep lacing $ f (t ) $ by $t f(t)$ and$( T-t)f ( T-t)$,o ne e a sil y derives from the latter est imate$$ \la b el{eq: Stron gB U CCo nv'}
\sup _{t\ in [0,T]} \|f(t) \ |_{H}
\ leq C_T \|f\|_ {H^ 1(0 ,T;X _ 1' )} ^{\ fr a c12 } \| f\| _ {L^ 2(0,T;X_ 1) }^ {\fra c12} $ $ f or s ome $C_ T>0$depending on$T> 0$.Inthe f ollow ing$\ mathc al{L}^ n(V)$ ,$n\in{\mathbb N }$,denotes t hesp ace o f all $n$-li nea r m appings $A\col o n V ^n \ t o { \mathbb R}$ for ave c t or space $ V$. Mo r eo ve r , if $A\ in
\m athc a l {L}^n (V)$ , $ n\geq 2$ , and$ x_ 1, \ldots, x _k\inV$ , $ 1\l eq k\ l eq n $, the n $A[x_1 ,\ldo t s, x_k] \in
\m a thcal{L}^{n-k } (V ) $ i s def ine d by $A[x_1 ,\ld o ts,x_k] ( x_ {k+ 1 },\ld ots,x_ n )= A(x_1,\ldots,
x_n)$ f or all $x_{ k+1},\ldots,x _n\in V$.W e introdu ce t h es caled inner pr oduct $$A:_h B= \frac1{ h^2}{{\rm sy m}\,}A :{ { \rm sym} \,} B + {{ \rm s ke w}\,}A:{{\rms k ew | in L^p(0,T;X_1)$_and $\frac{d}{dt}_f\in L^{p'}(0,T;X_1')$, $1< p_<\infty$, cf._Zeidler [@ZeidlerIIa_Proposition 23.23]._In_particular, (\[eq:HNormDifferential\]) implies_$$\label{eq:StrongBUCConv}
\sup_{t\in_[0,T]}\|f(t)\|_{H}^2
\leq 2\left(\|\partial_t f\|_{L^2(0,T;X_1')}\|f\|_{L^2(0,T;X_1)}_+ \|f(0)\|_H^2\right).$$ Replacing_$f(t)$_by $tf(t)$ and $(T-t)f(T-t)$, one easily derives from the latter estimate $$\label{eq:StrongBUCConv'}
\sup_{t\in_[0,T]}\|f(t)\|_{H}
_\leq C_T\|f\|_{H^1(0,T;X_1')}^{\frac12}\|f\|_{L^2(0,T;X_1)}^{\frac12}$$_for_some_$C_T>0$ depending on $T>0$.
In the_following $\mathcal{L}^n(V)$, $n\in{\mathbb N}$, denotes_the space_of all $n$-linear mappings $A\colon V^n\to {\mathbb R}$_for_a vector space_$V$. Moreover, if $A\in
\mathcal{L}^n(V)$, $n\geq 2$, and $x_1,\ldots, x_k\in_V$, $1\leq k\leq n$, then $A[x_1,\ldots,_x_k] \in
\mathcal{L}^{n-k}(V)$ is_defined_by_$A[x_1,\ldots, x_k](x_{k+1},\ldots, x_n)= A(x_1,\ldots,
x_n)$_for all $x_{k+1},\ldots,x_n\in V$.
We introduce the_scaled inner product $$A:_h B =_\frac1{h^2} {{\rm sym}\,}A : {{\rm sym}\,}B +_{{\rm skew}\,}A:{{\rm skew |
This can also be done by considering ${{{\mathrm{Dih}}}_2\triangleleft{{\mathrm{Dih}}}_4}$ such that the graded structure is $$\begin{matrix}
{\mathbb{R}}[{{\mathrm{Dih}}}_2]&=&{\mathbb{R}}^{{\oplus 2}}\oplus{\mathbb{R}}^{{\oplus 2}}&&\\
{\mathbin{\text{\rotatebox[origin=c]{-90}{$\hookrightarrow$}}}}&&{\mathbin{\text{\rotatebox[origin=c]{-90}{$\hookrightarrow$}}}}&&\\
{\mathbb{R}}[{{\mathrm{Dih}}}_4]&=&({\mathbb{R}}^{{\oplus 2}})^{{\oplus 2}}\oplus M_2({\mathbb{R}})&=&{{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{1,1},
\end{matrix}$$ $$\begin{aligned}
{{\mathrm{Cl}}}_{p,q}\hat{\otimes}{\mathbb{R}}[{\mathbb{Z}}_4]&={{\mathrm{Cl}}}_{p,q}\hat{\otimes}({{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{1,1})\\&={{\mathrm{Cl}}}_{p,q+1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{p+1,q+1}.
\end{aligned}$$
For ${n=3}$ we use ${{\mathrm{Dih}}}_6={\mathbb{Z}}_2\times{{\mathrm{Dih}}}_3$ and split our algebra to ${{{\mathrm{Cl}}}_{p,q}\hat\otimes{\mathbb{R}}\langle s,\sigma_v\rangle=({{\mathrm{Cl}}}_{p,q}\hat\otimes{\mathbb{R}}\langle s^3\rangle)\otimes{\mathbb{R}}\langle s^2,s\sigma_v\rangle}$ and get $$({\mathbb{R}}^{{\oplus 2}}\oplus M_2({\mathbb{R}}))\otimes{{\mathrm{Cl}}}_{q+7,3+1}\mapsto{\mathcal{R}}_{q-4}^3.$$
Prismatic Symmetry $D_{nh}$
---------------------------
{height="\fg"} ![The $D_{nh | This can also be done by considering $ { { { \mathrm{Dih}}}_2\triangleleft{{\mathrm{Dih}}}_4}$ such that the graded social organization is $ $ \begin{matrix }
{ \mathbb{R}}[{{\mathrm{Dih}}}_2]&=&{\mathbb{R}}^{{\oplus 2}}\oplus{\mathbb{R}}^{{\oplus 2}}&&\\
{ \mathbin{\text{\rotatebox[origin = c]{-90}{$\hookrightarrow$}}}}&&{\mathbin{\text{\rotatebox[origin = c]{-90}{$\hookrightarrow$}}}}&&\\
{ \mathbb{R}}[{{\mathrm{Dih}}}_4]&=&({\mathbb{R}}^{{\oplus 2}})^{{\oplus 2}}\oplus M_2({\mathbb{R}})&=&{{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{1,1 },
\end{matrix}$$ $ $ \begin{aligned }
{ { \mathrm{Cl}}}_{p, q}\hat{\otimes}{\mathbb{R}}[{\mathbb{Z}}_4]&={{\mathrm{Cl}}}_{p, q}\hat{\otimes}({{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{1,1})\\&={{\mathrm{Cl}}}_{p, q+1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{p+1,q+1 }.
\end{aligned}$$
For $ { n=3}$ we practice $ { { \mathrm{Dih}}}_6={\mathbb{Z}}_2\times{{\mathrm{Dih}}}_3 $ and burst our algebra to $ { { { \mathrm{Cl}}}_{p, q}\hat\otimes{\mathbb{R}}\langle s,\sigma_v\rangle=({{\mathrm{Cl}}}_{p, q}\hat\otimes{\mathbb{R}}\langle s^3\rangle)\otimes{\mathbb{R}}\langle s^2,s\sigma_v\rangle}$ and get $ $ ({ \mathbb{R}}^{{\oplus 2}}\oplus M_2({\mathbb{R}}))\otimes{{\mathrm{Cl}}}_{q+7,3 + 1}\mapsto{\mathcal{R}}_{q-4}^3.$$
Prismatic Symmetry $ D_{nh}$
---------------------------
! [ The $ D_{nh}$ prismatic symmetry degree groups: $ D_{1h},D_{2h},D_{3h},D_{4h},D_{6h}$.](C2v_3D.png " fig:"){height="\fg " } ! [ The $ D_{nh | Thls can also be done by cunsidering ${{{\mathtm{Eih}}}_2\trienglelert{{\mathrm{Aih}}}_4}$ such that the graded strnctuee is $$\begin{matrix}
{\mathbb{R}}[{{\matfrm{Dih}}}_2]&=&{\matjbb{R}}^{{\opluw 2}}\opoys{\mathbb{R}}^{{\o'mus 2}}&&\\
{\matmyin{\tesb{\rotaceuox[origin=c]{-90}{$\hookrlghtarrow$}}}}&&{\madhbin{\text{\rotatabux[lrigin=c]{-90}{$\hookrightarrow$}}}}&&\\
{\mathbb{R}}[{{\mathrm{Dyh}}}_4]&=&({\mathbn{R}}^{{\lplus 2}})^{{\oplus 2}}\opjus K_2({\iathgb{R}})&=&{{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{1,1},
\end{mztrix}$$ $$\btgin{aligned}
{{\mathrm{Ck}}}_{p,q}\hat{\otimes}{\mathbb{R}}[{\mathbb{Z}}_4]&={{\lathgm{Cl}}}_{p,q}\hat{\otimes}({{\matjrm{Cl}}}_{0,1}^{{\oplus 2}}\ipluf{{\nathrm{Cl}}}_{1,1})\\&={{\mathfm{Cl}}}_{p,q+1}^{{\oplus 2}}\oplus{{\mathrj{Cl}}}_{p+1,q+1}.
\end{aligned}$$
For ${n=3}$ we use ${{\matfrm{Dik}}}_6={\mathbb{Z}}_2\timgs{{\jahvrm{Dih}}}_3$ and wplit our algebra to ${{{\mathsm{Cl}}}_{p,q}\hst\otimes{\mathbb{V}}\langne w,\sigma_v\rangle=({{\mathrm{Cl}}}_{',q}\hat\otimes{\mathbb{R}}\laggle s^3\ranclz)\otimes{\mathbb{R}}\langle w^2,s\wigma_e\rancle}$ xbd eet $$({\methgb{R}}^{{\oplks 2}}\kplus M_2({\matgbb{R}}))\otimes{{\mqthrm{Cl}}}_{q+7,3+1}\mapsto{\mathcsl{W}}_{w-4}^3.$$
Prismatic Syjmetry $D_{gh}$
---------------------------
{heighr="\fg"} ![The $D_{nh | This can also be done by considering that graded structure $$\begin{matrix} {\mathbb{R}}[{{\mathrm{Dih}}}_2]&=&{\mathbb{R}}^{{\oplus 2}}\oplus{\mathbb{R}}^{{\oplus M_2({\mathbb{R}})&=&{{\mathrm{Cl}}}_{0,1}^{{\oplus \end{matrix}$$ $$\begin{aligned} {{\mathrm{Cl}}}_{p,q}\hat{\otimes}{\mathbb{R}}[{\mathbb{Z}}_4]&={{\mathrm{Cl}}}_{p,q}\hat{\otimes}({{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{p+1,q+1}. \end{aligned}$$ For we use ${{\mathrm{Dih}}}_6={\mathbb{Z}}_2\times{{\mathrm{Dih}}}_3$ and split our to ${{{\mathrm{Cl}}}_{p,q}\hat\otimes{\mathbb{R}}\langle s,\sigma_v\rangle=({{\mathrm{Cl}}}_{p,q}\hat\otimes{\mathbb{R}}\langle s^3\rangle)\otimes{\mathbb{R}}\langle s^2,s\sigma_v\rangle}$ and get $$({\mathbb{R}}^{{\oplus 2}}\oplus M_2({\mathbb{R}}))\otimes{{\mathrm{Cl}}}_{q+7,3+1}\mapsto{\mathcal{R}}_{q-4}^3.$$ Prismatic Symmetry $D_{nh}$ {height="\fg"} ![The $D_{nh | This can also be done by considEring ${{{\mathrM{Dih}}}_2\tRiaNglElEft{{\mAthrM{Dih}}}_4}$ such that thE GradEd structure is $$\begin{matrIx}
{\matHbB{r}}[{{\matHRm{dih}}}_2]&=&{\maThbb{R}}^{{\opLUs 2}}\OPLus{\MaThBb{R}}^{{\OpLUs 2}}&&\\
{\MathbIn{\tExt{\rotaTebox[origiN=c]{-90}{$\hOoKrightarrow$}}}}&&{\mAThBin{\text{\rotAteBox[origin=c]{-90}{$\hoOkrIghtarRoW$}}}}&&\\
{\maTHbb{R}}[{{\mAthRm{Dih}}}_4]&=&({\Mathbb{r}}^{{\Oplus 2}})^{{\oPlus 2}}\oplus m_2({\mAThbb{R}})&=&{{\mAThrm{Cl}}}_{0,1}^{{\oPLUs 2}}\OpluS{{\mathrm{Cl}}}_{1,1},
\end{matriX}$$ $$\BeGIn{aligned}
{{\mathrM{Cl}}}_{p,q}\hAt{\OTiMES}{\maThbB{R}}[{\mathbb{Z}}_4]&={{\mAtHrm{Cl}}}_{P,Q}\hat{\otiMEs}({{\MATHrm{cL}}}_{0,1}^{{\oplus 2}}\oplus{{\maThrm{Cl}}}_{1,1})\\&={{\mathrM{cl}}}_{p,Q+1}^{{\oplus 2}}\OpLus{{\MAthrm{CL}}}_{p+1,q+1}.
\enD{aLIgnEd}$$
For ${n=3}$ we use ${{\MathRm{Dih}}}_6={\mathBb{Z}}_2\timES{{\mathrm{dIh}}}_3$ and spLit our AlgEbrA to ${{{\mAThRm{cl}}}_{p,Q}\hAT\otIMeS{\maTHbb{r}}\langle s,\SiGmA_v\ranGle=({{\mATHRM{Cl}}}_{p,Q}\haT\otiMes{\maThbb{R}}\langle s^3\rAngLe)\otIMes{\MathbB{R}}\lanGle s^2,S\sIgma_v\Rangle}$ And geT $$({\mAthbb{R}}^{{\oplus 2}}\opluS M_2({\maThbb{R}}))\otimEs{{\mAtHrm{cl}}}_{Q+7,3+1}\mapsTO{\mathcAl{R}}_{Q-4}^3.$$
PrIsmatic symmetrY $d_{nh}$
---------------------------
{height="\Fg"} ![the $d_{nh | This can also be done byconsiderin g ${{ {\m ath rm {Dih }}}_ 2\triangleleft { {\ma thrm{Dih}}}_4}$ such t hat t he grad e dstruc ture is $$ \ b egi n{ ma tri x} {\ mathb b{R }}[{{\m athrm{Dih} }}_ 2] &=&{\mathbb{ R }} ^{{\oplus2}} \oplus{\math bb{ R}}^{{ \o plu s 2}}& &\\
{\ma thbin{ \ text{\ rotatebox [o r igin=c ] {-90}{$ \ h oo krig htarrow$}}}}&&{\m a th b in{\text{\rota tebox[ or i gi n = c]{ -90 }{$\hookri gh tarro w $}}}}&& \ \{ \ m ath b b{R}}[{{\math rm{Dih}}}_4 ] &=& ({\mat hb b{R } }^{{\o plus2} } )^{ {\oplus 2}} \opl us M_2({\ mathbb { R}})&=& { {\mathr m{Cl}} }_{ 0,1 }^{{ \ op lu s 2 }} \ opl u s{ {\m a thr m{Cl}}}_ {1 ,1 },
\e nd{m a t r i x}$$ $$ \beg in{al igned}
{{\mat hrm {Cl} } }_{ p,q}\ hat{\ otim es }{\ma thbb{R }}[{\ ma thbb{Z}}_4]&={{ \mat hrm{Cl}}} _{p ,q }\h at {\oti m es}({{ \ma thr m{Cl}}} _{0,1}^ { {\o pl u s 2} }\oplus{{\mathrm{C l} } } _{ 1,1})\\& ={{\ma t hr m{ C l}}}_{p, q+ 1}^ {{\o p l us 2} }\op l us {{\mathr m{Cl}} } _{ p+ 1,q+1}.
\ end{al ig ned }$$
For ${n= 3}$ we use ${{ \math r m{Dih}}}_6={\m a thbb{Z}}_2\ti m es { { \m a thrm {Di h}}}_3$ and spl i t ou r al g eb rat o ${{ {\mat hr m {C l }}}_{p,q}\hat\otime s{ \mathb b{R}} \langle s,\si gma_v\rang l e = ({{\math rm{C l }} } _{p,q}\hat\oti mes{\ mathbb{R}} \ langle s ^3\ra ngle)\ot imes{\mat h b b{R}}\la ngl e s ^2, s\s i g ma _v\rangle}$ a n d get $ $({\mat hbb {R}}^{{ \op lus 2} }\o pl us M_2({\ mathbb{R }} )) \o ti mes {{\ma t hrm{Cl}} }_ {q+ 7, 3+1 }\map s to{\ma thcal {R}} _{ q- 4 }^3 .$$
Pr i sm a t ic S ym me try$D_ {n h}$
- ---- - --- ------- --------- --{height= "\f g"} ![ The $D_{nh | This_can also_be done by considering_${{{\mathrm{Dih}}}_2\triangleleft{{\mathrm{Dih}}}_4}$ such_that_the graded_structure_is $$\begin{matrix}
{\mathbb{R}}[{{\mathrm{Dih}}}_2]&=&{\mathbb{R}}^{{\oplus 2}}\oplus{\mathbb{R}}^{{\oplus_2}}&&\\
{\mathbin{\text{\rotatebox[origin=c]{-90}{$\hookrightarrow$}}}}&&{\mathbin{\text{\rotatebox[origin=c]{-90}{$\hookrightarrow$}}}}&&\\
{\mathbb{R}}[{{\mathrm{Dih}}}_4]&=&({\mathbb{R}}^{{\oplus 2}})^{{\oplus 2}}\oplus_M_2({\mathbb{R}})&=&{{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{1,1},
\end{matrix}$$ $$\begin{aligned}
{{\mathrm{Cl}}}_{p,q}\hat{\otimes}{\mathbb{R}}[{\mathbb{Z}}_4]&={{\mathrm{Cl}}}_{p,q}\hat{\otimes}({{\mathrm{Cl}}}_{0,1}^{{\oplus 2}}\oplus{{\mathrm{Cl}}}_{1,1})\\&={{\mathrm{Cl}}}_{p,q+1}^{{\oplus_2}}\oplus{{\mathrm{Cl}}}_{p+1,q+1}.
\end{aligned}$$
For ${n=3}$ we_use_${{\mathrm{Dih}}}_6={\mathbb{Z}}_2\times{{\mathrm{Dih}}}_3$ and split our algebra to ${{{\mathrm{Cl}}}_{p,q}\hat\otimes{\mathbb{R}}\langle s,\sigma_v\rangle=({{\mathrm{Cl}}}_{p,q}\hat\otimes{\mathbb{R}}\langle s^3\rangle)\otimes{\mathbb{R}}\langle s^2,s\sigma_v\rangle}$ and get $$({\mathbb{R}}^{{\oplus 2}}\oplus_M_2({\mathbb{R}}))\otimes{{\mathrm{Cl}}}_{q+7,3+1}\mapsto{\mathcal{R}}_{q-4}^3.$$
Prismatic_Symmetry $D_{nh}$
---------------------------
{height="\fg"} ![The_$D_{nh |
left\langle{ -i}\right\rangle}_N}.$$ The multiplication is $$e_{ri}e_{sj}=\delta_{rs}e_{r{\left\langle{ i+j}\right\rangle}_N}.$$
This is the direct sum of $N$ group algebras of ${\mathbb{Z}}_N$. There are $N^2$ 1-dimensional simple modules. We use ${\lfloor{\cdots}\rceil}$ to denote a composite label. The simple modules can be labeled with two numbers ${\lfloor{ ri}\rceil}$. The basis is $$M_{{\lfloor{ ri}\rceil}}^r=\frac{1}{N}\sum_{k=0}^{N-1}
\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}i k}e_{rk}.$$ The matrix representations are $$M_{{\lfloor{
ri}\rceil},ss{\left\langle{s+j}\right\rangle}_N}^j=\delta_{rs}\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}ij}.$$ Then, we get $$\begin{aligned}
N_{{\lfloor{ ri}\rceil},s}&=\delta_{rs},\\
d_{{\lfloor{ ri}\rceil}}&=1,\\
T_{{\lfloor{ ri}\rceil}}&=\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}ri},\\
S_{{\lfloor{ ri}\rceil}{\lfloor{ sj}\rceil}}&=\frac{1}{N}\mathrm{e}^{\frac{2\pi \mathrm{i}}{N}\left(
rj+si \right)}.\end{aligned}$$
### Finite group *G* model {#finiteG}
Similar to the ${\mathbb{Z}}_N$ case we can define the rotation-invariant string-net model for a finite group $G$:
- $|G|$ types of strings, labeled by the group elements $g\in G$ and $g^*=g^{-1}$.
- The trivial string is now labeled by ${\textbf{1}}$, the identity element of $G$.
- $N_{g_1 g_2 g_3}=1$ iff $g_1 g_2 g_3={\textbf{1}}$.
- $F_{g_3 g_4 g_6}^{g_1 g_2 g_5}=1$ iff $ | left\langle { -i}\right\rangle}_N}.$$ The multiplication is $ $ e_{ri}e_{sj}=\delta_{rs}e_{r{\left\langle { i+j}\right\rangle}_N}.$$
This is the direct sum of $ N$ group algebras of $ { \mathbb{Z}}_N$. There be $ N^2 $ 1 - dimensional childlike modules. We use $ { \lfloor{\cdots}\rceil}$ to denote a composite label. The bare modules can be labeled with two number $ { \lfloor { ri}\rceil}$. The footing is $ $ M_{{\lfloor { ri}\rceil}}^r=\frac{1}{N}\sum_{k=0}^{N-1 }
\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}i k}e_{rk}.$$ The matrix representations are $ $ M_{{\lfloor {
ri}\rceil},ss{\left\langle{s+j}\right\rangle}_N}^j=\delta_{rs}\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}ij}.$$ Then, we get $ $ \begin{aligned }
N_{{\lfloor { ri}\rceil},s}&=\delta_{rs},\\
d_{{\lfloor { ri}\rceil}}&=1,\\
T_{{\lfloor { ri}\rceil}}&=\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}ri},\\
S_{{\lfloor { ri}\rceil}{\lfloor { sj}\rceil}}&=\frac{1}{N}\mathrm{e}^{\frac{2\pi \mathrm{i}}{N}\left (
rj+si \right)}.\end{aligned}$$
# # # Finite group * G * model { # finiteG }
Similar to the $ { \mathbb{Z}}_N$ case we can specify the rotation - invariant string - final model for a finite group $ G$:
- $ |G|$ types of strings, labeled by the group elements $ g\in G$ and $ g^*=g^{-1}$.
- The trivial string is now label by $ { \textbf{1}}$, the identity element of $ G$.
- $ N_{g_1 g_2 g_3}=1 $ iff $ g_1 g_2 g_3={\textbf{1}}$.
- $ F_{g_3 g_4 g_6}^{g_1 g_2 g_5}=1 $ iff $ | lefh\langle{ -i}\right\rangle}_N}.$$ Tht multiplication nw $$e_{ri}e_{vj}=\deltz_{rs}e_{r{\lefg\langle{ i+j}\right\rangle}_N}.$$
This id rhe durect sum of $N$ group augebras ov ${\mathbb{Z}}_N$. Tiere are $N^2$ 1-dimensional sliple lodunxs. We use ${\lfloot{\cdots}\rceil}$ do denote a cokpusnte label. The simple modules can be jabeled wlth two numberf ${\lfkjor{ dp}\rgeil}$. The basis is $$M_{{\lfloor{ ri}\rceim}}^r=\frac{1}{N}\vum_{k=0}^{N-1}
\mathrm{e}^{-\ftac{2\pi \mathrm{i}}{N}i k}e_{rk}.$$ The mwtrid representations wre $$M_{{\lfloor{
tj}\rcqul},ss{\left\langue{s+j}\right\rangle}_N}^j=\delta_{ts}\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}ij}.$$ Then, de gec $$\begin{aligbee}
J_{{\nfloor{ ri}\rcxil},s}&=\dejta_{rs},\\
d_{{\lfloor{ ri}\rcehl}}&=1,\\
T_{{\lfkoor{ ri}\rceil}}&=\mabhrm{e}^{-\hrac{2\pi \mathrm{i}}{N}ri},\\
S_{{\lfloor{ ci}\rceil}{\lfloor{ sj}\rceij}}&=\frac{1}{N}\matvrj{e}^{\frac{2\pi \mathrm{i}}{N}\oedt(
rj+sh \richt)}.\eve{alkgntd}$$
### Hinjte grlup *G* model {#fjniteG}
Similqr to the ${\mathbb{Z}}_N$ vafv we can defihe the rjtation-invariant string-net model for a xinjte group $G$:
- $|G|$ types od strings, labeled by jhe group qlements $g\in G$ and $g^*=g^{-1}$.
- The trivial string is now labemdd yn ${\gwxhbf{1}}$, the identity element of $G$.
- $N_{g_1 g_2 g_3}=1$ iff $g_1 d_2 g_3={\uexnbf{1}}$.
- $F_{g_3 g_4 g_6}^{g_1 g_2 g_5}=1$ iff $ | left\langle{ -i}\right\rangle}_N}.$$ The multiplication is $$e_{ri}e_{sj}=\delta_{rs}e_{r{\left\langle{ i+j}\right\rangle}_N}.$$ the sum of group algebras of simple We use ${\lfloor{\cdots}\rceil}$ denote a composite The simple modules can be labeled two numbers ${\lfloor{ ri}\rceil}$. The basis is $$M_{{\lfloor{ ri}\rceil}}^r=\frac{1}{N}\sum_{k=0}^{N-1} \mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}i k}e_{rk}.$$ The representations are $$M_{{\lfloor{ ri}\rceil},ss{\left\langle{s+j}\right\rangle}_N}^j=\delta_{rs}\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}ij}.$$ Then, we get $$\begin{aligned} N_{{\lfloor{ ri}\rceil},s}&=\delta_{rs},\\ d_{{\lfloor{ ri}\rceil}}&=1,\\ ri}\rceil}}&=\mathrm{e}^{-\frac{2\pi S_{{\lfloor{ sj}\rceil}}&=\frac{1}{N}\mathrm{e}^{\frac{2\pi rj+si \right)}.\end{aligned}$$ ### Finite group *G* model {#finiteG} Similar to the ${\mathbb{Z}}_N$ case we can define rotation-invariant string-net model for a finite group $G$: $|G|$ types of strings, by the group elements $g\in and - The string now by ${\textbf{1}}$, the element of $G$. - $N_{g_1 g_2 g_3}=1$ iff $g_1 g_2 g_3={\textbf{1}}$. - $F_{g_3 g_4 g_6}^{g_1 g_2 g_5}=1$ $ | left\langle{ -i}\right\rangle}_N}.$$ ThE multiplicAtion Is $$e_{Ri}e_{Sj}=\DeltA_{rs}e_{R{\left\langle{ i+j}\rIGht\rAngle}_N}.$$
This is the direct sUm of $N$ GrOUp alGEbRas of ${\Mathbb{Z}}_n$. thERE arE $N^2$ 1-DiMenSiONaL simpLe mOdules. WE use ${\lfloor{\CdoTs}\Rceil}$ to denotE A cOmposite laBel. the simple modUleS can be LaBelED with Two NumbeRs ${\lfloOR{ ri}\rceIl}$. The basiS iS $$m_{{\lflooR{ Ri}\rceil}}^R=\FRaC{1}{N}\suM_{k=0}^{N-1}
\mathrm{e}^{-\frac{2\pi \mAThRM{i}}{N}i k}e_{rk}.$$ The matRix repReSEnTATioNs aRe $$M_{{\lfloor{
rI}\rCeil},sS{\Left\lanGLe{S+J}\RIghT\Rangle}_N}^j=\delta_{Rs}\mathrm{e}^{-\frAC{2\pi \Mathrm{I}}{N}Ij}.$$ THEn, we geT $$\begiN{aLIgnEd}
N_{{\lfloor{ ri}\RceiL},s}&=\delta_{rs},\\
D_{{\lflooR{ Ri}\rceil}}&=1,\\
t_{{\Lfloor{ rI}\rceil}}&=\MatHrm{E}^{-\fraC{2\Pi \MaThrM{i}}{n}Ri},\\
S_{{\LFlOor{ RI}\rcEil}{\lflooR{ sJ}\rCeil}}&=\fRac{1}{N}\MATHRm{e}^{\fRac{2\Pi \maThrm{i}}{n}\left(
rj+si \righT)}.\enD{aliGNed}$$
### finitE grouP *G* moDeL {#finiTeG}
SimIlar tO tHe ${\mathbb{Z}}_N$ case wE can Define the RotAtIon-InVariaNT strinG-neT moDel for a Finite gROup $g$:
- $|G|$ TYPEs Of strings, labeled by ThE GRoUp elemenTs $g\in G$ ANd $G^*=g^{-1}$.
- tHe triviaL sTriNg is NOW labeLed bY ${\TeXtbf{1}}$, the iDentitY ElEmEnt of $G$.
- $N_{G_1 g_2 G_3}=1$ iff $g_1 g_2 G_3={\tExtBf{1}}$.
- $F_{G_3 g_4 g_6}^{g_1 g_2 G_5}=1$ Iff $ | left\langle{ -i}\right\ran gle}_N}.$$ Themul tip li cati on i s $$e_{ri}e_{s j }=\d elta_{rs}e_{r{\left\la ngle{ i + j}\r i gh t\ran gle}_N} . $$ Thi sis th ed ir ect s umof $N$group alge bra sof ${\mathbb { Z} }_N$. Ther e a re $N^2$ 1-d ime nsiona lsim p le mo dul es. W e use$ {\lflo or{\cdots }\ r ceil}$ to deno t e a com posite label. The si m ple modules ca n be l ab e le d wit h t wo numbers $ {\lfl o or{ ri} \ rc e i l }$. The basis is$$M_{{\lflo o r{ri}\rc ei l}} ^ r=\fra c{1}{ N} \ sum _{k=0}^{N-1 }
\m athrm{e}^ {-\fra c {2\pi \ m athrm{i }}{N}i k} e_{ rk}. $ $Th e m at r ixr ep res e nta tions ar e$$ M_{{\ lflo o r { ri}\ rce il}, ss{\l eft\langle{s+ j}\ righ t \ra ngle} _N}^j =\de lt a_{rs }\math rm{e} ^{ -\frac{2\pi \ma thrm {i}}{N}ij }.$ $The n, we g e t $$\b egi n{a ligned}
N_{{ \ lfl oo r { ri }\rceil},s}&=\delt a_ { r s} ,\\
d_ {{\lfl o or {r i}\rceil }} &=1 ,\\T_{{\ lflo o r{ ri}\rce il}}&= \ ma th rm{e}^{ -\ frac{2 \p i \ mat hrm{i } }{N} ri},\\
S_{{\lf loor{ ri}\rceil}{\lf l oor{ sj}\rcei l }} & = \f r ac{1 }{N }\mathrm{e} ^{\f r ac{2 \pi\ ma thr m {i}}{ N}\le ft (
r j +si \right)}.\end{a li gned}$ $
## # Finite grou p *G* mode l { #finiteG }
S i mi l ar to the ${\m athbb {Z}}_N$ ca s e we can defi ne the r otation-i n v ariant s tri ng- net mo d e lfor a finiteg r oup$G $:
- $| G|$ typ esofstr ing s, labeledby the g ro up e le men ts $g \ in G$ an d$g^ *= g^{ -1}$.
- T he tr ivia lst r ing is now la b e ledby $ {\te xtb f{ 1}}$, the ide ntity e lement of $G $ .
- $ N_{g_1g_2 g_3}=1$ i ff $g_1 g_2g_ 3={ \textb f { 1}}$.
- $F_{g_3 g_4 g_6}^{g_ 1 g_2 g_ 5}= 1$ if f $ | left\langle{ -i}\right\rangle}_N}.$$_The multiplication_is $$e_{ri}e_{sj}=\delta_{rs}e_{r{\left\langle{ i+j}\right\rangle}_N}.$$
This is_the direct_sum_of $N$_group_algebras of ${\mathbb{Z}}_N$._There are $N^2$_1-dimensional simple modules. We_use ${\lfloor{\cdots}\rceil}$ to_denote_a composite label. The simple modules can be labeled with two numbers ${\lfloor{ ri}\rceil}$._The_basis is_$$M_{{\lfloor{_ri}\rceil}}^r=\frac{1}{N}\sum_{k=0}^{N-1}
\mathrm{e}^{-\frac{2\pi_\mathrm{i}}{N}i k}e_{rk}.$$ The matrix representations_are $$M_{{\lfloor{
ri}\rceil},ss{\left\langle{s+j}\right\rangle}_N}^j=\delta_{rs}\mathrm{e}^{-\frac{2\pi \mathrm{i}}{N}ij}.$$ Then, we_get $$\begin{aligned}
_ N_{{\lfloor{ ri}\rceil},s}&=\delta_{rs},\\
d_{{\lfloor{ ri}\rceil}}&=1,\\
T_{{\lfloor{_ri}\rceil}}&=\mathrm{e}^{-\frac{2\pi_\mathrm{i}}{N}ri},\\
S_{{\lfloor{ ri}\rceil}{\lfloor{ sj}\rceil}}&=\frac{1}{N}\mathrm{e}^{\frac{2\pi_\mathrm{i}}{N}\left(
rj+si \right)}.\end{aligned}$$
### Finite group *G* model {#finiteG}
Similar to the_${\mathbb{Z}}_N$ case we can define the_rotation-invariant string-net model_for_a_finite group $G$:
- _ $|G|$ types of strings, labeled_by the group elements $g\in G$_and $g^*=g^{-1}$.
- The trivial string_is now labeled by ${\textbf{1}}$, the_identity element of $G$.
- _ $N_{g_1_g_2 g_3}=1$ iff $g_1 g_2_g_3={\textbf{1}}$.
- _$F_{g_3 g_4_g_6}^{g_1 g_2 g_5}=1$_iff $ |
}}$ are noted by ${\mathbb{E}}_\delta$ and ${\mathbb{E}}$, respectively.
The Markov chain $(Y_n: n\ge 0)$ can be also constructed from a probability space $({\widetilde{\Omega}}, {\widetilde{\cal F}},{\bf P})$ containing an independent family of random variables $\left(\delta^K_n: K\in {{\cal Y}}({{\cal J}}_\rho), n\ge 1 \right)$, where $\delta^K_n$ takes values in ${\mathbb{D}}_{1,2}(K)$ and $${\bf P}(\delta^K_n=\delta)=f^K_\delta,$$ where $f^K_\delta$ was defined in (\[e15\]). Thus, the random variables $(\delta^K_n: n\ge 1)$ are independent and identically distributed with law $f^K_\bullet$. It is easily checked that the random sequence given by $$Y_0=\delta, \;\; Y_n=(\delta^K_n: K\in Y_{n-1}) \;\, \forall n\ge 1,$$ defines a Markov chain $(Y_n)$ starting from $\delta$, and transition probability given by (\[e23\]).
Let us show that the Markov chain $(Y_n)$ fulfills the first claim of this section: after $n-$steps of time the probability of the event $\{Y_n=\delta\}$ is the weight of all the trees of length $j$ whose set of leaves is $\delta$.
\[lemma5\] For every $n\ge0$ and $\delta\in \partial({{\cal T}})$ it holds ${\mathbb{P}}(Y_n=\delta)=q^n_\delta$.
We will use a recurrence argument. For $n=0$ the property holds because $Y_0=\{I\}$ and the class of trees of length $0$ is the singleton ${{\cal T}}_0=\{\{I\}\}$. Assume the property holds up to $n$ let us show it for $n+1$. We have $$\begin{aligned}
{\mathbb{P}}(Y_{n+1}=\delta')&=&
\sum_{\delta\in \partial({{\cal T}}): \delta \rightarrow \delta'}
{\mathbb{P}}(Y_n=\delta) P_{\delta,\delta'}\\
&=&\sum_{\delta\in \partial({{\cal T}}): \delta \rightarrow \delta'}
q^n | } } $ are noted by $ { \mathbb{E}}_\delta$ and $ { \mathbb{E}}$, respectively.
The Markov chain $ (Y_n: n\ge 0)$ can be also constructed from a probability quad $ ({ \widetilde{\Omega } }, { \widetilde{\cal F}},{\bf P})$ control an independent family of random variables $ \left(\delta^K_n: K\in { { \cal Y}}({{\cal J}}_\rho), n\ge 1 \right)$, where $ \delta^K_n$ take values in $ { \mathbb{D}}_{1,2}(K)$ and $ $ { \bf P}(\delta^K_n=\delta)=f^K_\delta,$$ where $ f^K_\delta$ was defined in (\[e15\ ]). therefore, the random variables $ (\delta^K_n: n\ge 1)$ are independent and identically distributed with jurisprudence $ f^K_\bullet$. It is easily checked that the random succession given by $ $ Y_0=\delta, \;\; Y_n=(\delta^K_n: K\in Y_{n-1 }) \;\, \forall n\ge 1,$$ defines a Markov chain $ (Y_n)$ starting from $ \delta$, and passage probability given by (\[e23\ ]).
Let us show that the Markov chain $ (Y_n)$ fulfills the first title of this section: after $ n-$steps of time the probability of the event $ \{Y_n=\delta\}$ is the weight unit of all the trees of length $ j$ whose set of leaves is $ \delta$.
\[lemma5\ ] For every $ n\ge0 $ and $ \delta\in \partial({{\cal T}})$ it hold $ { \mathbb{P}}(Y_n=\delta)=q^n_\delta$.
We will use a recurrence argument. For $ n=0 $ the property holds because $ Y_0=\{I\}$ and the class of tree of length $ 0 $ is the singleton $ { { \cal T}}_0=\{\{I\}\}$. Assume the property holds up to $ n$ let us show it for $ n+1$. We have $ $ \begin{aligned }
{ \mathbb{P}}(Y_{n+1}=\delta')&= &
\sum_{\delta\in \partial({{\cal T } } ): \delta \rightarrow \delta' }
{ \mathbb{P}}(Y_n=\delta) P_{\delta,\delta'}\\
& = & \sum_{\delta\in \partial({{\cal T } } ): \delta \rightarrow \delta' }
q^n | }}$ arf noted by ${\mathbb{E}}_\delta$ xnd ${\mathbb{E}}$, reskextivelb.
The Madkov chakn $(Y_n: n\ge 0)$ can be also constcuctwd frim a probability space $({\widetildv{\Omega}}, {\wieetioee{\cal F}},{\bf '})$ contaiknng ah indzpxndent family on random vasiables $\left(\dentx^K_u: K\in {{\cal Y}}({{\cal J}}_\rho), n\ge 1 \right)$, where $\delta^K_m$ hakes values ig ${\mauhbf{D}}_{1,2}(K)$ znd $${\bf P}(\delta^K_n=\delta)=f^K_\delta,$$ where $r^K_\delta$ was defined on (\[e15\]). Thus, the random variahles $(\delta^K_n: n\ge 1)$ are lndependent and udentically aistributed with law $f^I_\bullet$. It is easily checked thxt thz random sewuwncf given by $$Y_0=\velta, \;\; Y_n=(\delta^K_n: K\lm Y_{n-1}) \;\, \forall n\ge 1,$$ defines s Merkoc chain $(Y_n)$ starting fcom $\delta$, and transijion probatimity given by (\[e23\]).
Ler ys shmw tvat gye Oariot cgain $(Y_j)$ fnlfills the first clain of this section: agtqg $n-$steps of tjme thq [robability of the event $\{Y_n=\delta\}$ is the wejght of all the trees od length $j$ whose set lf leaves is $\delta$.
\[lemma5\] For every $n\ge0$ and $\delta\in \partial({{\cdl T}})$ mt hoodf ${\nahhbb{P}}(Y_n=\delta)=q^n_\delta$.
We will use a recurrence awfukekt. For $n=0$ the proierty holds becausr $J_0=\{I\}$ wnd the class of trzss of length $0$ is the singlejon ${{\cao T}}_0=\{\{I\}\}$. Assuie tne property holds up to $n$ lwt us show in foe $n+1$. We have $$\begin{apigned}
{\mathby{P}}(Y_{n+1}=\dekta')&=&
\sim_{\delta\in \partial({{\cal T}}): \belta \dightarrow \felta'}
{\mathgc{P}}(Y_n=\delta) P_{\delta,\aelna'}\\
&=&\suk_{\delta\in \partial({{\cal T}}): \deltw \rightarcow \dzlta'}
q^n | }}$ are noted by ${\mathbb{E}}_\delta$ and ${\mathbb{E}}$, Markov $(Y_n: n\ge can be also $({\widetilde{\Omega}}, F}},{\bf P})$ containing independent family of variables $\left(\delta^K_n: K\in {{\cal Y}}({{\cal J}}_\rho), 1 \right)$, where $\delta^K_n$ takes values in ${\mathbb{D}}_{1,2}(K)$ and $${\bf P}(\delta^K_n=\delta)=f^K_\delta,$$ where $f^K_\delta$ defined in (\[e15\]). Thus, the random variables $(\delta^K_n: n\ge 1)$ are independent and distributed law It easily checked that the random sequence given by $$Y_0=\delta, \;\; Y_n=(\delta^K_n: K\in Y_{n-1}) \;\, \forall n\ge defines a Markov chain $(Y_n)$ starting from $\delta$, transition probability given by Let us show that the chain fulfills the claim this after $n-$steps of the probability of the event $\{Y_n=\delta\}$ is the weight of all the trees of length $j$ whose of leaves \[lemma5\] For $n\ge0$ $\delta\in T}})$ it holds will use a recurrence argument. For holds because $Y_0=\{I\}$ and the class of trees length $0$ the singleton ${{\cal T}}_0=\{\{I\}\}$. Assume the holds up to $n$ let us show it $n+1$. We have $$\begin{aligned} {\mathbb{P}}(Y_{n+1}=\delta')&=& \sum_{\delta\in \partial({{\cal T}}): \delta \rightarrow \delta'} {\mathbb{P}}(Y_n=\delta) P_{\delta,\delta'}\\ &=&\sum_{\delta\in \partial({{\cal \rightarrow \delta'} q^n | }}$ are noted by ${\mathbb{E}}_\delta$ and ${\Mathbb{E}}$, resPectiVelY.
ThE MArkoV chaIn $(Y_n: n\ge 0)$ can be alSO conStructed from a probabiliTy spaCe $({\WIdetILdE{\OmegA}}, {\widetiLDe{\CAL F}},{\bF P})$ CoNtaInINg An indEpeNdent faMily of randOm vArIables $\left(\deLTa^k_n: K\in {{\cal Y}}({{\cAl J}}_\Rho), n\ge 1 \right)$, wHerE $\delta^k_n$ TakES valuEs iN ${\mathBb{D}}_{1,2}(K)$ anD $${\Bf P}(\delTa^K_n=\delta)=F^K_\DElta,$$ whERe $f^K_\delTA$ WaS defIned in (\[e15\]). Thus, the ranDOm VAriables $(\delta^K_N: n\ge 1)$ arE iNDePENdeNt aNd identicaLlY distRIbuted wITh LAW $F^K_\bULlet$. It is easilY checked thaT The Random SeQueNCe giveN by $$Y_0=\dElTA, \;\; Y_n=(\Delta^K_n: K\in Y_{N-1}) \;\, \forAll n\ge 1,$$ defInes a MARkov chaIN $(Y_n)$ starTing frOm $\dEltA$, and TRaNsItiOn PRobABiLitY GivEn by (\[e23\]).
Let Us ShOw thaT the mARKOv chAin $(y_n)$ fuLfillS the first claiM of This SEctIon: afTer $n-$sTeps Of Time tHe probAbiliTy Of the event $\{Y_n=\delTa\}$ is The weight Of aLl The TrEes of LEngth $j$ WhoSe sEt of leaVes is $\deLTa$.
\[lEmMA5\] fOr Every $n\ge0$ and $\delta\in \PaRTIaL({{\cal T}})$ it hOlds ${\maTHbB{P}}(y_N=\delta)=q^n_\DeLta$.
we wiLL Use a rEcurREnCe argumeNt. For $n=0$ THe PrOperty hOlDs becaUsE $Y_0=\{I\}$ And The clASs of Trees oF length $0$ iS the sINgleton ${{\cal T}}_0=\{\{I\}\}$. AsSUme the propertY HoLDS uP To $n$ lEt uS show it for $n+1$. we haVE $$\begIn{alIGnEd}
{\mAThbb{P}}(y_{n+1}=\delTa')&=&
\SUm_{\DElta\in \partial({{\cal T}}): \deLtA \rightArrow \Delta'}
{\mathbb{P}}(Y_N=\delta) P_{\delTA,\DElta'}\\
&=&\sum_{\dElta\IN \pARtial({{\cal T}}): \delta \RightArrow \delta'}
Q^N | }}$ are noted by ${\mathbb {E}}_\delt a$ an d $ {\m at hbb{ E}}$ , respectively .
Th e Markov chain $(Y_n:n\ge0) $ can be also constr u ct e d fr om a pr ob a bi lityspa ce $({\ widetilde{ \Om eg a}}, {\widet i ld e{\cal F}} ,{\ bf P})$ cont ain ing an i nde p enden t f amily of ra n dom va riables $ \l e ft(\de l ta^K_n: K \i n {{ \cal Y}}({{\cal J } }_ \ rho), n\ge 1 \ right) $, wh e r e $ \de lta^K_n$ t ak es va l ues in$ {\ m a t hbb { D}}_{1,2}(K)$ and $${\bf P}( \delta ^K _n= \ delta) =f^K_ \d e lta ,$$ where $ f^K_ \delta$ w as def i ned in( \[e15\] ). Thu s,the ran d om v ari ab l es$ (\ del t a^K _n: n\ge 1 )$ areinde p e n d entand ide ntica lly distribut edwith law $f^K _\bul let$ .It is easil y che ck ed that the ran domsequencegiv en by $ $Y_0= \ delta, \; \;Y_n=(\d elta^K_ n : K \i n Y _{ n-1}) \;\, \foral ln \ ge 1,$$ de finesa M ar k ov chain $ (Y_ n)$s t artin g fr o m$\delta$ , andt ra ns ition p ro babili ty gi ven by ( \ [e23 \]).
Let us s how t h at the Markovc hain $(Y_n)$f ul f i ll s the fi rst claim o f th i s se ctio n :aft e r $n- $step so ft ime the probability o f theevent $\{Y_n=\delt a\}$ is th e w eight of all th e trees of leng th $j $ whose se t of leav es is $\delta $.
\[lem m a 5\] Foreve ry$n\ ge0 $ an d $\delta\in\ p arti al ({{\cal T} })$ ithol ds${\ mat hb b{P}}(Y_n =\delta) =q ^n _\ de lta $.
W e will us ea r ec urr encea rgumen t. Fo r $n =0 $t hepropert y h o l ds b ec au se $ Y_0 =\ {I\}$ and the classof treesofl engt h$0 $ is th e singleton $ {{ \cal T}}_0 =\ {\{ I\}\}$ . Assume t he property holds up to $n$ let us show itfor $n+1$ . W e have $$ \ begin{ aligne d}
{\ ma thb b { P}}(Y _ { n+ 1}= \d elta')&=&\ sum _{\de lt a\in \parti al({{\cal T}}): \d e lta \rightarrow\de lta' } {\ mat h bb { P}} (Y _ n=\ d e lta) P_{\delta, \delta'}\\
& = &\ sum_{\delt a \in \ partial ({{\cal T}}) : \delta \rightar row \delt a' }
q^ n | }}$ are_noted by_${\mathbb{E}}_\delta$ and ${\mathbb{E}}$, respectively.
The_Markov chain_$(Y_n:_n\ge 0)$_can_be also constructed_from a probability_space $({\widetilde{\Omega}}, {\widetilde{\cal F}},{\bf_P})$ containing an_independent_family of random variables $\left(\delta^K_n: K\in {{\cal Y}}({{\cal J}}_\rho), n\ge 1 \right)$, where $\delta^K_n$_takes_values in_${\mathbb{D}}_{1,2}(K)$_and_$${\bf P}(\delta^K_n=\delta)=f^K_\delta,$$ where $f^K_\delta$ was_defined in (\[e15\]). Thus, the_random variables_$(\delta^K_n: n\ge 1)$ are independent and identically distributed_with_law $f^K_\bullet$. It_is easily checked that the random sequence given by_$$Y_0=\delta, \;\; Y_n=(\delta^K_n: K\in Y_{n-1}) _\;\, \forall n\ge_1,$$_defines_a Markov chain $(Y_n)$_starting from $\delta$, and transition probability_given by (\[e23\]).
Let us show that_the Markov chain $(Y_n)$ fulfills the first_claim of this section: after $n-$steps_of time the probability of_the event_$\{Y_n=\delta\}$ is the weight of_all the trees_of length_$j$ whose set_of leaves is $\delta$.
\[lemma5\] For every_$n\ge0$ and $\delta\in_\partial({{\cal T}})$ it holds ${\mathbb{P}}(Y_n=\delta)=q^n_\delta$.
We will_use_a recurrence argument._For_$n=0$_the property_holds because $Y_0=\{I\}$_and_the class_of_trees of length $0$ is the_singleton_${{\cal T}}_0=\{\{I\}\}$. Assume the property holds up_to $n$ let us_show_it for $n+1$. We_have $$\begin{aligned}
{\mathbb{P}}(Y_{n+1}=\delta')&=&
\sum_{\delta\in \partial({{\cal T}}):_\delta \rightarrow \delta'}
{\mathbb{P}}(Y_n=\delta) P_{\delta,\delta'}\\
&=&\sum_{\delta\in \partial({{\cal T}}):_\delta \rightarrow_\delta'}
q^n |
= 2\sqrt{\rho_R}$, $\rho$ can vanish (there is a so-called vacuum point) and a modification of the solution is required. To find a simple wave solution for large piston velocities, we must derive new conditions for the parameters $r_i$. We modify the DSW solution by introducing a locally periodic TW between the piston and the trailing edge of the DSW. When $v_p = 2\sqrt{\rho_R}$, the DSW forms a vacuum point [@El1995; @Hoefer2006] at the piston. The *vacuum condition*, $\rho = 0$, is satisfied in eq. when $$\label{eq:13}
\lambda_1 = 0 ~ \Rightarrow ~ r_1-r_2-r_3+r_4 = 0 ~ \Rightarrow ~
r_4 - r_3 = 4\sqrt{\rho_R}.$$ Note that the fluid velocity is undefined at a vacuum point, even though the vacuum points have a well-defined propagation speed through the fluid. We assume that this condition holds for $v_p >
2\sqrt{\rho_R}$ as well. One more condition is required to completely determine $r_4$ and $r_3(x=0,t=0^+)$; $r_1$, $r_2$, and $r_3(x>0,0^+)$ are determined by the initial data. Because there is a locally periodic TW between the piston and DSW, we assume that the velocity of the TW equals the piston velocity. This is the *TW velocity condition* $$\label{eq:14}
V = v_p ~ \Rightarrow ~ r_1+r_2+r_3 + r_4 = 4 v_p ~ \Rightarrow ~
r_3 + r_4 = 4 v_p.$$ Given the initial data in eq. and the two conditions and, only $r_3$ and $\sigma$ in the initial data of eq. are altered $$r_3(x,0^+) = 2v_p \mp 2\sqrt{\rho_R}, ~
\sigma(x,0^+) = \mp 1, ~ x \genfrac{}{}{0pt}{2}{=}{>} 0.$$ Note that $\sigma = -1$ in the locally periodic region. Figure \[fig:piston\_dsw\], right | = 2\sqrt{\rho_R}$, $ \rho$ can vanish (there is a so - called void item) and a modification of the solution is required. To witness a bare wave solution for large piston velocities, we must deduce new conditions for the parameters $ r_i$. We change the DSW solution by inaugurate a locally periodic TW between the piston and the trail edge of the DSW. When $ v_p = 2\sqrt{\rho_R}$, the DSW forms a vacuum distributor point [ @El1995; @Hoefer2006 ] at the piston. The * vacuum condition *, $ \rho = 0 $, is satisfied in eq. when $ $ \label{eq:13 }
\lambda_1 = 0 ~ \Rightarrow ~ r_1 - r_2 - r_3+r_4 = 0 ~ \Rightarrow ~
r_4 - r_3 = 4\sqrt{\rho_R}.$$ Note that the fluid speed is undefined at a vacuum point, even though the vacuum point have a well - defined generation speed through the fluid. We simulate that this condition have for $ v_p >
2\sqrt{\rho_R}$ as well. One more condition is required to wholly determine $ r_4 $ and $ r_3(x=0,t=0^+)$; $ r_1 $, $ r_2 $, and $ r_3(x>0,0^+)$ are determined by the initial data. Because there is a locally periodic TW between the piston and DSW, we assume that the velocity of the TW equal the piston velocity. This is the * TW velocity condition * $ $ \label{eq:14 }
V = v_p ~ \Rightarrow ~ r_1+r_2+r_3 + r_4 = 4 v_p ~ \Rightarrow ~
r_3 + r_4 = 4 v_p.$$ Given the initial data in eq. and the two conditions and, only $ r_3 $ and $ \sigma$ in the initial data of eq. are altered $ $ r_3(x,0^+) = 2v_p \mp 2\sqrt{\rho_R }, ~
\sigma(x,0^+) = \mp 1, ~ x \genfrac{}{}{0pt}{2}{= } { > } 0.$$ notice that $ \sigma = -1 $ in the locally periodic area. Figure \[fig: piston\_dsw\ ], right | = 2\sert{\rho_R}$, $\rho$ can vanish (tmere is a so-callgd vacuuk poinf) and a oodification of the solution iw reqyired. To find a simple wave solltion for larje piston velocivjes, we must dsvive uex conditions fot the parameders $r_i$. We modhfh che DSW solution by introducing a losally prrlodic TW betwegn tht pystoh and the trailing edge of the DSW. When $v_k = 2\sqrt{\rho_R}$, the DSE forms a vacuum point [@El1995; @Joeffr2006] at the piston. Tje *vacuum cindieuon*, $\rho = 0$, is satisfied in eq. when $$\mabel{eq:13}
\lambda_1 = 0 ~ \Rightarrow ~ f_1-r_2-r_3+r_4 = 0 ~ \Rightartos ~
r_4 - r_3 = 4\sqrt{\cho_R}.$$ Njte that the fluid valocity is undefined st e vaxuum point, even thougi the vacuum points rave a wenl-befined propagation spewd thtough the dlukd. Se aasume hhav this condjtion holds for $v_p >
2\sqrt{\rho_R}$ as wqol. One more cknditijn is required to completely determine $r_4$ dnd $r_3(x=0,t=0^+)$; $r_1$, $r_2$, and $r_3(x>0,0^+)$ are detwrmined by the initiap data. Besause there is a locally periodic TW between the [istoi xnd DFD, wf assume that the velocity of the TW equals trs kisnon velocity. This is the *TW fepovyty condition* $$\label{zs:14}
V = v_p ~ \Rightarrow ~ r_1+r_2+r_3 + t_4 = 4 v_p ~ \Rightarwow ~
r_3 + r_4 = 4 v_p.$$ Given the initiql data in ez. and the two condicions and, onuy $r_3$ and $\xigma$ in the initial daca of sq. are alteged $$r_3(x,0^+) = 2v_l \mp 2\sqrt{\rho_R}, ~
\skgms(x,0^+) = \mp 1, ~ x \genfrac{}{}{0pt}{2}{=}{>} 0.$$ Note trat $\sigma = -1$ iu the lozalli perioqic region. Figuva \[fig:piston\_dsw\], rigjt | = 2\sqrt{\rho_R}$, $\rho$ can vanish (there is vacuum and a of the solution simple solution for large velocities, we must new conditions for the parameters $r_i$. modify the DSW solution by introducing a locally periodic TW between the piston the trailing edge of the DSW. When $v_p = 2\sqrt{\rho_R}$, the DSW forms vacuum [@El1995; at piston. The *vacuum condition*, $\rho = 0$, is satisfied in eq. when $$\label{eq:13} \lambda_1 = 0 \Rightarrow ~ r_1-r_2-r_3+r_4 = 0 ~ \Rightarrow ~ - r_3 = 4\sqrt{\rho_R}.$$ that the fluid velocity is at vacuum point, though vacuum have a well-defined speed through the fluid. We assume that this condition holds for $v_p > 2\sqrt{\rho_R}$ as well. One condition is completely determine and $r_1$, and $r_3(x>0,0^+)$ are the initial data. Because there is TW between the piston and DSW, we assume the velocity the TW equals the piston velocity. is the *TW velocity condition* $$\label{eq:14} V = ~ \Rightarrow ~ r_1+r_2+r_3 + r_4 = 4 v_p ~ \Rightarrow ~ r_3 + r_4 v_p.$$ Given the initial in eq. and two and, $r_3$ $\sigma$ in initial data of eq. are altered $$r_3(x,0^+) = 2v_p \mp 2\sqrt{\rho_R}, \sigma(x,0^+) = \mp 1, ~ x \genfrac{}{}{0pt}{2}{=}{>} 0.$$ Note that -1$ the locally periodic Figure \[fig:piston\_dsw\], right | = 2\sqrt{\rho_R}$, $\rho$ can vanish (there Is a so-calleD vacuUm pOinT) aNd a mOdifIcation of the soLUtioN is required. To find a simpLe wavE sOLutiON fOr larGe pistoN VeLOCitIeS, wE muSt DErIve neW coNditionS for the parAmeTeRs $r_i$. We modify THe dSW solutioN by Introducing a LocAlly peRiOdiC tW betWeeN the pIston aND the trAiling edgE oF The DSW. wHen $v_p = 2\sqRT{\RhO_R}$, thE DSW forms a vacuum pOInT [@el1995; @Hoefer2006] at the pIston. THe *VAcUUM coNdiTion*, $\rho = 0$, is sAtIsfieD In eq. wheN $$\LaBEL{Eq:13}
\lAMbda_1 = 0 ~ \RightarroW ~ r_1-r_2-r_3+r_4 = 0 ~ \RightaRRow ~
R_4 - r_3 = 4\sqrt{\RhO_R}.$$ NOTe that The flUiD VelOcity is undeFineD at a vacuuM point, EVen thouGH the vacUum poiNts HavE a weLL-dEfIneD pROpaGAtIon SPeeD through ThE fLuid. WE assUME THat tHis CondItion Holds for $v_p >
2\sqrT{\rhO_R}$ as WEll. one moRe conDitiOn Is reqUired tO compLeTely determine $r_4$ aNd $r_3(x=0,T=0^+)$; $r_1$, $r_2$, and $r_3(x>0,0^+)$ aRe dEtErmInEd by tHE initiAl dAta. because There is A LocAlLY PErIodic TW between the pIsTON aNd DSW, we aSsume tHAt ThE Velocity Of The tW eqUALs the PistON vElocity. THis is tHE *Tw vElocity CoNditioN* $$\lAbeL{eq:14}
v = v_p ~ \RiGHtarRow ~ r_1+r_2+r_3 + R_4 = 4 v_p ~ \RightArrow ~
R_3 + R_4 = 4 v_p.$$ Given the iniTIal data in eq. anD ThE TWo COndiTioNs and, only $r_3$ aNd $\siGMa$ in The iNItIal DAta of Eq. are AlTErED $$r_3(x,0^+) = 2v_p \mp 2\sqrt{\rho_R}, ~
\sigmA(x,0^+) = \Mp 1, ~ x \genFrac{}{}{0pT}{2}{=}{>} 0.$$ Note that $\sigmA = -1$ in the locaLLY Periodic RegiON. FIGure \[fig:piston\_dSw\], rigHt | = 2\sqrt{\rho_R}$, $\rho$ can vanis h (th ere is a so- call ed vacuum poin t ) an d a modification of th e sol ut i on i s r equir ed. Tof in d a s im pl e w av e s oluti onfor lar ge pistonvel oc ities, we mu s tderive new co nditions for th e para me ter s $r_i $.We mo dify t h e DSWsolutionby introd u cing al o ca llyperiodic TW betwe e nt he piston andthe tr ai l in g edg e o f the DSW. W hen $ v _p = 2\ s qr t { \ rho _ R}$, the DSWforms a vac u umpoint[@ El1 9 95; @H oefer 20 0 6]at the pist on.The *vacu um con d ition*, $\rho = 0$, i s s ati sfie d i neq. w hen $$ \la b el{ eq:13}
\ la mbda_ 1 =0 ~ \Rig hta rrow ~ r_ 1-r_2-r_3+r_4 =0 ~\ Rig htarr ow ~ r_ 4- r_3 = 4\s qrt{\ rh o_R}.$$ Note th at t he fluidvel oc ity i s und e finedata v acuum p oint, e v enth o u g hthe vacuum pointsha v e a well-de finedp ro pa g ation sp ee d t hrou g h theflui d .We assum e that th is condit io n hold sfor $v _p >2 \sqr t{\rho _R}$ aswell. One more condi t ion is requir e dt o c o mple tel y determine $r_ 4 $ an d $r _ 3( x=0 , t=0^+ )$; $ r_ 1 $, $r_2$, and $r_3(x>0 ,0 ^+)$ a re de termined by t he initial d a ta. Beca uset he r e is a locally peri odic TW be t ween the pist on and D SW, we as s u me thatthe ve loc ity o fthe TW equals t he p is ton vel oci ty. Thi s i s t he*TW v elocity c ondition *$$ \l ab el{ eq:14 }
V = v _p ~\R igh tarro w ~ r_1 +r_2+ r_3+r_ 4 =4 v_p ~ \R i g htar ro w~
r_3 + r_4= 4v _p. $$ Give n the ini tia l dat ain eq. a nd the two co nd itions and ,onl y $r_3 $ and $\si gma$ in the initial dat a of eq. a re al tere d $$r_3(x ,0^ +) = 2 v_p \mp 2\ sqrt{\ rho_R }, ~\sigm a ( x, 0^+ )= \mp 1, ~ x \g enfra c{ }{}{ 0pt}{2} {=}{>} 0.$$ Note t h at$\sigma = -1$ in the l oc all y p e rio di c re g i on. Figure \[fi g:piston\_ ds w \] , right | =_2\sqrt{\rho_R}$, $\rho$_can vanish (there is_a so-called_vacuum_point) and_a_modification of the_solution is required._To find a simple_wave solution for_large_piston velocities, we must derive new conditions for the parameters $r_i$. We modify the_DSW_solution by_introducing_a_locally periodic TW between the_piston and the trailing edge_of the_DSW. When $v_p = 2\sqrt{\rho_R}$, the DSW forms_a_vacuum point [@El1995;_@Hoefer2006] at the piston. The *vacuum condition*, $\rho =_0$, is satisfied in eq. when_$$\label{eq:13}
\lambda_1_=_0_~ \Rightarrow ~ r_1-r_2-r_3+r_4_= 0 ~ \Rightarrow ~
_r_4 - r_3 = 4\sqrt{\rho_R}.$$ Note_that the fluid velocity is undefined at_a vacuum point, even though the_vacuum points have a well-defined_propagation speed_through the fluid. We assume_that this condition_holds for_$v_p >
2\sqrt{\rho_R}$ as_well. One more condition is required_to completely determine_$r_4$ and $r_3(x=0,t=0^+)$; $r_1$, $r_2$, and_$r_3(x>0,0^+)$_are determined by_the_initial_data. Because_there is a_locally_periodic TW_between_the piston and DSW, we assume_that_the velocity of the TW equals the_piston velocity. This is_the_*TW velocity condition* $$\label{eq:14}
_ V = v_p ~_\Rightarrow ~ r_1+r_2+r_3 + r_4 =_4 v_p_~ \Rightarrow_~
r_3 + r_4 = 4 v_p.$$ Given the initial_data in eq. and the two_conditions and, only $r_3$_and $\sigma$_in_the initial data_of_eq. are_altered $$r_3(x,0^+) = 2v_p \mp 2\sqrt{\rho_R}, ~
_ \sigma(x,0^+)_= \mp 1, ~ x \genfrac{}{}{0pt}{2}{=}{>}_0.$$ Note that $\sigma_=_-1$ in the locally periodic region._Figure \[fig:piston\_dsw\], right |
model.
Then, we obtain an equation for $x$ and $y$: $$F(x)=\int_0^x\frac{f(y)dy}{\sqrt{x-y}},$$ where $$F(x)=T\frac{\Delta\omega}{\omega}\frac{1}{\sqrt{x}},$$ $$f(y)=\frac{\Delta c}{c}\frac{1}{2y^{3/2}\left(\dfrac{d\log c}{d\log r}+1\right)}.$$
To solve for $f(y)$ we multiply both sides of Eq.(12) by ${dx}/{\sqrt{z-x}}$ and integrate with respect to $x$ from 0 to $z$: $$\int_0^z\frac{F(x)dx}{\sqrt{z-x}}=\int_0^z\frac{dx}{\sqrt{z-x}}
\int_0^x\frac{f(y)dy}{\sqrt{x-y}}=$$ $$=\int_0^x f(y)dy\int_y^z\frac{dx}{\sqrt{(z-x)(x-y)}}.$$ Here we changed the order of integration.
Note that $$\int_y^z\frac{dx}{\sqrt{(z-x)(x-y)}}=\pi,$$ then $$\int_0^z\frac{F(x)dx}{\sqrt{z-x}}=\pi\int_0^x f(y)dy.$$ Differentiating with respect to $x$, we obtain the final solution: $$f(y)=\frac{1}{\pi}\frac{d}{dx}\int_0^z\frac{F(x)dx}{\sqrt{z-x}}.$$ Then, from $f(y)$ we find the sound-speed correction $\Delta c/c$.
This method based on linearization of the asymptotic Abel integral is called “differential asymptotic sound-speed inversion” [@Christensen-Dalsgaard1988]. It provides estimates of the sound-speed deviations from a reference solar model.
Alternatively, the sound-speed profile inside the Sun can be found from a implicit solution of the Abel obtained by differentiating the Duvall’s law equation (\[eq-osc26\]) with respect to variable $y=L/\omega$. Then, this equation can be solved analytically. The solution provides an implicit relationship between the solar radius and sound speed [@Gough1986 | model.
Then, we obtain an equation for $ x$ and $ y$: $ $ F(x)=\int_0^x\frac{f(y)dy}{\sqrt{x - y}},$$ where $ $ F(x)=T\frac{\Delta\omega}{\omega}\frac{1}{\sqrt{x}},$$ $ $ f(y)=\frac{\Delta c}{c}\frac{1}{2y^{3/2}\left(\dfrac{d\log c}{d\log r}+1\right)}.$$
To solve for $ f(y)$ we breed both slope of Eq.(12) by $ { dx}/{\sqrt{z - x}}$ and integrate with respect to $ x$ from 0 to $ z$: $ $ \int_0^z\frac{F(x)dx}{\sqrt{z - x}}=\int_0^z\frac{dx}{\sqrt{z - x } }
\int_0^x\frac{f(y)dy}{\sqrt{x - y}}=$$ $ $ = \int_0^x f(y)dy\int_y^z\frac{dx}{\sqrt{(z - x)(x - y)}}.$$ Here we change the order of integration.
notice that $ $ \int_y^z\frac{dx}{\sqrt{(z - x)(x - y)}}=\pi,$$ then $ $ \int_0^z\frac{F(x)dx}{\sqrt{z - x}}=\pi\int_0^x f(y)dy.$$ Differentiating with respect to $ x$, we obtain the concluding solution: $ $ f(y)=\frac{1}{\pi}\frac{d}{dx}\int_0^z\frac{F(x)dx}{\sqrt{z - x}}.$$ Then, from $ f(y)$ we find oneself the legal - speed correction $ \Delta c / c$.
This method free-base on linearization of the asymptotic Abel integral is called “ differential asymptotic sound - speed inversion ” [ @Christensen - Dalsgaard1988 ]. It supply estimates of the sound - speed deviation from a reference solar model.
Alternatively, the sound - speed visibility inside the Sun can be found from a implicit solution of the Abel receive by differentiating the Duvall ’s law equation (\[eq - osc26\ ]) with deference to variable $ y = L/\omega$. Then, this equation can be solved analytically. The solution provides an implicit relationship between the solar spoke and healthy speed [ @Gough1986 | mofel.
Then, we obtain an equxtion for $x$ and $y$: $$F(x)=\inv_0^x\frac{f(g)dy}{\sqrt{x-h}},$$ where $$F(x)=T\frac{\Delta\omega}{\omeja}\frqc{1}{\sqru{q}},$$ $$f(y)=\frac{\Delta c}{c}\frac{1}{2y^{3/2}\ueft(\dfrac{f\log c}{d\lig r}+1\cight)}.$$
To solve foc $f(y)$ we multipmn botk wides of Eq.(12) by ${dx}/{\sqrt{z-x}}$ dnd integrate figh respect to $x$ from 0 to $z$: $$\int_0^z\frac{F(x)qx}{\sqrt{z-c}}=\ijt_0^z\frac{dx}{\sqrt{z-v}}
\int_0^q\fwac{f(g)dy}{\sqrt{x-y}}=$$ $$=\int_0^x f(y)dy\int_y^z\frac{dx}{\sqrt{(a-x)(x-y)}}.$$ Hege we changed the order of integration.
Note hhat $$\int_y^z\frac{dx}{\sqrt{(z-x)(d-y)}}=\pi,$$ then $$\ibt_0^z\fwqc{F(x)dx}{\sqrt{z-x}}=\oi\int_0^x f(y)dj.$$ Differentjating with respect to $x$, we obtxin tke final soourioj: $$f(y)=\frac{1}{\pi}\frec{d}{dx}\igt_0^z\frac{F(x)dx}{\sqrt{z-x}}.$$ Than, from $f(y)$ we find thc souid-spwed correction $\Delta r/c$.
This method based jn linearhzction of the asymptotuc Abel intagrau is camlxd “sifferfntmal asymptofic sound-spwed inversion” [@Chrisuenfvm-Dalsgaard1988]. If provydqs estimates of the sound-speed deviatiots rrom a reference solar nodel.
Alternatively, thg sound-speqd profile inside the Sun can be found from a impnicit rolbblon ud hhe Abel obtained by differentiating the Duvajm’s lsw equation (\[eq-jsc26\]) with rexpfcy to variable $i=L/\omega$. Thsn, this equation cwn be sjlved analyticwlly. The solution provides an inplicit relanionwhip between the sllar radius and sounc speed [@Gough1986 | model. Then, we obtain an equation for $y$: where $$F(x)=T\frac{\Delta\omega}{\omega}\frac{1}{\sqrt{x}},$$ c}{c}\frac{1}{2y^{3/2}\left(\dfrac{d\log c}{d\log r}+1\right)}.$$ multiply sides of Eq.(12) ${dx}/{\sqrt{z-x}}$ and integrate respect to $x$ from 0 to $$\int_0^z\frac{F(x)dx}{\sqrt{z-x}}=\int_0^z\frac{dx}{\sqrt{z-x}} \int_0^x\frac{f(y)dy}{\sqrt{x-y}}=$$ $$=\int_0^x f(y)dy\int_y^z\frac{dx}{\sqrt{(z-x)(x-y)}}.$$ Here we changed the order of integration. Note that then $$\int_0^z\frac{F(x)dx}{\sqrt{z-x}}=\pi\int_0^x f(y)dy.$$ Differentiating with respect to $x$, we obtain the final solution: Then, $f(y)$ find sound-speed correction $\Delta c/c$. This method based on linearization of the asymptotic Abel integral is called asymptotic sound-speed inversion” [@Christensen-Dalsgaard1988]. It provides estimates of sound-speed deviations from a solar model. Alternatively, the sound-speed inside Sun can found a solution of the obtained by differentiating the Duvall’s law equation (\[eq-osc26\]) with respect to variable $y=L/\omega$. Then, this equation can solved analytically. provides an relationship the radius and sound | model.
Then, we obtain an equatiOn for $x$ and $y$: $$f(x)=\int_0^X\frAc{f(Y)dY}{\sqrT{x-y}},$$ wHere $$F(x)=T\frac{\DelTA\omeGa}{\omega}\frac{1}{\sqrt{x}},$$ $$f(y)=\frac{\delta C}{c}\FRac{1}{2y^{3/2}\LEfT(\dfraC{d\log c}{d\LOg R}+1\RIghT)}.$$
TO sOlvE fOR $f(Y)$ we muLtiPly both Sides of Eq.(12) bY ${dx}/{\SqRt{z-x}}$ and integRAtE with respeCt tO $x$ from 0 to $z$: $$\int_0^Z\frAc{F(x)dx}{\SqRt{z-X}}=\Int_0^z\fRac{Dx}{\sqrT{z-x}}
\int_0^X\Frac{f(y)Dy}{\sqrt{x-y}}=$$ $$=\iNt_0^X F(y)dy\inT_Y^z\frac{dX}{\SQrT{(z-x)(x-Y)}}.$$ Here we changed the ORdER of integration.
note thAt $$\INt_Y^Z\FraC{dx}{\Sqrt{(z-x)(x-y)}}=\pi,$$ ThEn $$\int_0^Z\Frac{F(x)dX}{\SqRT{Z-X}}=\pi\INt_0^x f(y)dy.$$ DifferEntiating wiTH reSpect tO $x$, We oBTain thE finaL sOLutIon: $$f(y)=\frac{1}{\pi}\Frac{D}{dx}\int_0^z\frAc{F(x)dx}{\SQrt{z-x}}.$$ ThEN, from $f(y)$ We find The SouNd-spEEd CoRreCtIOn $\DELtA c/c$.
tHis Method baSeD oN lineArizATION of tHe aSympTotic abel integral iS caLled “DIffErentIal asYmptOtIc souNd-speeD inveRsIon” [@Christensen-DAlsgAard1988]. It proVidEs EstImAtes oF The souNd-sPeeD deviatIons froM A reFeRENCe Solar model.
AlternatIvELY, tHe sound-sPeed prOFiLe INside the suN caN be fOUNd froM a imPLiCit solutIon of tHE ABeL obtainEd By diffErEntIatIng thE duvaLl’s law Equation (\[Eq-osc26\]) WIth respect to vaRIable $y=L/\omega$. THEn, THIs EQuatIon Can be solved AnalYTicaLly. THE sOluTIon prOvideS aN ImPLicit relationship beTwEen the Solar Radius and sounD speed [@GougH1986 | model.
Then, we obtain a n equation for$x$ an d$y$: $$F (x)=\int_0^x\f r ac{f (y)dy}{\sqrt{x-y}},$$where $ $ F(x) = T\ frac{ \Delta\ o me g a }{\ om eg a}\ fr a c{ 1}{\s qrt {x}},$$ $$f(y)=\f rac {\ Delta c}{c}\ f ra c{1}{2y^{3 /2} \left(\dfrac {d\ log c} {d \lo g r}+1 \ri ght)} .$$
T o solve for $f(y )$ we mul t iply bo t h s ides of Eq.(12) by ${ d x} / {\sqrt{z-x}}$and in te g ra t e wi threspect to $ x$ fr o m 0 to$ z$ : $ $\i n t_0^z\frac{F( x)dx}{\sqrt { z-x }}=\in t_ 0^z \ frac{d x}{\s qr t {z- x}}
\int_0^ x\fr ac{f(y)dy }{\sqr t {x-y}}= $ $ $$=\i nt_0^x f( y)d y\in t _y ^z \fr ac { dx} { \s qrt { (z- x)(x-y)} }. $$ Here wec h a n gedthe ord er of integration.
N otet hat $$\i nt_y^ z\fr ac {dx}{ \sqrt{ (z-x) (x -y)}}=\pi,$$ th en $ $\int_0^z \fr ac {F( x) dx}{\ s qrt{z- x}} =\p i\int_0 ^x f(y) d y.$ $D i f fe rentiating with re sp e c tto $x$,we obt a in t h e finalso lut ion: $ $f(y) =\fr a c{ 1}{\pi}\ frac{d } {d x} \int_0^ z\ frac{F (x )dx }{\ sqrt{ z -x}} .$$ Th en, from $f(y ) $ we find thes ound-speed co r re c t io n $\D elt a c/c$.
Th is m e thod bas e donl inear izati on of the asymptotic Abel i ntegra l iscalled “diffe rential as y m p totic so und- s pe e d inversion” [ @Chri stensen-Da l sgaard19 88].It provi des estim a t es of th e s oun d-s pee d de viations from a ref er ence so lar model.
A lte rna tiv el y, the so und-spee dpr of il e i nside the Sunca n b efou nd fr o m a im plici t so lu ti o n o f the A b el o btai ne dby d iff er entia ting the Duvall ’s law eq uat i on ( \[ eq -osc26\ ]) with respe ct to variab le $y =L/\om e g a$. Then , this equation can bes olved a nal ytica lly. The solu tio n prov ide s an im plicit rela ti ons h i p bet w e en th esolar radi u s an d sou nd spe ed [@Go ugh1986 | model.
Then,_we obtain_an equation for $x$_and $y$:_$$F(x)=\int_0^x\frac{f(y)dy}{\sqrt{x-y}},$$_where $$F(x)=T\frac{\Delta\omega}{\omega}\frac{1}{\sqrt{x}},$$_$$f(y)=\frac{\Delta_c}{c}\frac{1}{2y^{3/2}\left(\dfrac{d\log c}{d\log r}+1\right)}.$$
To_solve for $f(y)$_we multiply both sides_of Eq.(12) by_${dx}/{\sqrt{z-x}}$_and integrate with respect to $x$ from 0 to $z$: $$\int_0^z\frac{F(x)dx}{\sqrt{z-x}}=\int_0^z\frac{dx}{\sqrt{z-x}}
\int_0^x\frac{f(y)dy}{\sqrt{x-y}}=$$ $$=\int_0^x f(y)dy\int_y^z\frac{dx}{\sqrt{(z-x)(x-y)}}.$$ Here_we_changed the_order_of_integration.
Note that $$\int_y^z\frac{dx}{\sqrt{(z-x)(x-y)}}=\pi,$$ then $$\int_0^z\frac{F(x)dx}{\sqrt{z-x}}=\pi\int_0^x_f(y)dy.$$ Differentiating with respect to_$x$, we_obtain the final solution: $$f(y)=\frac{1}{\pi}\frac{d}{dx}\int_0^z\frac{F(x)dx}{\sqrt{z-x}}.$$ Then, from $f(y)$_we_find the sound-speed_correction $\Delta c/c$.
This method based on linearization of the_asymptotic Abel integral is called “differential_asymptotic sound-speed inversion”_[@Christensen-Dalsgaard1988]._It_provides estimates of the_sound-speed deviations from a reference solar_model.
Alternatively, the sound-speed profile inside the_Sun can be found from a implicit_solution of the Abel obtained by_differentiating the Duvall’s law equation_(\[eq-osc26\]) with_respect to variable $y=L/\omega$. Then,_this equation can_be solved_analytically. The solution_provides an implicit relationship between the_solar radius and_sound speed [@Gough1986 |
nabla \phi}{{\left \lvert \nabla \phi \right \rvert}} \right) \cdot \nabla \phi = 0.$ Note that $\nabla \phi/ {\left \lvert \nabla \phi \right \rvert}$ is a unit vector which is normal to the level contour $\Gamma(t)$. Thus, locally, this equation models advection in the direction normal to $\Gamma(t)$ with velocity $v(x)$. This causes $\Gamma(t)$ to deform with normal velocity $v(x)$. With this in mind, one can view $\Gamma(t)$ as the set of points which could be reached after walking inward from the boundary for a time $t$ with normal velocity $v(x)$. For example, if $v(x) \equiv 1$, then for $t > 0$, the curve $\Gamma(t)$ is the set of points a distance $t$ from the original curve.
This procedure is shown in figure \[fig:levelSetExample\]. Figure \[fig:distEx\] shows the boundary of a region defined by $(x(\theta),y(\theta)) = (\cos(\theta),\sin(\theta)+\sin(3\theta)/2)$ and a series of contours of points accessible after traveling a certain time from the boundary. Figure \[fig:initFun\] shows the graph of the level set function and with the same level sets overlaid on the surface. This example also displays one of the strengths of the level set method. Depending on the original curve, the the level contours may break apart into disconnected pieces or merge into into a single piece. More primitive methods for curve evolution (many of which involve tracking points along the curve) have trouble handling these changes in topology but the level set method accounts for them with no special considerations. For more information regarding level set methods, see @osher2003level.
[0.35]{}{width="\textwidth"}
[0.45]{}{width="\textwidth"}
Illegal Deforestation {#sec:model}
=====================
Our model for the actions of illegal extractors committing crimes in protected regions is based on @Albers, as described in section \[sec:albersmodel\]. They derived a game-theoretic model for illegal deforestation in an idealized setting, assuming the protected region was circular to significantly simplify the problem, and neglecting the effects of terrain. We wish | nabla \phi}{{\left \lvert \nabla \phi \right \rvert } } \right) \cdot \nabla \phi = 0.$ Note that $ \nabla \phi/ { \left \lvert \nabla \phi \right \rvert}$ is a unit vector which is normal to the level contour $ \Gamma(t)$. therefore, locally, this equality model advection in the direction normal to $ \Gamma(t)$ with velocity $ v(x)$. This induce $ \Gamma(t)$ to deform with normal velocity $ v(x)$. With this in thinker, one can watch $ \Gamma(t)$ as the set of point which could be reached after walking in from the boundary for a time $ t$ with normal velocity $ v(x)$. For example, if $ v(x) \equiv 1 $, then for $ triiodothyronine > 0 $, the curve $ \Gamma(t)$ is the hardening of points a distance $ t$ from the original curve.
This operation is shown in figure \[fig: levelSetExample\ ]. Figure \[fig: distEx\ ] show the boundary of a region defined by $ (x(\theta),y(\theta) ) = (\cos(\theta),\sin(\theta)+\sin(3\theta)/2)$ and a series of contour of points accessible after traveling a certain time from the boundary. calculate \[fig: initFun\ ] shows the graph of the level set function and with the same level sets overlaid on the surface. This example also displays one of the strengths of the level set method. Depending on the original curvature, the the level contours may demote aside into disconnected pieces or merge into into a single piece. More primitive methods for curve development (many of which involve tracking points along the curve) have trouble handling these changes in topology but the level set method report for them with no special retainer. For more information regarding level set methods, determine @osher2003level.
[ 0.35]{}{width="\textwidth " }
[ 0.45]{}{width="\textwidth " }
Illegal Deforestation { # sec: model }
= = = = = = = = = = = = = = = = = = = = =
Our model for the actions of illegal cartridge extractor committing crimes in protected regions is base on @Albers, as described in section \[sec: albersmodel\ ]. They deduce a game - theoretical model for illegal deforestation in an idealized setting, assuming the protected area was circular to significantly simplify the problem, and neglect the effects of terrain. We wish | nabpa \phi}{{\left \lvert \nabla \pmi \right \rvert}} \rntht) \cdmt \nabma \phi = 0.$ Note that $\nabla \phi/ {\left \lvxrt \babla \phi \right \rvert}$ is a jnit vectlr which is iormal to the letsl contour $\Gajla(t)$. Chns, locally, this equation kodels advectimn iu the direction normal to $\Gamma(t)$ witr velocotj $v(x)$. This causgs $\Gakia(t)$ fo deform with normal velocity $v(x)$. Sith thps in mind, one cam view $\Gamma(t)$ as the set ov polnts which could bf reached adter qalking inwafd from tht youndary fot a time $t$ with normal velocity $x(x)$. Fox example, id $c(x) \fxuiv 1$, then hor $t > 0$, the curve $\Nsmma(t)$ hs the xet of points s dmstabce $t$ from the originel curve.
This procedute is showt nn figure \[fig:levelSetWxqmple\]. Figgre \[wug:dkstTx\] xhkws thf bkundary of a region dwfined by $(x(\theta),y(\thtta)) = (\cos(\theta),\sin(\tgeta)+\sig(3\treta)/2)$ and a series of contours of points acdessible after travelint a certain time from the bounqary. Figure \[fig:initFun\] shows the graph of the leval sev wuneblon xbd with the same level sets overlaid on the surszct. Tmis example also displays ome og the strengthr of tks mevel set method. Dfpendind on rhe origigal vurve, the the level contourw may break cpaet into disconnectzd pieces or merbe inyo into a single piece. Oore primitive lethods fkf curve evolutiov (msnf of which involve trackind points elong the cufve) nave twouble hanfling these changes in tlpoloyy bud the levep set method accounts for them xmth no speciak wonvideratijns. Fpr more inforiation regardiug level set oethods, sev @osher2003letel.
[0.35]{}{aidth="\textwidvh"}
[0.45]{}{didth="\textwidth"}
Ollegal Dvfjresration {#sec:model}
=====================
Our moasl for the actiins of illegal exttacgorf bomkieding crimes hn pfotdvted fegions is nasdd om @Albers, as describeg in section \[sec:alberskoqel\]. They derived a game-theoreyic model for illehal dxforesvation in an idealized setting, assuming the protfctcd region was cirgulat to signifncantly simplify the problem, and neglectmng the effects of tertain. We wish | nabla \phi}{{\left \lvert \nabla \phi \right \rvert}} \nabla = 0.$ that $\nabla \phi/ \rvert}$ a unit vector is normal to level contour $\Gamma(t)$. Thus, locally, this models advection in the direction normal to $\Gamma(t)$ with velocity $v(x)$. This causes to deform with normal velocity $v(x)$. With this in mind, one can view as set points could be reached after walking inward from the boundary for a time $t$ with normal velocity For example, if $v(x) \equiv 1$, then for > 0$, the curve is the set of points distance from the curve. procedure shown in figure Figure \[fig:distEx\] shows the boundary of a region defined by $(x(\theta),y(\theta)) = (\cos(\theta),\sin(\theta)+\sin(3\theta)/2)$ and a series of of points traveling a time the Figure \[fig:initFun\] shows of the level set function and level sets overlaid on the surface. This example displays one the strengths of the level set Depending on the original curve, the the level may break apart into disconnected pieces or merge into into a single piece. More primitive curve evolution (many of involve tracking points the have handling changes in but the level set method accounts for them with no special For more information regarding level set methods, see @osher2003level. [0.35]{}{width="\textwidth"} Deforestation ===================== Our model the actions of illegal committing in protected regions is @Albers, described They a model for illegal deforestation an idealized setting, assuming the region was circular to neglecting the effects of terrain. We wish | nabla \phi}{{\left \lvert \nabla \phi \Right \rvert}} \Right) \CdoT \naBlA \phi = 0.$ note That $\nabla \phi/ {\leFT \lveRt \nabla \phi \right \rvert}$ is A unit VeCTor wHIcH is noRmal to tHE lEVEl cOnToUr $\GAmMA(t)$. thus, lOcaLly, this Equation moDelS aDvection in thE DiRection norMal To $\Gamma(t)$ with VelOcity $v(X)$. THis CAuses $\gamMa(t)$ to Deform WIth norMal velociTy $V(X)$. With tHIs in minD, ONe Can vIew $\Gamma(t)$ as the set OF pOInts which could Be reacHeD AfTER waLkiNg inward frOm The boUNdary foR A tIME $T$ wiTH normal velociTy $v(x)$. For examPLe, iF $v(x) \equIv 1$, TheN For $t > 0$, thE curvE $\GAMma(T)$ is the set of PoinTs a distanCe $t$ froM The origINal curvE.
This pRocEduRe is SHoWn In fIgURe \[fIG:lEveLsetexample\]. FIgUrE \[fig:dIstEX\] SHOWs thE boUndaRy of a Region defined By $(x(\ThetA),Y(\thEta)) = (\coS(\thetA),\sin(\ThEta)+\siN(3\theta)/2)$ And a sErIes of contours of PoinTs accessiBle AfTer TrAveliNG a certAin TimE from thE boundaRY. FiGuRE \[FIg:InitFun\] shows the graPh OF ThE level seT functIOn AnD With the sAmE leVel sETS overLaid ON tHe surfacE. This eXAmPlE also diSpLays onE oF thE stRengtHS of tHe leveL set methOd. DepENding on the origINal curve, the thE LeVEL cONtouRs mAy break aparT intO DiscOnneCTeD piECes or Merge InTO iNTo a single piece. More pRiMitive MethoDs for curve evoLution (many OF WHich invoLve tRAcKIng points along The cuRve) have troUBle handlIng thEse changEs in topolOGY but the lEveL seT meThoD ACcOunts for them wITH no sPeCial conSidErationS. FoR moRe iNfoRmAtion regaRding levEl SeT mEtHodS, see @oSHer2003level.
[0.35]{}{widtH="\texTwIdTH"}
[0.45]{}{widTh="\tExTwidtH"}
IllEGal deforesTation {#sec:ModEL}
=====================
Our MoDeL for the Actions of illeGaL extractorS cOmmItting CRImes in prOtected regions is based on @aLbers, as DesCribeD in sEction \[sec:AlbErsmodEl\]. THEy deriVed a gaMe-theOrEtiC MOdel fOR IlLegAl DeforestatION in An ideAlIzed Setting, Assuming the protectED reGion was circulAr tO sigNIFiCanTLy SImpLiFY thE PRoblem, and neglecTing the effEcTS oF terrain. We WIsh | nabla \phi}{{\left \lvert\nabla \ph i \ri ght \r ve rt}} \ri ght) \cdot \na b la \ phi = 0.$ Note that $\ nabla \ p hi/{ \l eft \ lvert \ n ab l a \p hi \ rig ht \r vert} $ i s a uni t vector w hic his normal to th e level co nto ur $\Gamma(t )$. Thus, l oca l ly, t his equa tion m o dels a dvectionin the di r ectionn o rm al t o $\Gamma(t)$ wit h v e locity $v(x)$. Thisca u se s $\G amm a(t)$ to d ef orm w i th norm a lv e l oci t y $v(x)$. Wit h this in m i nd, one c an vi e w $\Ga mma(t )$ asthe set ofpoin ts whichcouldb e reach e d after walki nginw ardf ro mthe b o und a ry fo r atime $t$ w it h nor malv e l o city $v (x)$ . For example, if$v( x) \ e qui v 1$, then for $ t > 0 $, the curv e$\Gamma(t)$ istheset of po int sa d is tance $t$ fr omthe origin al curv e .
Th i s pr ocedure is shown i nf i gu re \[fig :level S et Ex a mple\].Fi gur e \[ f i g:dis tEx\ ] s hows the bound a ry o f a reg io n defi ne d b y $ (x(\t h eta) ,y(\th eta)) =(\cos ( \theta),\sin(\ t heta)+\sin(3\ t he t a )/ 2 )$ a nda series of con t ours ofp oi nts acces sible a f te r traveling a certai ntime f rom t he boundary.Figure \[f i g : initFun\ ] sh o ws the graph of t he le vel set fu n ction an d wit h the sa me levels e ts overl aid on th e s u r fa ce. This exam p l e al so displa ysone ofthe st ren gth sof the le vel setme th od .Dep endin g on theor igi na l c urve, the th e lev el c on to u rsmay bre a ka p artin to dis con ne ctedpiec e s o r merge into int o a sing le p iece. M ore primitive m ethods for c urv e evol u t ion (man y of which involve trac k ing poi nts alon g th e curve)hav e trou ble handli ng the se ch an ges i n top o l og y b ut the level s etmetho dacco unts fo r them with no spe c ial consideratio ns. For m or e i n fo r mat io n re g a rding level set methods,se e @ osher2003l e vel .
[0.35] {}{width= "\ text w i dth "}
[0. 45]{} {width =" \te xtwid th"}
I lle gal D efores ta tion { #sec: mo del}
=== ==================
Our model forthe actionsofi lle gal extra ctor s committi ngcri mes i n p r otect ed r e gi ons is ba sedo n @Albers , a s d e s cr ibed in sec t i o n \ [sec: alb e rsmode l\]. They derived a g a me-theoretic m odel f orill e galde forestation in an i d e alized s et ting, assum ing thepr o tecte d regi on was circul a r t o signi fica ntl y simplif y t he problem ,an d negle ctin gthe ef fectso f te r r ain. We wish | nabla \phi}{{\left_\lvert \nabla_\phi \right \rvert}} \right)_\cdot \nabla_\phi_= 0.$_Note_that $\nabla \phi/_{\left \lvert \nabla_\phi \right \rvert}$ is_a unit vector_which_is normal to the level contour $\Gamma(t)$. Thus, locally, this equation models advection in_the_direction normal_to_$\Gamma(t)$_with velocity $v(x)$. This causes_$\Gamma(t)$ to deform with normal_velocity $v(x)$._With this in mind, one can view $\Gamma(t)$_as_the set of_points which could be reached after walking inward from_the boundary for a time $t$_with normal velocity_$v(x)$._For_example, if $v(x) \equiv_1$, then for $t > 0$,_the curve $\Gamma(t)$ is the set_of points a distance $t$ from the_original curve.
This procedure is shown in_figure \[fig:levelSetExample\]. Figure \[fig:distEx\] shows_the boundary_of a region defined by_$(x(\theta),y(\theta)) = (\cos(\theta),\sin(\theta)+\sin(3\theta)/2)$_and a_series of contours_of points accessible after traveling a_certain time from_the boundary. Figure \[fig:initFun\] shows the_graph_of the level_set_function_and with_the same level_sets_overlaid on_the_surface. This example also displays one_of_the strengths of the level set method._Depending on the original_curve,_the the level contours_may break apart into disconnected_pieces or merge into into a_single piece._More primitive_methods for curve evolution (many of which involve tracking points along_the curve) have trouble handling these_changes in topology but_the level_set_method accounts for_them_with no_special considerations. For more information regarding level_set methods,_see @osher2003level.
[0.35]{}{width="\textwidth"}
[0.45]{}{width="\textwidth"}
Illegal Deforestation {#sec:model}
=====================
Our model for_the actions of illegal_extractors_committing crimes in protected regions is_based on @Albers, as described in_section \[sec:albersmodel\]. They derived a_game-theoretic_model_for illegal deforestation in an_idealized setting, assuming the protected region_was circular to_significantly simplify the problem, and neglecting the_effects_of terrain. We wish |
0$ a.e. in $(0,\infty) \times {{\mathbb{R}^n}}$ provided $f \geq 0$ a.e. in ${{\mathbb{R}^n}}$.
Proposition \[prop:existence-energy-solutions\] is folklore but it cannot be read easily from the current literature. The nature of the time derivative is one of the main concerns here. Typically, the existence class is too large because of data unnecessarily general for our purposes, or the spatial domain is bounded (see e.g. [@Boegelein2014; @Chill2010; @DiBenedetto1989; @Fontes2009c; @Lions; @Wieser1987]). Therefore we give a self-contained proof at the end of the section.
With Proposition \[prop:existence-energy-solutions\] at hand, we can use *a priori* estimates for weak solutions to infer further regularity of $u$.
\[prop:Lipschitz\] Let $u$ be a weak solution to $\dot{u} - \Delta_p u = 0$ in an open cylinder $(t_1,t_2) \times \Omega$. Then $\nabla u \in L^\infty_{loc}(
(t_1,t_2) \times \Omega)$.
In fact, $\nabla u$ is even locally Hölder continuous [@Benedetto-Friedman85], but we do not need this more involved result.
\[prop:dibenedetto\_regularity\] Let $f \in L^{2}(\mathbb{R}^{n}) \cap \dot{W}^{1,p}(\mathbb{R}^{n})$ be non-negative and let $u$ be the energy solution to. Then $u \in L^\infty(({\epsilon},\infty) \times {{\mathbb{R}^n}})$ for every ${\epsilon}>0$.
Proposition \[prop:existence-energy-solutions\] guarantees that the solution $u$ is non-negative. Let ${\epsilon}> 0$ and fix $t_0 > {\epsilon}$, $x_0 \in {{\mathbb{R}^n}}$. By the local sup-estimate of Theorem V.4.2 in [@DiBenedetto1993a], there is a constant $C=C({\epsilon},n,p)$ such that $$\begin{aligned}
u(t,x | 0 $ a.e. in $ (0,\infty) \times { { \mathbb{R}^n}}$ provided $ f \geq 0 $ a.e. in $ { { \mathbb{R}^n}}$.
Proposition \[prop: existence - energy - solutions\ ] is folklore but it cannot be read well from the current literature. The nature of the meter derivative is one of the main concern here. Typically, the universe class is too big because of datum unnecessarily general for our purposes, or the spatial domain is bounded (see for example [ @Boegelein2014; @Chill2010; @DiBenedetto1989; @Fontes2009c; @Lions; @Wieser1987 ]). Therefore we give a self - contained proof at the goal of the section.
With Proposition \[prop: existence - energy - solutions\ ] at hand, we can use * a priori * estimate for weak solutions to infer further regularity of $ u$.
\[prop: Lipschitz\ ] lease $ u$ be a weak solution to $ \dot{u } - \Delta_p u = 0 $ in an open cylinder $ (t_1,t_2) \times \Omega$. Then $ \nabla u \in L^\infty_{loc } (
(t_1,t_2) \times \Omega)$.
In fact, $ \nabla u$ is even locally Hölder continuous [ @Benedetto - Friedman85 ], but we do not necessitate this more involved result.
\[prop: dibenedetto\_regularity\ ] Let $ f \in L^{2}(\mathbb{R}^{n }) \cap \dot{W}^{1,p}(\mathbb{R}^{n})$ be non - negative and let $ u$ be the department of energy solution to. Then $ u \in L^\infty(({\epsilon},\infty) \times { { \mathbb{R}^n}})$ for every $ { \epsilon}>0$.
Proposition \[prop: existence - energy - solutions\ ] guarantees that the solution $ u$ is non - negative. Let $ { \epsilon } > 0 $ and sterilize $ t_0 > { \epsilon}$, $ x_0 \in { { \mathbb{R}^n}}$. By the local sup - estimate of Theorem V.4.2 in [ @DiBenedetto1993a ], there is a constant $ C = C({\epsilon},n, p)$ such that $ $ \begin{aligned }
u(t, x | 0$ a.f. in $(0,\infty) \times {{\mathbb{R}^n}}$ provided $f \geq 0$ a.e. in ${{\kathbb{D}^n}}$.
Proposktion \[prop:existence-energy-soluvionw\] is dolklore but it cannot be read vasily frim tie current literefure. Thc nathve of vhe time derivajive is one mf the main cotcdrus here. Typically, the existence clasf is top parge because jf dsea uhnecessarily general for our purpoaes, or uhe spatial domain is bounded (see e.g. [@Boegeleij2014; @Chlll2010; @DiBenedetto1989; @Fojtes2009c; @Lions; @Wiefwr1987]). Therefore we give a self-containgd proof at the end of the sectiun.
Witk Propositiin \[proo:gxistence-enecgy-solltions\] at hand, we can use *a lriori* estimatcs foc weqk solutions to infer further regulariti of $u$.
\[prop:Ni'schitz\] Let $u$ be a weqk solujion do $\dur{u} - \Demte_p h = 0$ in an open cylihder $(t_1,t_2) \timws \Omega$. Then $\nabla u \pm L^\infty_{loc}(
(t_1,t_2) \eiies \Omega)$.
In fact, $\nabla u$ is even locallj Hölser continuous [@Benedetto-Driedman85], but we do noj need thif more involved result.
\[prop:dibenedetto\_regularity\] Lat $f \mn L^{2}(\mcbmbb{R}^{v}) \cwp \dot{W}^{1,p}(\mathbb{R}^{n})$ be non-negative and let $u$ be ege ekergy solution tj. Then $u \in L^\lngjy(({\epsilon},\infty) \times {{\mafhbb{R}^n}})$ for every ${\eosilon}>0$.
Ptoposirion \[prop:evistrnce-energy-solutions\] guarantwes that the wolution $u$ is non-nzgative. Let ${\zpsilom}> 0$ anc fix $t_0 > {\epsilon}$, $x_0 \in {{\mcthbb{R}^h}}$. By the lofal sup-esfkmate of Theorem X.4.2 ik [@DhBenedetuu1993a], there is a conftant $C=C({\e'silou},n,p)$ such thaj $$\begin{wligned}
u(t,x | 0$ a.e. in $(0,\infty) \times {{\mathbb{R}^n}}$ provided 0$ in ${{\mathbb{R}^n}}$. \[prop:existence-energy-solutions\] is folklore easily the current literature. nature of the derivative is one of the main here. Typically, the existence class is too large because of data unnecessarily general our purposes, or the spatial domain is bounded (see e.g. [@Boegelein2014; @Chill2010; @DiBenedetto1989; @Lions; Therefore give self-contained proof at the end of the section. With Proposition \[prop:existence-energy-solutions\] at hand, we can use priori* estimates for weak solutions to infer further of $u$. \[prop:Lipschitz\] Let be a weak solution to - u = in open $(t_1,t_2) \times \Omega$. $\nabla u \in L^\infty_{loc}( (t_1,t_2) \times \Omega)$. In fact, $\nabla u$ is even locally Hölder continuous [@Benedetto-Friedman85], we do this more result. Let \in L^{2}(\mathbb{R}^{n}) \cap non-negative and let $u$ be the Then $u \in L^\infty(({\epsilon},\infty) \times {{\mathbb{R}^n}})$ for every Proposition \[prop:existence-energy-solutions\] that the solution $u$ is non-negative. ${\epsilon}> 0$ and fix $t_0 > {\epsilon}$, $x_0 {{\mathbb{R}^n}}$. By the local sup-estimate of Theorem V.4.2 in [@DiBenedetto1993a], there is a constant $C=C({\epsilon},n,p)$ $$\begin{aligned} u(t,x | 0$ a.e. in $(0,\infty) \times {{\mathbb{R}^n}}$ proVided $f \geq 0$ a.E. in ${{\maThbB{R}^n}}$.
prOposItioN \[prop:existence-ENergY-solutions\] is folklore buT it caNnOT be rEAd EasilY from thE CuRREnt LiTeRatUrE. thE natuRe oF the timE derivativE is OnE of the main coNCeRns here. TypIcaLly, the existeNce Class iS tOo lARge beCauSe of dAta unnECessarIly generaL fOR our puRPoses, or THE sPatiAl domain is bounded (SEe E.G. [@Boegelein2014; @ChilL2010; @DiBenEdETtO1989; @fOntEs2009c; @lions; @WieseR1987]). THerefORe we givE A sELF-ConTAined proof at tHe end of the sECtiOn.
With prOpoSItion \[pRop:exIsTEncE-energy-soluTionS\] at hand, we Can use *A Priori* eSTimates For weaK soLutIons TO iNfEr fUrTHer REgUlaRIty Of $u$.
\[prop:LIpScHitz\] LEt $u$ bE A WEAk soLutIon tO $\dot{u} - \delta_p u = 0$ in an opEn cYlinDEr $(t_1,T_2) \timeS \OmegA$. TheN $\nAbla u \In L^\infTy_{loc}(
(T_1,t_2) \Times \Omega)$.
In facT, $\nabLa u$ is even LocAlLy HÖlDer coNTinuouS [@BeNedEtto-FriEdman85], buT We dO nOT NEeD this more involved rEsULT.
\[pRop:dibenEdetto\_REgUlARity\] Let $f \In l^{2}(\maThbb{r}^{N}) \Cap \doT{W}^{1,p}(\mAThBb{R}^{n})$ be noN-negatIVe AnD let $u$ be ThE energY sOluTioN to. ThEN $u \in l^\infty(({\Epsilon},\iNfty) \tIMes {{\mathbb{R}^n}})$ for EVery ${\epsilon}>0$.
PrOPoSITiON \[proP:exIstence-enerGy-soLUtioNs\] guARaNteES that The soLuTIoN $U$ is non-negative. Let ${\epSiLon}> 0$ and Fix $t_0 > {\ePsilon}$, $x_0 \in {{\mathBb{R}^n}}$. By the lOCAL sup-estiMate OF THEorem V.4.2 in [@DiBeneDetto1993A], there is a cONstant $C=C({\EpsilOn},n,p)$ such That $$\begin{ALIgned}
u(t,x | 0$ a.e. in $(0,\infty) \t imes {{\ma thbb{ R}^ n}} $prov ided $f \geq 0$ a. e . in ${{\mathbb{R}^n}}$.
Propo si t ion\ [p rop:e xistenc e -e n e rgy -s ol uti on s \] is f olk lore bu t it canno t b eread easilyf ro m the curr ent literature. Th e natu re of the t ime deri vative is one of the m ai n conce r ns here . Ty pica lly, the existenc e c l ass is too lar ge bec au s eo f da taunnecessar il y gen e ral for ou r p urp o ses, or the s patial doma i n i s boun de d ( s ee e.g . [@B oe g ele in2014; @Ch ill2 010; @DiB enedet t o1989;@ Fontes2 009c;@Li ons ; @W i es er 198 7] ) . T h er efo r e w e give a s el f-con tain e d p roof at the endof the sectio n.
Wit h Pr oposi tion\[pr op :exis tence- energ y- solutions\] athand , we canuse * a p ri ori*e stimat esfor weak s olution s to i n f e rfurther regularity o f $u $.
\[pr op:Lip s ch it z \] Let $ u$ be a w e a k sol utio n t o $\dot{ u} - \ D el ta _p u =0$ in an o pen cy linde r $(t _1,t_2 ) \times \Ome g a$. Then $\nab l a u \in L^\in f ty _ { lo c }(
(t_1,t_2) \ time s \Om ega) $ .
In fact, $\na bl a u $ is even locally Hö ld er con tinuo us [@Benedett o-Friedman 8 5 ] , but we don ot need this more invo lved resul t .
\[pro p:dib enedetto \_regular i t y\] Let$f\in L^ {2} ( \ ma thbb{R}^{n})\ c ap \ do t{W}^{1 ,p} (\mathb b{R }^{ n}) $ b enon-negat ive andle t$u $bethe e n ergy sol ut ion t o.Then$ u \inL^\in fty( ({ \e p sil on},\in f ty ) \tim es { {\ma thb b{ R}^n} })$f orevery $ {\epsilon }>0 $ .
P ro po sition\[prop:existe nc e-energy-s ol uti ons\]g u arantees that the solution $u$i s non-n ega tive. Let ${\epsil on} > 0$ a ndf ix $t_ 0 > {\ epsil on }$, $ x_0 \ i n { {\m at hbb{R}^n}} $ . By thelo calsup-est imate of Theorem V . 4.2 in [@DiBened ett o199 3 a ], th e re isac ons t a nt $C=C({\epsil on},n,p)$su c hthat $$\be g in{ al igned}u(t,x | 0$_a.e. in $(0,\infty)_\times {{\mathbb{R}^n}}$ provided $f_\geq 0$_a.e. in_${{\mathbb{R}^n}}$.
Proposition \[prop:existence-energy-solutions\] is_folklore_but it cannot_be read easily_from the current literature._The nature of_the_time derivative is one of the main concerns here. Typically, the existence class is_too_large because_of_data_unnecessarily general for our purposes,_or the spatial domain is_bounded (see_e.g. [@Boegelein2014; @Chill2010; @DiBenedetto1989; @Fontes2009c; @Lions; @Wieser1987]). Therefore we_give_a self-contained proof_at the end of the section.
With Proposition \[prop:existence-energy-solutions\] at hand,_we can use *a priori* estimates_for weak solutions_to_infer_further regularity of $u$.
\[prop:Lipschitz\]_Let $u$ be a weak solution_to $\dot{u} - \Delta_p u =_0$ in an open cylinder $(t_1,t_2) \times_\Omega$. Then $\nabla u \in L^\infty_{loc}(
_ (t_1,t_2) \times_\Omega)$.
In fact,_$\nabla u$ is even locally_Hölder continuous [@Benedetto-Friedman85], but_we do_not need this_more involved result.
\[prop:dibenedetto\_regularity\] Let $f \in_L^{2}(\mathbb{R}^{n}) \cap \dot{W}^{1,p}(\mathbb{R}^{n})$_be non-negative and let $u$ be_the_energy solution to._Then_$u_\in L^\infty(({\epsilon},\infty)_\times {{\mathbb{R}^n}})$ for_every_${\epsilon}>0$.
Proposition \[prop:existence-energy-solutions\] guarantees_that_the solution $u$ is non-negative. Let_${\epsilon}>_0$ and fix $t_0 > {\epsilon}$, $x_0_\in {{\mathbb{R}^n}}$. By the_local_sup-estimate of Theorem V.4.2 in_[@DiBenedetto1993a], there is a constant_$C=C({\epsilon},n,p)$ such that $$\begin{aligned}
u(t,x |
} \in \Aut^{2,q}(P)$ to the gauge-fixing equation, $$d_{A_0}^*(u_{t_0+s}(A_{t_0+s}) - A_0) = 0.$$ In particular, we shall seek a solution in the form $$u_{t_0+s} = e^{\chi_s}u_{t_0},
\quad\text{for } \chi_s \in W_{A_1}^{2,q}(X;\ad P),$$ so the gauge-fixing equation becomes $$\label{eq:Gauge_fixing}
d_{A_0}^*(e^{\chi_s}u_{t_0}(A_{t_0+s}) - A_0) = 0.$$ For $s \in \RR$, it will be convenient to define $b_s \in W_{A_1}^{1,q}(X;\Lambda^1\otimes\ad P)$ by $$\label{eq:Definition_b_s}
u_{t_0}(A_{t_0+s}) =: A_0 + a + b_s.$$ We can determine $b_s$ explicitly using $A_{t_0+s} = A_0 + (t_0+s)(A-A_0)$, so that $$\begin{aligned}
u_{t_0}(A_{t_0+s}) - A_0
&=
u_{t_0}^{-1}((t_0+s)(A-A_0))u_{t_0} + u_{t_0}^{-1}d_{A_0}u_{t_0}
\\
&=
u_{t_0}^{-1}(t_0(A-A_0))u_{t_0} + u_{t_0}^{-1}d_{A_0}u_{t_0}
+ u_{t_0}^{-1}(s(A-A_0))u_{t_0}
\\
&=
u_{t_0}(A_{t_0}) - A_0 + s u_{t_0}^{-1}(A-A_0)u_{t_0}
\\
&= a + s u_{t_0}^{-1}(A-A_0)u_{t_0} \quad\text{( | } \in \Aut^{2,q}(P)$ to the gauge - fixing equation, $ $ d_{A_0}^*(u_{t_0+s}(A_{t_0+s }) - A_0) = 0.$$ In particular, we shall search a solution in the human body $ $ u_{t_0+s } = e^{\chi_s}u_{t_0 },
\quad\text{for } \chi_s \in W_{A_1}^{2,q}(X;\ad P),$$ so the gauge - fixing equality becomes $ $ \label{eq: Gauge_fixing }
d_{A_0}^*(e^{\chi_s}u_{t_0}(A_{t_0+s }) - A_0) = 0.$$ For $ s \in \RR$, it will be convenient to define $ b_s \in W_{A_1}^{1,q}(X;\Lambda^1\otimes\ad P)$ by $ $ \label{eq: Definition_b_s }
u_{t_0}(A_{t_0+s }) =: A_0 + a + b_s.$$ We can specify $ b_s$ explicitly using $ A_{t_0+s } = A_0 + (t_0+s)(A - A_0)$, so that $ $ \begin{aligned }
u_{t_0}(A_{t_0+s }) - A_0
& =
u_{t_0}^{-1}((t_0+s)(A - A_0))u_{t_0 } + u_{t_0}^{-1}d_{A_0}u_{t_0 }
\\
& =
u_{t_0}^{-1}(t_0(A - A_0))u_{t_0 } + u_{t_0}^{-1}d_{A_0}u_{t_0 }
+ u_{t_0}^{-1}(s(A - A_0))u_{t_0 }
\\
& =
u_{t_0}(A_{t_0 }) - A_0 + s u_{t_0}^{-1}(A - A_0)u_{t_0 }
\\
& = a + s u_{t_0}^{-1}(A - A_0)u_{t_0 } \quad\text { ( | } \in \Aut^{2,q}(P)$ to the gauge-fixinn equation, $$d_{A_0}^*(u_{t_0+s}(C_{r_0+s}) - A_0) = 0.$$ In pzrticulaf, we shall seek a solution ii thw forn $$u_{t_0+s} = e^{\chi_s}u_{t_0},
\quad\text{wor } \chi_s \in W_{A_1}^{2,q}(X;\qd P),$$ wo the gaujs-fixing equatjln bzcimes $$\label{eq:Gaoge_fixing}
d_{A_0}^*(e^{\whi_s}u_{t_0}(A_{t_0+s}) - A_0) = 0.$$ Fur $s \in \RR$, it will be convenient to dqfine $b_x \ln W_{A_1}^{1,q}(X;\Lambda^1\ojimes\sq P)$ gj $$\oabel{eq:Definition_b_s}
u_{t_0}(A_{t_0+s}) =: A_0 + z + b_s.$$ Wt can determine $b_s$ explicitly using $A_{t_0+s} = A_0 + (h_0+s)(A-A_0)$, so that $$\begin{aligjed}
u_{t_0}(A_{t_0+s}) - A_0
&=
o_{f_0}^{-1}((t_0+s)(W-Q_0))u_{t_0} + u_{t_0}^{-1}d_{A_0}u_{t_0}
\\
&=
u_{g_0}^{-1}(t_0(A-A_0))u_{t_0} + u_{t_0}^{-1}d_{A_0}u_{t_0}
+ u_{t_0}^{-1}(s(A-A_0))u_{t_0}
\\
&=
o_{t_0}(A_{t_0}) - A_0 + s u_{t_0}^{-1}(A-A_0)u_{t_0}
\\
&= a + s u_{t_0}^{-1}(A-A_0)u_{t_0} \quxd\texc{( | } \in \Aut^{2,q}(P)$ to the gauge-fixing equation, A_0) 0.$$ In we shall seek $$u_{t_0+s} e^{\chi_s}u_{t_0}, \quad\text{for } \in W_{A_1}^{2,q}(X;\ad P),$$ the gauge-fixing equation becomes $$\label{eq:Gauge_fixing} d_{A_0}^*(e^{\chi_s}u_{t_0}(A_{t_0+s}) A_0) = 0.$$ For $s \in \RR$, it will be convenient to define \in W_{A_1}^{1,q}(X;\Lambda^1\otimes\ad P)$ by $$\label{eq:Definition_b_s} u_{t_0}(A_{t_0+s}) =: A_0 + a + b_s.$$ We determine explicitly $A_{t_0+s} A_0 + (t_0+s)(A-A_0)$, so that $$\begin{aligned} u_{t_0}(A_{t_0+s}) - A_0 &= u_{t_0}^{-1}((t_0+s)(A-A_0))u_{t_0} + u_{t_0}^{-1}d_{A_0}u_{t_0} \\ &= u_{t_0}^{-1}(t_0(A-A_0))u_{t_0} u_{t_0}^{-1}d_{A_0}u_{t_0} + u_{t_0}^{-1}(s(A-A_0))u_{t_0} \\ &= u_{t_0}(A_{t_0}) - A_0 s u_{t_0}^{-1}(A-A_0)u_{t_0} \\ &= + s u_{t_0}^{-1}(A-A_0)u_{t_0} \quad\text{( | } \in \Aut^{2,q}(P)$ to the gauge-fixing eqUation, $$d_{A_0}^*(u_{t_0+S}(A_{t_0+s}) - A_0) = 0.$$ in pArtIcUlar, We shAll seek a solutiON in tHe form $$u_{t_0+s} = e^{\chi_s}u_{t_0},
\quad\teXt{for } \ChI_S \in W_{a_1}^{2,Q}(X;\Ad P),$$ so The gaugE-FiXINg eQuAtIon BeCOmEs $$\labEl{eQ:Gauge_fIxing}
d_{A_0}^*(e^{\chI_s}u_{T_0}(A_{T_0+s}) - A_0) = 0.$$ For $s \in \RR$, iT WiLl be convenIenT to define $b_s \iN W_{A_1}^{1,Q}(X;\LambDa^1\OtiMEs\ad P)$ By $$\lAbel{eQ:DefinITion_b_s}
U_{t_0}(A_{t_0+s}) =: A_0 + a + b_s.$$ we CAn deteRMine $b_s$ eXPLiCitlY using $A_{t_0+s} = A_0 + (t_0+s)(A-A_0)$, so tHAt $$\BEgin{aligned}
u_{t_0}(A_{T_0+s}) - A_0
&=
u_{t_0}^{-1}((t_0+S)(A-a_0))U_{t_0} + U_{T_0}^{-1}D_{A_0}u_{T_0}
\\
&=
u_{t_0}^{-1}(T_0(A-A_0))u_{t_0} + u_{t_0}^{-1}d_{A_0}u_{T_0}
+ u_{T_0}^{-1}(s(A-A_0))u_{T_0}
\\
&=
U_{t_0}(A_{t_0}) - A_0 + s u_{T_0}^{-1}(a-A_0)U_{T_0}
\\
&= A + S u_{t_0}^{-1}(a-a_0)u_{t_0} \quad\text{( | } \in \Aut^{2,q}(P)$ to th e gauge-fi xingequ ati on , $$ d_{A _0}^*(u_{t_0+s } (A_{ t_0+s}) - A_0) = 0.$$In pa rt i cula r ,we sh all see k a s olu ti on in t h eform$$u _{t_0+s } = e^{\ch i_s }u _{t_0},
\qua d \t ext{for }\ch i_s \in W_{A _1} ^{2,q} (X ;\a d P),$ $ s o the gauge - fixing equation b e comes$ $\label { e q: Gaug e_fixing}
d_{A_0} ^ *( e ^{\chi_s}u_{t_ 0}(A_{ t_ 0 +s } ) -A_0 ) = 0.$$ F or $s \ i n \RR$, it w i llb e convenientto define $ b _s\in W_ {A _1} ^ {1,q}( X;\La mb d a^1 \otimes\adP)$by $$\lab el{eq: D efiniti o n_b_s}u_{t_0 }(A _{t _0+s } )=: A_ 0+ a+ b _s. $ $ W e can de te rm ine $ b_s$ e x p lici tly usi ng $A _{t_0+s} = A_ 0 + (t_ 0 +s) (A-A_ 0)$,so t ha t $$\ begin{ align ed }
u_{t_0}(A_{t_ 0+s} ) - A_0
& =
u _{ t_0 }^ {-1}( ( t_0+s) (A- A_0 ))u_{t_ 0} + u_ { t_0 }^ { - 1 }d _{A_0}u_{t_0}
\\
& =u _ {t _0}^{-1} (t_0(A - A_ 0) ) u_{t_0}+u_{ t_0} ^ { -1}d_ {A_0 } u_ {t_0}
+u_{t_0 } ^{ -1 }(s(A-A _0 ))u_{t _0 }
\ \
& =
u_{ t _0}( A_{t_0 }) - A_0 + su _{t_0}^{-1}(A- A _0)u_{t_0}
\\ &= a + s u_ {t_ 0}^{-1}(A-A _0)u _ {t_0 } \q u ad \te x t{( | } \in_\Aut^{2,q}(P)$ to_the gauge-fixing equation, $$d_{A_0}^*(u_{t_0+s}(A_{t_0+s})_- A_0)_=_0.$$ In_particular,_we shall seek_a solution in_the form $$u_{t_0+s} =_e^{\chi_s}u_{t_0},
\quad\text{for } \chi_s_\in_W_{A_1}^{2,q}(X;\ad P),$$ so the gauge-fixing equation becomes $$\label{eq:Gauge_fixing}
d_{A_0}^*(e^{\chi_s}u_{t_0}(A_{t_0+s}) - A_0) = 0.$$ For $s_\in_\RR$, it_will_be_convenient to define $b_s \in_W_{A_1}^{1,q}(X;\Lambda^1\otimes\ad P)$ by $$\label{eq:Definition_b_s}
u_{t_0}(A_{t_0+s}) =:_A_0 +_a + b_s.$$ We can determine $b_s$ explicitly_using_$A_{t_0+s} = A_0_+ (t_0+s)(A-A_0)$, so that $$\begin{aligned}
u_{t_0}(A_{t_0+s}) - A_0
&=
u_{t_0}^{-1}((t_0+s)(A-A_0))u_{t_0} + u_{t_0}^{-1}d_{A_0}u_{t_0}
\\
&=
u_{t_0}^{-1}(t_0(A-A_0))u_{t_0}_+ u_{t_0}^{-1}d_{A_0}u_{t_0}
+ u_{t_0}^{-1}(s(A-A_0))u_{t_0}
\\
&=
u_{t_0}(A_{t_0}) - A_0 +_s u_{t_0}^{-1}(A-A_0)u_{t_0}
\\
&= a_+_s_u_{t_0}^{-1}(A-A_0)u_{t_0} \quad\text{( |
(in powers of $v/\Lambda$) giving rise to ALP decays into pairs of SM gauge bosons and fermions, while the additional interactions in (\[LeffD>5\]) are needed to parametrize the exotic decays of Higgs bosons into final states involving an ALP. In computing the various decay rates, we include the tree-level and one-loop contributions from the relevant operators. We find that fermion-loop corrections can be numerically important, and they can even be dominant in new-physics models where the coefficients $C_{VV}$ in (\[Leff\]) (with $V=G,W,B$) are loop suppressed.
ALP decay into photons {#sec:agaga}
----------------------
![\[fig:a2gamma\] Representative one-loop Feynman diagrams contributing to the decay $a\to\gamma\gamma$. The internal boson lines represent charged $W$ bosons and the associated charged Goldstone fields. The last diagram contains the (gauge-dependent) self-energy $\Pi_{\gamma Z}(0)$. One also needs to include the on-shell wave-function renormalization factors for the external photon fields.](graphs_ato2gamma){width="\textwidth"}
In many scenarios, the di-photon decay is the dominant decay mode of a light ALP. Because of its special importance, we have calculated the corresponding decay rate from the effective Lagrangian (\[Leff\]) including the complete set of one-loop corrections. The relevant Feynman diagrams are shown in Figure \[fig:a2gamma\]. We define an effective coefficient $C_{\gamma\gamma}^{\rm eff}$ such that $$\label{di-photonrate1}
\Gamma(a\to\gamma\gamma)
\equiv \frac{4\pi\alpha^2 m_a^3}{\Lambda^2}\,\big| C_{\gamma\gamma}^{\rm eff} \big|^2 \,.$$ To an excellent approximation (apart from a mild mass dependence in the loop corrections) the $a\to\gamma\gamma$ decay rate scales with the third power of the ALP mass. For a very light ALP with $m_a<2m_e$ this is the only SM decay mode allowed, and with decreasing ALP mass the decay rate will eventually become so small that the ALP will leave the detector and appear as an invisible particle.
The expression for $C_{\gamma\gamma}^{\rm eff}$ depends on the ALP mass | (in powers of $ v/\Lambda$) giving rise to ALP decays into pair of samarium gauge bosons and fermions, while the extra interactions in (\[LeffD>5\ ]) are needed to parametrize the exotic decay of Higgs bosons into final state involving an ALP. In computing the versatile decay rates, we include the tree - degree and one - loop contributions from the relevant operators. We find that fermion - loop corrections can be numerically significant, and they can even be dominant in new - physics model where the coefficients $ C_{VV}$ in (\[Leff\ ]) (with $ V = G, W, B$) are loop suppressed.
ALP decay into photon { # sec: agaga }
----------------------
! [ \[fig: a2gamma\ ] Representative one - loop Feynman diagram contributing to the decay $ a\to\gamma\gamma$. The internal boson lines represent charge $ W$ bosons and the associated charged Goldstone fields. The last diagram contain the (bore - dependent) self - energy $ \Pi_{\gamma Z}(0)$. One also needs to include the on - shell wave - function renormalization gene for the external photon fields.](graphs_ato2gamma){width="\textwidth " }
In many scenario, the di - photon decay is the dominant decay mode of a faint ALP. Because of its limited importance, we have calculated the corresponding decay rate from the effective Lagrangian (\[Leff\ ]) including the complete set of one - loop correction. The relevant Feynman diagrams are shown in Figure \[fig: a2gamma\ ]. We define an effective coefficient $ C_{\gamma\gamma}^{\rm eff}$ such that $ $ \label{di - photonrate1 }
\Gamma(a\to\gamma\gamma)
\equiv \frac{4\pi\alpha^2 m_a^3}{\Lambda^2}\,\big| C_{\gamma\gamma}^{\rm eff } \big|^2 \,.$$ To an excellent approximation (apart from a mild aggregate dependence in the loop corrections) the $ a\to\gamma\gamma$ decay rate scales with the third power of the ALP mass. For a very idle ALP with $ m_a<2m_e$ this is the only SM decay mode allowed, and with decreasing ALP bulk the decay rate will eventually become so humble that the ALP will leave the detector and appear as an invisible particle.
The construction for $ C_{\gamma\gamma}^{\rm eff}$ depends on the ALP multitude | (in powers of $v/\Lambda$) givinn rise to ALP deeqys invo paira of SM eauge bosons and fermions, whmle rhe aeditional interactions in (\[LeffD&ht;5\]) are nwedev to parametrize the exobnc dedwys mh Higgs bosons lnto final vtates involvitg au ALP. In computing the various decay rates, ee include the ttee-lenej ans one-loop contributions from the rslevant operators. We find that fermion-loop corgectlons can be numerifally imporjznt, qnd they can even be dominant in ngw-physics models where the coeffkciencs $C_{VV}$ in (\[Lgyd\]) (wljh $V=G,W,B$) are ooop fuppressed.
ALI decay into pnotons {#sec:agags}
----------------------
![\[fij:a2ganma\] Representative onx-loop Feynman diagrais contritucing to the decay $a\to\tanma\gakma$. Dhe kbtefnam uoskn linfs cepresent cgarged $W$ bowons and the associstqe charged Golsstone fyelds. The last diagram contains the (gauce-dspendent) self-energy $\Pi_{\gqmma Z}(0)$. One also needs to incluqe the on-shell wave-function renormalization factoss foc ghe ewtervql photon fields.](graphs_ato2gamma){width="\textwidth"}
In iznu xcenarios, the bi-photon decay ix hhr dominant decxy modz or a light ALP. Becakse of yts special imkortamce, we have calculated the xorrespondiny dwcay rate from the effective Uagrsngiam (\[Leff\]) including the cooplefe set of oje-loop codfections. The reldvakt Xeynman diagrams are shown in Figurx \[fig:a2yamma\]. We defone an effective coefnhcient $C_{\gamma\gamma}^{\gm efy}$ sucv that $$\labfl{di-photonrate1}
\Gamma(a\to\gamma\gekma)
\equiv \fsac{4\ii\alpha^2 m_c^3}{\Lambds^2}\,\big| C_{\gamma\gaima}^{\rm eff} \big|^2 \,.$$ To an zxcelldnt approxpmation (a'art from a iild mass depajdence in thx loop cowrecrionw) the $a\gu\gamma\gamma$ devay rate scales wity the third power pf jhs ALP mass. For c cery light ALP eitf $m_w<2m_v$ tiis if the only SM decxy opde aulowed, and cigh drcreasing ALP mass tve dscay rate will evemtmally becime so siall that the ALP will leave tht detertor aid apprar as an invisible particle.
The espression fov $C_{\gamma\gamma}^{\wm enf}$ dgpends on tke ALP mass | (in powers of $v/\Lambda$) giving rise to into of SM bosons and fermions, (\[LeffD>5\]) needed to parametrize exotic decays of bosons into final states involving an In computing the various decay rates, we include the tree-level and one-loop contributions the relevant operators. We find that fermion-loop corrections can be numerically important, and can be in models where the coefficients $C_{VV}$ in (\[Leff\]) (with $V=G,W,B$) are loop suppressed. ALP decay into photons ---------------------- ![\[fig:a2gamma\] Representative one-loop Feynman diagrams contributing to decay $a\to\gamma\gamma$. The internal lines represent charged $W$ bosons the charged Goldstone The diagram the (gauge-dependent) self-energy Z}(0)$. One also needs to include the on-shell wave-function renormalization factors for the external photon fields.](graphs_ato2gamma){width="\textwidth"} In scenarios, the is the decay of light ALP. Because special importance, we have calculated the from the effective Lagrangian (\[Leff\]) including the complete of one-loop The relevant Feynman diagrams are shown Figure \[fig:a2gamma\]. We define an effective coefficient $C_{\gamma\gamma}^{\rm such that $$\label{di-photonrate1} \Gamma(a\to\gamma\gamma) \equiv \frac{4\pi\alpha^2 m_a^3}{\Lambda^2}\,\big| C_{\gamma\gamma}^{\rm eff} \big|^2 \,.$$ To an excellent approximation a mild mass dependence the loop corrections) $a\to\gamma\gamma$ rate with third power the ALP mass. For a very light ALP with $m_a<2m_e$ this the only SM decay mode allowed, and with decreasing ALP decay will eventually become small that the ALP leave detector and appear as particle. expression depends the mass | (in powers of $v/\Lambda$) giving riSe to ALP decAys inTo pAirS oF SM gAuge Bosons and fermiONs, whIle the additional interaCtionS iN (\[leffd≫5\]) Are neEded to pARaMETriZe ThE exOtIC dEcays Of HIggs bosOns into finAl sTaTes involving AN AlP. In computIng The various deCay Rates, wE iNclUDe the TreE-leveL and onE-Loop coNtributioNs FRom the RElevant OPErAtorS. We find that fermioN-LoOP corrections caN be numErICaLLY imPorTant, and theY cAn eveN Be dominANt IN NEw-pHYsics models whEre the coeffICieNts $C_{VV}$ In (\[lefF\]) (With $V=G,w,B$) are LoOP suPpressed.
ALP DecaY into photOns {#sec:AGaga}
----------------------
![\[fig:A2Gamma\] RePresenTatIve One-lOOp feYnmAn DIagRAmS coNTriButing to ThE dEcay $a\To\gaMMA\GAmma$. the InteRnal bOson lines reprEseNt chARgeD $W$ bosOns anD the AsSociaTed chaRged GOlDstone fields. The Last Diagram coNtaInS thE (gAuge-dEPendenT) seLf-eNergy $\Pi_{\Gamma Z}(0)$. ONE alSo NEEDs To include the on-shelL wAVE-fUnction rEnormaLIzAtIOn factorS fOr tHe exTERnal pHotoN FiElds.](grapHs_ato2gAMmA){wIdth="\texTwIdth"}
In MaNy sCenArios, THe di-Photon Decay is tHe domINant decay mode oF A light ALP. BecaUSe OF ItS SpecIal Importance, wE havE CalcUlatED tHe cORrespOndinG dECaY Rate from the effectivE LAgrangIan (\[LeFf\]) including thE complete sET OF one-loop CorrECtIOns. The relevant feynmAn diagrams ARe shown iN FiguRe \[fig:a2gaMma\]. We defiNE An effectIve CoeFfiCieNT $c_{\gAmma\gamma}^{\rm efF}$ SUch tHaT $$\label{dI-phOtonratE1}
\GaMma(A\to\GamMa\Gamma)
\equiV \frac{4\pi\aLpHa^2 M_a^3}{\laMbdA^2}\,\big| C_{\GAmma\gammA}^{\rM efF} \bIg|^2 \,.$$ TO an exCEllent ApproXimaTiOn (AParT from a mILd MASs dePeNdEnce In tHe Loop cOrreCTioNs) the $a\tO\gamma\gamMa$ dECay rAtE sCales wiTh the third powEr Of the ALP maSs. for A very lIGHt ALP witH $m_a<2m_e$ this is the only SM decAY mode alLowEd, and With DecreasinG ALp mass tHe dECay ratE will eVentuAlLy bECOme so SMAlL thAt The ALP will LEAve The deTeCtor And appeAr as an invisible parTIclE.
The expressioN foR $C_{\gaMMA\gAmmA}^{\Rm EFf}$ dEpENds ON The ALP mass | (in powers of $v/\Lambda$ ) giving r ise t o A LPde cays int o pairs of SMg auge bosons and fermions,while t h e ad d it ional intera c ti o n s i n(\ [Le ff D &g t;5\] ) a re need ed to para met ri ze the exoti c d ecays of H igg s bosons int o f inal s ta tes invol vin g anALP. I n compu ting theva r ious d e cay rat e s ,we i nclude the tree-l e ve l and one-loopcontri bu t io n s fr omthe releva nt oper a tors. W e f i n d th a t fermion-loo p correctio n s c an benu mer i callyimpor ta n t,and they ca n ev en be dom inanti n new-p h ysics m odelswhe rethec oe ff ici en t s $ C _{ VV} $ in (\[Leff \] )(with $V= G , W , B$)are loo p sup pressed.
ALP de cayi nto phot ons { #sec :a gaga}
----- ----- -- ----------
![\ [fig :a2gamma\ ] R ep res en tativ e one-l oop Fe ynman d iagrams con tr i b u ti ng to the decay $a \t o \ ga mma\gamm a$. Th e i nt e rnal bos on li nesr e prese nt c h ar ged $W$bosons an dthe ass oc iatedch arg edGolds t onefields . The la st di a gram containst he (gauge-dep e nd e n t) self -en ergy $\Pi_{ \gam m a Z} (0)$ . O nea lso n eedsto in c lude the on-shell w av e-func tionrenormalizati on factors f o r the ex tern a lp hoton fields.] (grap hs_ato2gam m a){width ="\te xtwidth" }
In man y scenario s,the di -ph o t on decay is the d omin an t decay mo de of a li ght AL P.Be cause ofits spec ia lim po rta nce,w e have c al cul at edthe c o rrespo nding dec ay r a tefrom th e e f f ecti ve L agra ngi an (\[L eff\ ] ) i ncludin g the com ple t e se tof one-lo op correction s. The relev an t F eynman d iagramsare shown in Figure \[f i g:a2gam ma\ ]. We def ine an ef fec tive c oef f icient $C_{\ gamma \g amm a } ^{\rm e ff }$su ch that $$ \ l abe l{di- ph oton rate1} \Gamma(a\to\gam m a\g amma)
\eq uiv \fr a c {4 \pi \ al p ha^ 2m _a^ 3 } {\Lambda^2}\,\b ig| C_{\ga mm a \g amma}^{\rm eff }\big|^2 \,.$$To an excelle nt approx imation ( ap artf r oma mild mas s depend ence in t h e loo p c orrec tio ns) th e$a\ to\ga mma\ga m ma$ deca y rate s caleswithth e thirdpower of the ALP mass.For averylig ht ALP wi th$ m_a <2m_e$ th is i s the only SM de cay m ode allow ed,a nd wi t h dec reas i ng ALP ma s sthe d ec ay rate wil l e ven tuall y b e come s o sm all that the ALPw ill leave thedete c t orand appe ar as an invisib lepa r t icle.
T he expression for $C_ {\ g amma\ gamma} ^{\rmeff}$ d e p en d s on t he A LPmass | (in_powers of_$v/\Lambda$) giving rise to_ALP decays_into_pairs of_SM_gauge bosons and_fermions, while the_additional interactions in (\[LeffD>5\])_are needed to_parametrize_the exotic decays of Higgs bosons into final states involving an ALP. In computing_the_various decay_rates,_we_include the tree-level and one-loop_contributions from the relevant operators._We find_that fermion-loop corrections can be numerically important, and_they_can even be_dominant in new-physics models where the coefficients $C_{VV}$ in_(\[Leff\]) (with $V=G,W,B$) are loop suppressed.
ALP_decay into photons_{#sec:agaga}
----------------------
![\[fig:a2gamma\]_Representative_one-loop Feynman diagrams contributing_to the decay $a\to\gamma\gamma$. The internal_boson lines represent charged $W$ bosons_and the associated charged Goldstone fields. The_last diagram contains the (gauge-dependent) self-energy_$\Pi_{\gamma Z}(0)$. One also needs_to include_the on-shell wave-function renormalization factors_for the external_photon fields.](graphs_ato2gamma){width="\textwidth"}
In_many scenarios, the_di-photon decay is the dominant decay_mode of a_light ALP. Because of its special_importance,_we have calculated_the_corresponding_decay rate_from the effective_Lagrangian_(\[Leff\]) including_the_complete set of one-loop corrections. The_relevant_Feynman diagrams are shown in Figure \[fig:a2gamma\]. We_define an effective coefficient_$C_{\gamma\gamma}^{\rm_eff}$ such that $$\label{di-photonrate1}
_ \Gamma(a\to\gamma\gamma)
_ \equiv \frac{4\pi\alpha^2 m_a^3}{\Lambda^2}\,\big| C_{\gamma\gamma}^{\rm eff}_\big|^2 \,.$$_To an_excellent approximation (apart from a mild mass dependence in the loop_corrections) the $a\to\gamma\gamma$ decay rate scales_with the third power_of the_ALP_mass. For a_very_light ALP_with $m_a<2m_e$ this is the only SM_decay mode_allowed, and with decreasing ALP mass_the decay rate will_eventually_become so small that the ALP_will leave the detector and appear_as an invisible particle.
The expression_for_$C_{\gamma\gamma}^{\rm_eff}$ depends on the ALP_mass |
OPERA collaboration, which claimed to have discovered Lorentz violation in the neutrino sector [@OPERA:2011zb]. A large number of theoretical models emerged trying to explain the observed anomaly, for example by Fermi point splitting [@Klinkhamer:2011mf], spontaneous symmetry breaking caused by the existence of a fermionic condensate [@Klinkhamer:2011iz], or a multiple Lorentz group structure [@Schreck:2011ni]. However, the physics community remained sceptical and articles were published trying to explain the result by an error source that had not been taken into account [@Contaldi:2011; @Besida:2011fi; @vanElburg:2011ze]. Unfortunately, at the 25th International Conference on Neutrino Physics and Astrophysics OPERA announced that their new measurement yields a deviation of the neutrino velocity from the speed of light, which is consistent with zero. Now again all laws of nature seem to obey Lorentz invariance.
[^2]: It is evident that the so-called nonbirefringent *Ansatz* is only nonbirefringent to first order in $\widetilde{\kappa}^{\mu\nu}$. Nevertheless we will still use the term “nonbirefringent” in order to distinguish from the nine-dimensional parameter sector of modified Maxwell theory, which shows no birefringence at least to first-order Lorentz violation, from the remaining ten coefficients. In the latter parameter region birefringent modes emerge already at first order with respect to the Lorentz-violating parameters [@KosteleckyMewes2002].
[^3]: assuming $k^2\neq 0$
[^4]: This means dropping terms that are proportional to at least one external four-momentum $k^{\mu}$, which we denote by the word “truncated”
[^5]: Also in the isotropic and the parity-even anisotropic sector of modified Maxwell theory the polarization tensor of one single transversal mode cannot be decomposed covariantly [@Klinkhamer:2010zs].
[^6]: We assume the four-vector $\zeta^{\mu}$ to be spacelike, since timelike MCS theory is expected to be nonunitary and noncausal [@AdamKlinkhamer2001].
[^7]: with nonzero parity-odd parameters $\kappa^{0213}$, $\kappa^{0123}$ (corresponding to the first two entries of the ten-dimensional vector from Eq. (8) in [@Kost | OPERA collaboration, which claimed to have discovered Lorentz violation in the neutrino sector [ @OPERA:2011zb ]. A large numeral of theoretical mannequin emerged judge to excuse the observed anomaly, for example by Fermi point splitting [ @Klinkhamer:2011mf ], ad-lib isotropy breaking caused by the existence of a fermionic condensate [ @Klinkhamer:2011iz ], or a multiple Lorentz group structure [ @Schreck:2011ni ]. However, the purgative community remained sceptical and articles were publish trying to explain the result by an error source that had not been learn into account [ @Contaldi:2011; @Besida:2011fi; @vanElburg:2011ze ]. Unfortunately, at the 25th International Conference on Neutrino Physics and Astrophysics OPERA announced that their modern measurement yields a deviation of the neutrino velocity from the speed of luminosity, which is consistent with zero. Now again all laws of nature seem to obey Lorentz invariance.
[ ^2 ]: It is apparent that the so - called nonbirefringent * Ansatz * is only nonbirefringent to first order in $ \widetilde{\kappa}^{\mu\nu}$. Nevertheless we will still use the term “ nonbirefringent ” in order to distinguish from the nine - dimensional parameter sector of modified Maxwell theory, which shows no birefringence at least to first - ordering Lorentz violation, from the remaining ten coefficient. In the latter argument region birefringent modes emerge already at first orderliness with esteem to the Lorentz - violating parameters [ @KosteleckyMewes2002 ].
[ ^3 ]: simulate $ k^2\neq 0 $
[ ^4 ]: This means dropping terms that are proportional to at least one external four - momentum $ k^{\mu}$, which we denote by the word “ truncated ”
[ ^5 ]: Also in the isotropic and the parity - even anisotropic sector of modified Maxwell theory the polarization tensor of one single cross mode cannot be decompose covariantly [ @Klinkhamer:2010zs ].
[ ^6 ]: We assume the four - vector $ \zeta^{\mu}$ to be spacelike, since timelike MCS theory is expected to be nonunitary and noncausal [ @AdamKlinkhamer2001 ].
[ ^7 ]: with nonzero parity - curious parameters $ \kappa^{0213}$, $ \kappa^{0123}$ (corresponding to the first two entry of the ten - dimensional vector from Eq. (8) in [ @Kost | OPFRA collaboration, which glaimed to have buscoveced Lorsntz viouation in the neutrino sectoc [@OPWRA:2011zb]. A large number of theuretical lodels energtd trying to explejn the observsf anmnaly, for exampke by Fermh point splitthne [@Nlinkhamer:2011mf], spontaneous symmetry bweaking cwused by the evisttncq of a fermionic condensate [@Klinkhamer:2011jz], or a multiple Lorrntz group structure [@Schrefk:2011ni]. However, the physifs communiti reiqined sceptizal and aruieles were poblished trying to explain the rdsult by an errir soktce that had not feen taken ikno accognt [@Conyaldi:2011; @Besida:2011fi; @vaiElbyrg:2011ze]. Unfortunately, av the 25th Internationwl Conferanee on Neutrino Physicw qnd Avtro[hyskxs UPEDA ahnouncfd vhat their hew measurenent yields a deviauiog of the neutrjno vejosity from the speed of light, which is cmnsjstent with zero. Now agqin all laws of naturg seem to jbey Lorentz invariance.
[^2]: It is evident that the so-wallev vonylvefrkbgfnt *Ansatz* is only nonbirefringent to first owset pn $\widetilde{\kappa}^{\iu\nu}$. Nevertnepexf we will stiul use ths term “nonbirefrinhent” in ordee to distynguosh from the nine-dimensionao parameter fwctor of modified Laxwell theury, ehich shows no birefringence at meast to figst-order Murentz violation, frpm the remaining ten coefficyents. In vhe lctter pafamejer regyon birefrlngenb modes emerge alrewdy aj firsd order wihh respect to the Lorentz-violatmig parameters [@KmstvleckyMewzs2002].
[^3]: assmming $k^2\neq 0$
[^4]: Thif means droppiug terms that are propogtional tm at least jne external xlur-momentum $n^{\mu}$, whicr we denite by gfe word “truncayed”
[^5]: Also pn the isotropic and the patify-even anisotro'nc sector of modigiea Mwxaekl dheory the pmlarkzagoon tdnsor of onc sknglr transversal mode cdnnof be decomposed cofaviantly [@Koinkhamew:2010zs].
[^6]: We assume the four-vector $\zeua^{\mu}$ tm bx spacrlihe, since timelike MCS theory ia expectef tj be nonunitwry qnd noncausak [@AdamKlinkhamer2001].
[^7]: with nonzero parity-odd parameters $\kappa^{0213}$, $\jappa^{0123}$ (corresponding uo the first vwo eneries of dhe ten-dimensional vwctor from Eq. (8) in [@Lost | OPERA collaboration, which claimed to have discovered in neutrino sector A large number to the observed anomaly, example by Fermi splitting [@Klinkhamer:2011mf], spontaneous symmetry breaking caused the existence of a fermionic condensate [@Klinkhamer:2011iz], or a multiple Lorentz group structure However, the physics community remained sceptical and articles were published trying to explain result an source had not been taken into account [@Contaldi:2011; @Besida:2011fi; @vanElburg:2011ze]. Unfortunately, at the 25th International Conference on Physics and Astrophysics OPERA announced that their new yields a deviation of neutrino velocity from the speed light, is consistent zero. again laws of nature to obey Lorentz invariance. [^2]: It is evident that the so-called nonbirefringent *Ansatz* is only nonbirefringent to order in we will use term in order to the nine-dimensional parameter sector of modified shows no birefringence at least to first-order Lorentz from the ten coefficients. In the latter parameter birefringent modes emerge already at first order with to the Lorentz-violating parameters [@KosteleckyMewes2002]. [^3]: assuming $k^2\neq 0$ [^4]: This means dropping terms that to at least one four-momentum $k^{\mu}$, which denote the “truncated” Also in isotropic and the parity-even anisotropic sector of modified Maxwell theory the tensor of one single transversal mode cannot be decomposed covariantly We the four-vector $\zeta^{\mu}$ be spacelike, since timelike theory expected to be nonunitary [@AdamKlinkhamer2001]. with $\kappa^{0213}$, (corresponding the first two entries the ten-dimensional vector from Eq. in [@Kost | OPERA collaboration, which clAimed to havE discOveRed loRentZ vioLation in the neuTRino Sector [@OPERA:2011zb]. A large numBer of ThEOretICaL modeLs emergED tRYIng To ExPlaIn THe ObserVed Anomaly, For example By FErMi point splitTInG [@KlinkhameR:2011mf], Spontaneous sYmmEtry brEaKinG CauseD by The exIstencE Of a ferMionic conDeNSate [@KlINkhamer:2011IZ], Or A mulTiple Lorentz group STrUCture [@Schreck:2011ni]. howeveR, tHE pHYSicS coMmunity remAiNed scEPtical aND aRTICleS Were published Trying to expLAin The resUlT by AN error SourcE tHAt hAd not been taKen iNto accounT [@ContaLDi:2011; @BesidA:2011Fi; @vanElBurg:2011ze]. unfOrtUnatELy, At The 25Th iNteRNaTioNAl COnferencE oN NEutriNo PhYSICS and astRophYsics oPERA announceD thAt thEIr nEw meaSuremEnt yIeLds a dEviatiOn of tHe Neutrino velocitY froM the speed Of lIgHt, wHiCh is cONsisteNt wIth Zero. Now Again alL LawS oF NATuRe seem to obey LorentZ iNVArIance.
[^2]: It iS evideNT tHaT The so-calLeD noNbirEFRingeNt *AnSAtZ* is only nOnbireFRiNgEnt to fiRsT order In $\WidEtiLde{\kaPPa}^{\mu\Nu}$. NeveRtheless We wilL Still use the terM “NonbirefringeNT” iN ORdER to dIstInguish from The nINe-diMensIOnAl pARametEr secToR Of MOdified Maxwell theorY, wHich shOws no Birefringence At least to fIRST-order LoRentZ ViOLation, from the rEmainIng ten coefFIcients. IN the lAtter parAmeter regION birefriNgeNt mOdeS emERGe Already at firsT ORder WiTh respeCt tO the LorEntZ-viOlaTinG pArameters [@kostelecKymeWeS2002].
[^3]: aSsuMing $k^2\NEq 0$
[^4]: This meAnS drOpPinG termS That arE propOrtiOnAl TO at Least onE ExTERnal FoUr-MomeNtuM $k^{\Mu}$, whiCh we DEnoTe by the Word “truncAteD”
[^5]: also In ThE isotroPic and the pariTy-Even anisotRoPic Sector OF Modified maxwell theory the polarizATion tenSor Of one SingLe transveRsaL mode cAnnOT be decOmposeD covaRiAntLY [@klinkHAMeR:2010zs].
[^6]: we Assume the fOUR-veCtor $\zEtA^{\mu}$ tO be spacElike, since timelike mcS tHeory is expectEd tO be nONUnItaRY aND noNcAUsaL [@aDamKlinkhamer2001].
[^7]: wiTh nonzero pArITy-Odd parametERs $\kApPa^{0213}$, $\kappa^{0123}$ (CorrespOndinG To the fiRst two entRies of the TeN-dimENSioNal vector fRom Eq. (8) in [@KOst | OPERA collaboration, whic h claimedto ha vedis co vere d Lo rentz violatio n inthe neutrino sector [@ OPERA :2 0 11zb ] .A lar ge numb e ro f th eo re tic al mo delseme rged tr ying to ex pla in the observe d a nomaly, fo r e xample by Fe rmi point s pli t ting[@K linkh amer:2 0 11mf], spontane ou s symme t ry brea k i ng cau sed by the existe n ce of a fermionic conde ns a te [ @Kl ink hamer:2011 iz ], or a multi p le L o ren t z group struc ture [@Schr e ck: 2011ni ]. Ho w ever,the p hy s ics communityrema ined scep ticala nd arti c les wer e publ ish edtryi n gto ex pl a int he re s ult by an e rr or sour ce t h a t hadnot bee n tak en into accou nt[@Co n tal di:20 11; @ Besi da :2011 fi; @v anElb ur g:2011ze]. Unfo rtun ately, at th e25t hInter n ationa l C onf erenceon Neut r ino P h y s ic s and Astrophysics O P E RA announc ed tha t t he i r new me as ure ment y ields a d e vi ation of the n e ut ri no velo ci ty fro mthe sp eed o f lig ht, wh ich is c onsis t ent with zero. Now again all la w s o f nat ure seem to ob ey L o rent z in v ar ian c e.
[ ^2]:It is evident that the so -c allednonbi refringent *A nsatz* iso n l y nonbir efri n ge n t to first ord er in $\widetil d e{\kappa }^{\m u\nu}$.Neverthel e s s we wil l s til l u set h eterm “nonbire f r inge nt ” in or der to dis tin gui shfro mthe nine- dimensio na lpa ra met er se c tor of m od ifi ed Ma xwell theory , whi ch s ho ws nobirefri n ge n c e at l ea st t o f ir st-or derL ore ntz vio lation, f rom there ma ining t en coefficien ts . In the l at ter param e t er regio n birefringent modes em e rge alr ead y atfirs t order w ith respe ctt o theLorent z-vio la tin g param e t er s [ @K osteleckyM e w es2 002].
[^3] : assum ing $k^2\neq 0$
[ ^ 4]: This means d rop ping t er mst ha t ar ep rop o r tional to at le ast one ex te r na l four-mom e ntu m$k^{\mu }$, whi ch we denoteby the wo rd “trunc at ed”[ ^5] : Also inthe isot ropic and the p a ri ty-ev enanisot ro pic sect or ofm odi fiedMaxwel ltheory thepo larizati on tensor of one single trans versa l m ode canno t b e de composedcova riantly [@ Kli nkh amer: 201 0 zs].
[^6 ] :Wea ssume the four-vect o r$\z e t a^ {\mu}$ to b e s pac elike , s i nce ti meli ke MCS theory ise xpected to benonu n i tar y a n d no nc ausal [@AdamKl ink ha m e r2001].
[ ^7]: with n onzero p ar i ty-od d para meters $\kapp a ^ {0 2 13}$,$\ka ppa ^{0123}$(co rr e spondin gto the fi rsttw o entr ies of thet e n-dimensional ve ctorf r om Eq . (8 ) in[@ Kost | OPERA_collaboration, which_claimed to have discovered_Lorentz violation_in_the neutrino_sector_[@OPERA:2011zb]. A large_number of theoretical_models emerged trying to_explain the observed_anomaly,_for example by Fermi point splitting [@Klinkhamer:2011mf], spontaneous symmetry breaking caused by the existence_of_a fermionic_condensate_[@Klinkhamer:2011iz],_or a multiple Lorentz group_structure [@Schreck:2011ni]. However, the physics_community remained_sceptical and articles were published trying to explain_the_result by an_error source that had not been taken into account_[@Contaldi:2011; @Besida:2011fi; @vanElburg:2011ze]. Unfortunately, at the_25th International Conference_on_Neutrino_Physics and Astrophysics OPERA_announced that their new measurement yields_a deviation of the neutrino velocity_from the speed of light, which is_consistent with zero. Now again all_laws of nature seem to_obey Lorentz_invariance.
[^2]: It is evident that_the so-called nonbirefringent_*Ansatz* is_only nonbirefringent to_first order in $\widetilde{\kappa}^{\mu\nu}$. Nevertheless we_will still use_the term “nonbirefringent” in order to_distinguish_from the nine-dimensional_parameter_sector_of modified_Maxwell theory, which_shows_no birefringence_at_least to first-order Lorentz violation, from_the_remaining ten coefficients. In the latter parameter_region birefringent modes emerge_already_at first order with_respect to the Lorentz-violating parameters_[@KosteleckyMewes2002].
[^3]: assuming $k^2\neq 0$
[^4]: This means_dropping terms_that are_proportional to at least one external four-momentum $k^{\mu}$, which we denote_by the word “truncated”
[^5]: Also in_the isotropic and the_parity-even anisotropic_sector_of modified Maxwell_theory_the polarization_tensor of one single transversal mode cannot_be decomposed_covariantly [@Klinkhamer:2010zs].
[^6]: We assume the four-vector_$\zeta^{\mu}$ to be spacelike,_since_timelike MCS theory is expected to_be nonunitary and noncausal [@AdamKlinkhamer2001].
[^7]: with_nonzero parity-odd parameters $\kappa^{0213}$, $\kappa^{0123}$_(corresponding_to_the first two entries of_the ten-dimensional vector from Eq. (8) in_[@Kost |
can be found. The fact that the construction of the candidate polyhedron fails can mean two things. Either the group is not discrete, or the choice of the basepoint is bad: it gives a Dirichlet domain with a very large number of faces (possibly even infinite see for instance [@GP2]).
Since Mostow’s work, different techniques have been developed to produce fundamental domains. In particular different classes of hypersurfaces have been used to produce faces of polyhedron. One natural idea is to generalise bisectors by replacing them by hypersurfaces that are foliated by totally geodesic subspaces. This leads to the notion of [*${\mathbb C}$-surface*]{} or [*${\mathbb R}$-surface*]{}. These surfaces are typically diffeomorphic to ${\mathbb R}\times {\bold H^{1}_{\mathbb C}}$ or ${\mathbb R}\times {\bold H^{2}_{\mathbb R}}$. Examples of these were developed in [@FZ; @FJP; @FK; @FK1; @S; @S1; @S3; @Schbook; @Wi2; @Wi7]. The ${\mathbb C}$-surfaces are somewhat easier to handle than ${\mathbb R}$-spheres because of the duality between the Grasmanian of complex lines in ${ \mbox{{\bf H}}^{2}_{\mathbb C}}$ and ${\mathbb C}P^2\setminus \overline{{ \mbox{{\bf H}}^{2}_{\mathbb C}}}$ induced by the Hermitian form. Any complex line $L$ in ${ \mbox{{\bf H}}^{2}_{\mathbb C}}$ is the projectivisation of a complex plane $P$ in ${\mathbb C}^3$, which is orthogonal for the Hermitian form to a linear subspace ${\mathbb C}v\subset {\mathbb C}^3$. The vector $v$ is called polar to $L$. As a consequence, any ${\mathbb C}$-surface corresponds to a curve in the outside of ${ \mbox{{\bf H}}^{2}_{\mathbb C}}$.
\[dualcurve\] For any ${\mathbb C}$-surface $\Sigma$, there exists a curve $\gamma: {\mathbb R}\longrightarrow {\mathbb C}P^2\setminus \overline{{ \mbox{{\bf H}}^{2}_{\mathbb C}}}$ such that $$\Sigma = \underset{t\in{\mathbb R}}{\bigcup}\gamma(t)^{\perp}$$
As seen above, bisectors are the first examples of ${\mathbb C}$-surfaces. In this case, the real spine $\sigma$ | can be found. The fact that the construction of the candidate polyhedron fails can think of two thing. Either the group is not discrete, or the choice of the basepoint is bad: it gives a Dirichlet world with a very large number of boldness (possibly even countless see for instance [ @GP2 ]).
Since Mostow ’s work, different techniques have been develop to produce cardinal domains. In particular different classes of hypersurfaces have been used to produce faces of polyhedron. One lifelike idea is to generalise bisectors by replacing them by hypersurfaces that are foliate by totally geodesic subspace. This leads to the impression of [ * $ { \mathbb C}$-surface * ] { } or [ * $ { \mathbb R}$-surface * ] { }. These surfaces are typically diffeomorphic to $ { \mathbb R}\times { \bold H^{1}_{\mathbb C}}$ or $ { \mathbb R}\times { \bold H^{2}_{\mathbb R}}$. Examples of these were developed in [ @FZ; @FJP; @FK; @FK1; @S; @S1; @S3; @Schbook; @Wi2; @Wi7 ]. The $ { \mathbb C}$-surfaces are somewhat easier to handle than $ { \mathbb R}$-spheres because of the duality between the Grasmanian of complex line in $ { \mbox{{\bf H}}^{2}_{\mathbb C}}$ and $ { \mathbb C}P^2\setminus \overline { { \mbox{{\bf H}}^{2}_{\mathbb C}}}$ induced by the Hermitian form. Any complex line $ L$ in $ { \mbox{{\bf H}}^{2}_{\mathbb C}}$ is the projectivisation of a complex plane $ P$ in $ { \mathbb C}^3 $, which is orthogonal for the Hermitian mannequin to a linear subspace $ { \mathbb C}v\subset { \mathbb C}^3$. The vector $ v$ is called polar to $ L$. As a consequence, any $ { \mathbb C}$-surface corresponds to a curve in the outside of $ { \mbox{{\bf H}}^{2}_{\mathbb C}}$.
\[dualcurve\ ] For any $ { \mathbb C}$-surface $ \Sigma$, there exists a curve $ \gamma: { \mathbb R}\longrightarrow { \mathbb C}P^2\setminus \overline { { \mbox{{\bf H}}^{2}_{\mathbb C}}}$ such that $ $ \Sigma = \underset{t\in{\mathbb R}}{\bigcup}\gamma(t)^{\perp}$$
As seen above, bisectors are the first examples of $ { \mathbb C}$-surfaces. In this event, the real spine $ \sigma$ | caj be found. The fact that the construction of tie candjdate pouyhedron fails can mean two vhints. Eiuker the group is not discrete, or the xhoire of the basepomht is bad: it flves c Virichlet domaik with a vesy large numbes uf faces (possibly even infinite see fjr instsnfe [@GP2]).
Since Mosjow’s ejrk, spfnerent techniques have been devemoped tm produce funcamental domains. In particklar different classes of hypersutracqw have been jsed to produce faces kf polyhedron. One natural idea ks to generalisg gidgctors by re'lacind them by hyivrsurfawes thay are foliated by torally geodesic subspares. This leads to thg notion ox [*${\jathbb C}$-surface*]{} oe [*${\nathbt R}$-sgrfazw*]{}. Tfest snrfzces age vypically djffeomorphix to ${\mathbb R}\times {\noje H^{1}_{\mathbb C}}$ od ${\mathfb R}\times {\bold H^{2}_{\mathbb R}}$. Examples of thest wers developed in [@FZ; @FJP; @FJ; @FK1; @S; @S1; @S3; @Schbook; @Wi2; @Wi7]. The ${\mwthbb C}$-surfaces are somewhat easier to handle that ${\matibc R}$-wpmerer bfcause of the duality between the Grasmanian jr vokplex lines in ${ \mbox{{\bf H}}^{2}_{\mstjbn C}}$ and ${\mathbb C}P^2\setminhs \overline{{ \mbox{{\bf H}}^{2}_{\mathbf C}}}$ ibduced by the Hermitian form. Any complex line $L$ in ${ \ivox{{\bf H}}^{2}_{\mathbb C}}$ is the projeccivisajion og a complex plane $P$ in ${\oathgb C}^3$, which ls orthogkval for the Hermktisn form to a linear subspace ${\mathbb C}t\subszt {\mathbc C}^3$. Jhe veceor $v$ is cwlled polar to $L$. As a cojsequgnce, aty ${\mathbb F}$-surface corresponds to a curve in the outsidg ox ${ \kbox{{\bf H}}^{2}_{\iathbn C}}$.
\[dualcurve\] Fow any ${\mathbb C}$-xurface $\Sigmx$, there expsts a cucve $\gamma: {\maehbb R}\longrigvjarrow {\mathbb C}P^2\setmigus \iveroine{{ \mbub{{\bf H}}^{2}_{\mathbb C}}}$ xuch that $$\Sigma = \uneerset{t\in{\mathbb R}}{\blgcup}\ezmma(t)^{\perp}$$
As seeu qbove, bisectors ard tre fmrst qfamples of ${\mdthbc C}$-rirfacds. In this gasd, thr real spine $\sigma$ | can be found. The fact that the the polyhedron fails mean two things. discrete, the choice of basepoint is bad: gives a Dirichlet domain with a large number of faces (possibly even infinite see for instance [@GP2]). Since Mostow’s different techniques have been developed to produce fundamental domains. In particular different classes hypersurfaces been to faces of polyhedron. One natural idea is to generalise bisectors by replacing them by hypersurfaces that foliated by totally geodesic subspaces. This leads to notion of [*${\mathbb C}$-surface*]{} [*${\mathbb R}$-surface*]{}. These surfaces are diffeomorphic ${\mathbb R}\times H^{1}_{\mathbb or R}\times {\bold H^{2}_{\mathbb Examples of these were developed in [@FZ; @FJP; @FK; @FK1; @S; @S1; @S3; @Schbook; @Wi2; @Wi7]. The C}$-surfaces are to handle ${\mathbb because the duality between of complex lines in ${ \mbox{{\bf ${\mathbb C}P^2\setminus \overline{{ \mbox{{\bf H}}^{2}_{\mathbb C}}}$ induced by Hermitian form. complex line $L$ in ${ \mbox{{\bf C}}$ is the projectivisation of a complex plane in ${\mathbb C}^3$, which is orthogonal for the Hermitian form to a linear subspace ${\mathbb C}^3$. The vector $v$ called polar to As consequence, ${\mathbb corresponds to curve in the outside of ${ \mbox{{\bf H}}^{2}_{\mathbb C}}$. \[dualcurve\] For ${\mathbb C}$-surface $\Sigma$, there exists a curve $\gamma: {\mathbb R}\longrightarrow \overline{{ H}}^{2}_{\mathbb C}}}$ such $$\Sigma = \underset{t\in{\mathbb R}}{\bigcup}\gamma(t)^{\perp}$$ seen bisectors are the first ${\mathbb In real $\sigma$ | can be found. The fact that the cOnstructioN of thE caNdiDaTe poLyheDron fails can meAN two Things. Either the group is Not diScREte, oR ThE choiCe of the BAsEPOinT iS bAd: iT gIVeS a DirIchLet domaIn with a verY laRgE number of facES (pOssibly eveN inFinite see for InsTance [@Gp2]).
SIncE mostoW’s wOrk, diFferenT TechniQues have bEeN DeveloPEd to proDUCe FundAmental domains. In pARtICular different ClasseS oF HyPERsuRfaCes have beeN uSed to PRoduce fACeS OF PolYHedron. One natuRal idea is to GEneRalise BiSecTOrs by rEplacInG TheM by hypersurFaceS that are fOliateD By totalLY geodesIc subsPacEs. THis lEAdS tO thE nOTioN Of [*${\MatHBb C}$-Surface*]{} oR [*${\mAtHbb R}$-sUrfaCE*]{}. tHEse sUrfAces Are tyPically diffeoMorPhic TO ${\maThbb R}\Times {\Bold h^{1}_{\mAthbb c}}$ or ${\matHbb R}\tImEs {\bold H^{2}_{\mathbb R}}$. EXampLes of thesE weRe DevElOped iN [@fZ; @FJP; @Fk; @FK1; @s; @S1; @S3; @schbook; @wi2; @Wi7]. The ${\MAthBb c}$-SURfAces are somewhat easIeR TO hAndle thaN ${\mathbB r}$-sPhERes becauSe Of tHe duALIty beTweeN ThE GrasmanIan of cOMpLeX lines iN ${ \mBox{{\bf H}}^{2}_{\MaThbB C}}$ aNd ${\matHBb C}P^2\SetminUs \overliNe{{ \mboX{{\Bf H}}^{2}_{\mathbb C}}}$ induCEd by the HermitIAn FORm. aNy coMplEx line $L$ in ${ \mbOx{{\bf h}}^{2}_{\MathBb C}}$ iS ThE prOJectiVisatIoN Of A Complex plane $P$ in ${\mathBb c}^3$, which Is ortHogonal for the hermitian fORM To a lineaR subSPaCE ${\mathbb C}v\subseT {\mathBb C}^3$. The vectOR $v$ is callEd polAr to $L$. As a ConsequenCE, Any ${\mathbB C}$-sUrfAce CorRESpOnds to a curve iN THe ouTsIde of ${ \mbOx{{\bF H}}^{2}_{\mathbB C}}$.
\[dUalCurVe\] FOr Any ${\mathbb c}$-surface $\siGmA$, tHeRe eXists A Curve $\gamMa: {\MatHbB R}\lOngriGHtarroW {\mathBb C}P^2\SeTmINus \OverlinE{{ \MbOX{{\Bf H}}^{2}_{\mAtHbB C}}}$ suCh tHaT $$\SigmA = \undERseT{t\in{\matHbb R}}{\bigcuP}\gaMMa(t)^{\pErP}$$
AS seen abOve, bisectors aRe The first exAmPleS of ${\matHBB C}$-surfacEs. In this case, the real spinE $\Sigma$ | can be found. The fact th at the con struc tio n o fthecand idate polyhedr o n fa ils can mean two thing s. Ei th e r th e g roupis notd is c r ete ,or th ec ho ice o f t he base point is b ad: i t gives a Di r ic hlet domai n w ith a very l arg e numb er of faces (p ossib ly eve n infin ite see f or instan c e [@GP2 ] ) .
Sin ce Mostow’s work, di f ferent techniq ues ha ve be e n de vel oped to pr od uce f u ndament a ld o m ain s . In particul ar differen t cl assesof hy p ersurf acesha v e b een used to pro duce face s of p o lyhedro n . One n atural id eais t o g en era li s e b i se cto r s b y replac in gthemby h y p e r surf ace s th at ar e foliated by to tall y ge odesi c sub spac es . Thi s lead s toth e notion of [*$ {\ma thbb C}$- sur fa ce* ]{ } or[ *${\ma thb b R }$-surf ace*]{} . Th es e s ur faces are typicall yd i ff eomorphi c to $ { \m at h bb R}\ti me s { \bol d H^{1} _{\m a th bb C}}$or ${\ m at hb b R}\ti me s {\bo ld H^ {2} _{\ma t hbbR}}$.Examples of t h ese were devel o ped in [@FZ;@ FJ P ; @ F K; @ FK1 ; @S; @S1;@S3; @Sch book ; @ Wi2 ; @Wi7 ]. Th e$ {\ m athbb C}$-surfacesar e some whateasier to han dle than $ { \ m athbb R} $-sp h er e s because of t he du ality betw e en the G rasma nian ofcomplex l i n es in ${ \m box {{\ bfH } }^ {2}_{\mathbbC } }$ a nd ${\mat hbb C}P^2\ set min us\ov er line{{ \m box{{\bf H }} ^{ 2} _{\ mathb b C}}}$ i nd uce dbythe H e rmitia n for m. A ny c o mpl ex line $L $ in $ {\m box{ {\b fH}}^{ 2}_{ \ mat hbb C}} $ is thepro j ecti vi sa tion of a complex pl an e $P$ in $ {\ mat hbb C} ^ 3 $, which is orthogonal for theH ermitia n f orm t o alinear su bsp ace ${ \ma t hbb C} v\subs et {\ ma thb b C}^3$ . Th e v ec tor $v$ is c all ed po la r to $L$. A s a consequence, a n y $ {\mathbb C}$- sur face c or res p on d s t oa cu r v e in the outsid e of ${ \m bo x {{ \bf H}}^{2 } _{\ ma thbb C} }$.
\[ dualc u rve\] F or any ${ \mathbb C }$ -sur f a ce$\Sigma$,there ex ists a cu r ve $\ g am ma: { \ma thbb R }\ lon grigh tarrow {\m athbb C}P^2 \s etminu s \ov er line{{ \ mbox{{\bf H}}^{2}_{\mat hbb C} }}$ s uch that $$\ Sig m a = \underse t{t\ in{\mathbb R} }{\ bigcu p}\ g amma( t)^{ \ pe rp} $ $
As see n above, b i se cto r s a re the firs t e xam plesof$ {\math bb C }$-surfaces. In t h is case, the r eals p ine $\ s igma $ | can_be found._The fact that the_construction of_the_candidate polyhedron_fails_can mean two_things. Either the_group is not discrete,_or the choice_of_the basepoint is bad: it gives a Dirichlet domain with a very large number_of_faces (possibly_even_infinite_see for instance [@GP2]).
Since Mostow’s_work, different techniques have been_developed to_produce fundamental domains. In particular different classes of_hypersurfaces_have been used_to produce faces of polyhedron. One natural idea is_to generalise bisectors by replacing them_by hypersurfaces that_are_foliated_by totally geodesic subspaces._This leads to the notion of_[*${\mathbb C}$-surface*]{} or [*${\mathbb R}$-surface*]{}. These_surfaces are typically diffeomorphic to ${\mathbb R}\times_{\bold H^{1}_{\mathbb C}}$ or ${\mathbb R}\times_{\bold H^{2}_{\mathbb R}}$. Examples of_these were_developed in [@FZ; @FJP; @FK;_@FK1; @S; @S1;_@S3; @Schbook;_@Wi2; @Wi7]. The_${\mathbb C}$-surfaces are somewhat easier to_handle than ${\mathbb_R}$-spheres because of the duality between_the_Grasmanian of complex_lines_in_${ \mbox{{\bf_H}}^{2}_{\mathbb C}}$ and_${\mathbb_C}P^2\setminus \overline{{_\mbox{{\bf_H}}^{2}_{\mathbb C}}}$ induced by the Hermitian_form._Any complex line $L$ in ${ \mbox{{\bf_H}}^{2}_{\mathbb C}}$ is the_projectivisation_of a complex plane_$P$ in ${\mathbb C}^3$, which_is orthogonal for the Hermitian form_to a_linear subspace_${\mathbb C}v\subset {\mathbb C}^3$. The vector $v$ is called polar to_$L$. As a consequence, any ${\mathbb_C}$-surface corresponds to a_curve in_the_outside of ${_\mbox{{\bf_H}}^{2}_{\mathbb C}}$.
\[dualcurve\]_For any ${\mathbb C}$-surface $\Sigma$, there exists_a curve_$\gamma: {\mathbb R}\longrightarrow {\mathbb C}P^2\setminus \overline{{_\mbox{{\bf H}}^{2}_{\mathbb C}}}$ such_that_$$\Sigma = \underset{t\in{\mathbb R}}{\bigcup}\gamma(t)^{\perp}$$
As seen above,_bisectors are the first examples of_${\mathbb C}$-surfaces. In this case,_the_real_spine $\sigma$ |
system weakly bound (1.475 MeV) to breaking up as $\alpha$+d, it is not obvious that a two-body Lee-Suzuki transformation would treat the relevant Jacobi coordinate responsible for the dominant energy dependence. This should be explored.\
$\bullet$ I believe the conclusions about the CG expansion will apply to other effective interactions. For example, V-low-k [@achim], a soft potential obtained by integrating out high-momentum states, is derived in a plane-wave basis, where $T$ is diagonal. Thus it should be analogous to our CG interaction, requiring a similar renormalization when embedded into a HO SM space. It would be interesting to test this conclusion.
Summary
=======
These results show that the effective interaction in the HO SM must have a very sharp dependence on the binding energy, defined as the energy of the bound state relative to the first open channel. This is typically 0 to 10 MeV for the bound states of most nuclei (and 2.22 MeV for the deuteron ground state explored here). Once this energy dependence is identified, the set of effective interaction matrix elements can be represented quite well by a CG expansion, and the results for successive LO, NLO, and NNLO calculations improve systematically.
This result suggests that the explicit energy dependence of the BH equation is almost entirely due to $QT$ – though this inference, based on the behavior of matrix elements between states with different number of HO quanta, must be tested in a case where multiple bound states exist.
We also presented a simple redefinition of the gradients associated with CG expansions, viewing the expansion as one around a momentum scale $\sim 1/b$. This definition removes operator mixing, making NNLO and higher-order fits very simple. The expansion then becomes one in nodal quantum numbers, with the coefficients of the expansion related to Talmi integrals, generalized for nonlocal interactions.
While our exploration here has been based on “data" obtained from an exact BH calculation of the effective interaction for the $av18$ potential, this raises the question, is such a potential necessary to the SM? That is, now that the success of an NNLO description of $Q$-space contributions is established, could one start with $PHP$ and determine the coefficients for such a potential directly from data, without knowledge of matrix elements of $H$ outside of $P$? I believe the answer is yes, even in cases (like the deuteron) when insufficient information is available from | system weakly bound (1.475 MeV) to breaking up as $ \alpha$+d, it is not obvious that a two - body Lee - Suzuki transformation would regale the relevant Jacobi coordinate creditworthy for the dominant energy addiction. This should be explored.\
$ \bullet$ I believe the conclusion about the CG expansion will apply to other effective interaction. For example, V - abject - k [ @achim ], a soft electric potential obtained by integrating out high - momentum states, is derive in a plane - wave footing, where $ T$ is diagonal. therefore it should be analogous to our CG interaction, want a similar renormalization when embedded into a HO SM distance. It would be interesting to test this conclusion.
Summary
= = = = = = =
These consequence show that the effective interaction in the HO samarium must have a very sharp dependence on the binding energy, define as the energy of the bound state relative to the first clear channel. This is typically 0 to 10 MeV for the bound states of most nuclei (and 2.22 MeV for the deuteron ground state explored here). Once this energy dependence is identified, the stage set of effective interaction matrix elements can be represented quite well by a CG expansion, and the resultant role for consecutive LO, NLO, and NNLO calculation improve systematically.
This result suggests that the explicit energy dependence of the BH equation is about entirely due to $ QT$ – though this inference, based on the behavior of matrix elements between states with different number of HO quanta, must be tested in a case where multiple bound country exist.
We also salute a simple redefinition of the gradients associated with CG expansions, viewing the expansion as one around a momentum plate $ \sim 1 / b$. This definition removes operator mixing, making NNLO and higher - order fits very simple. The expansion then becomes one in nodal quantum number, with the coefficients of the expansion relate to Talmi integrals, generalize for nonlocal interactions.
While our exploration here has been establish on “ data " obtain from an exact bohrium calculation of the effective interaction for the $ av18 $ potential, this raises the question, is such a likely necessary to the SM? That is, now that the success of an NNLO description of $ Q$-space contributions is established, could one start with $ PHP$ and determine the coefficients for such a potential directly from data, without knowledge of matrix component of $ H$ outside of $ P$? I believe the answer is yes, even in case (like the deuteron) when insufficient information is available from | sydtem weakly bound (1.475 MeV) tu breaking up as $\alpha$+v, it is not obvkous that a two-body Lee-Suzukm trqnsfoemation would treat thd relevann Jacobi xoorvinate responsible for tmz domjkant znxrgy dependence. This shound be explored.\
$\tuulzt$ I believe the conclusions about tre CG ecpwnsion will apkly tp othsg tffective interactions. For exampls, V-low-k [@achim], a soft potential obtained by intfgrahing out high-momenhum states, us dqeived in a puane-wave basis, where $T$ is diagonal. Thus it should be xnaloyous to our CT ijjeraction, rewuirigg a similar renormanizatiom when embeddec iito q HO SM space. It woulv be interesting to jest this wouclusion.
Summary
=======
These eewults shof thxr tfe tffxctjve inherection in tge HO SM muwt have a very shark dqirndence on tge binqigg energy, defined as the energy of the touhd state relative to thw first open channel. Jhis is ty[ically 0 to 10 MeV for the bound states of most nucnei (aid 2.22 MtV for gye deuteron ground state explored here). Once thif emeggy dependence is identified, tje fet of effectkve incsrzction matrix elemfnts cag be eepresenttd quote well by a CG expansion, qnd the resujrs for successive PO, NLO, and UNLO cslculstions improve systematncally.
Fhis result suggests ghat the explicig ekercy dependence of the BH eqtation is almpst entkreli due tj $QT$ – thouhh thlv inference, based ln thg behaeior of mahrix elements between states wivi different nombar mf HO qucnta, mmst be tested ig a case where multipje bojnd states exist.
Wx also presegted a simple tedefinition mf the gwadiwnts associxged with CG exlansions, nitwing the wxpansion as one avound z momentum scalz $\wim 1/b$. This defimitkon rvmotes o[arator mixinc, mayine NNLO and higher-ifder fits very simple. Tha exlansion then becomrs one in bodal quwntum numbers, with the coefficitnts oh the xxpansoon related to Talmi integrals, geheralized fov nonlocal ineeragtiogs.
While oux exploration here has been based on “date" obtained from an exacr BH calculation of uhe effective intewaction fmr the $av18$ potential, rhis raises the qmestion, is such a potehtial tecesdary to the SM? That is, now that the success of an NNLO description of $Q$-spqce coitwibutions ia esyablivhzd, eould oge svact with $PHP$ and dvtermine the coefficients for surh a potendicl directly from data, without kvowledge of mxtrix elements of $H$ outsjde of $P$? I believe the answer is yes, evrn in cases (like the deuberon) wheb inshfficivht information is avsilablz from | system weakly bound (1.475 MeV) to breaking $\alpha$+d, is not that a two-body relevant coordinate responsible for dominant energy dependence. should be explored.\ $\bullet$ I believe conclusions about the CG expansion will apply to other effective interactions. For example, [@achim], a soft potential obtained by integrating out high-momentum states, is derived in plane-wave where is Thus it should be analogous to our CG interaction, requiring a similar renormalization when embedded into HO SM space. It would be interesting to this conclusion. Summary ======= results show that the effective in HO SM have very dependence on the energy, defined as the energy of the bound state relative to the first open channel. This is 0 to for the states most (and 2.22 MeV deuteron ground state explored here). Once is identified, the set of effective interaction matrix can be quite well by a CG expansion, the results for successive LO, NLO, and NNLO improve systematically. This result suggests that the explicit energy dependence of the BH equation is due to $QT$ – this inference, based the of elements states with number of HO quanta, must be tested in a case where bound states exist. We also presented a simple redefinition of associated CG expansions, viewing expansion as one around momentum $\sim 1/b$. This definition mixing, NNLO very The then becomes one in quantum numbers, with the coefficients the expansion related to interactions. While our exploration here has been based “data" obtained from an exact BH calculation the effective interaction for the $av18$ potential, this raises the question, is a potential the SM? That is, now that the success an NNLO description of contributions is established, could one start with $PHP$ and the for such potential directly from without knowledge of elements of $H$ $P$? I the is cases (like the deuteron) when insufficient is from | system weakly bound (1.475 MeV) to breAking up as $\aLpha$+d, It iS noT oBvioUs thAt a two-body Lee-SUZuki Transformation would treAt the ReLEvanT jaCobi cOordinaTE rESPonSiBlE foR tHE dOminaNt eNergy dePendence. ThIs sHoUld be exploreD.\
$\BuLlet$ I belieVe tHe conclusionS abOut the cG ExpANsion WilL applY to othER effecTive interAcTIons. FoR Example, v-LOw-K [@achIm], a soft potential oBTaINed by integratiNg out hIgH-MoMENtuM stAtes, is deriVeD in a pLAne-wave BAsIS, WHerE $t$ is diagonal. ThUs it should bE AnaLogous To Our cg interActioN, rEQuiRing a similaR renOrmalizatIon wheN EmbeddeD Into a HO sM spacE. It WouLd be INtErEstInG To tESt ThiS ConClusion.
SUmMaRy
=======
TheSe reSULTS shoW thAt thE effeCtive interactIon In thE hO Sm must Have a Very ShArp dePendenCe on tHe Binding energy, deFineD as the eneRgy Of The BoUnd stATe relaTivE to The firsT open chANneL. THIS Is Typically 0 to 10 MeV for tHe BOUnD states oF most nUClEi (ANd 2.22 MeV for ThE deUterON GrounD staTE eXplored hEre). OncE ThIs Energy dEpEndencE iS idEntIfied, THe seT of effEctive inTeracTIon matrix elemeNTs can be represENtED QuITe weLl bY a CG expansiOn, anD The rEsulTS fOr sUCcessIve LO, nLo, AnD nNLO calculations impRoVe systEmatiCally.
This resuLt suggests THAT the explIcit ENeRGy dependence of The BH Equation is ALmost entIrely Due to $QT$ – tHough this INFerence, bAseD on The BehAVIoR of matrix elemENTs beTwEen statEs wIth diffEreNt nUmbEr oF Ho quanta, muSt be testEd In A cAsE whEre muLTiple bouNd StaTeS exIst.
We ALso preSenteD a siMpLe REdeFinitioN Of THE graDiEnTs asSocIaTed wiTh CG EXpaNsions, vIewing the ExpANsioN aS oNe arounD a momentum scaLe $\Sim 1/b$. This deFiNitIon remOVEs operatOr mixing, making NNLO and hiGHer-ordeR fiTs verY simPle. The expAnsIon theN beCOmes onE in nodAl quaNtUm nUMBers, wITH tHe cOeFficients oF THe eXpansIoN relAted to TAlmi integrals, generALizEd for nonlocal IntEracTIOnS.
WhILe OUr eXpLOraTIOn here has been baSed on “data" oBtAInEd from an exACt Bh cAlculatIon of thE effeCTive intEraction fOr the $av18$ poTeNtiaL, THis Raises the qUestion, iS such a potENtial NEcEssarY to The SM? THaT is, Now thAt the sUCceSs of aN NNLO dEsCriptiOn of $Q$-SpAce contrIbutions is established, coUld one Start WitH $PHP$ and deTerMIne The coeffiCienTs for such a PotEntIal diRecTLy froM datA, WiThoUT knowLedgE Of matrix eLEmEntS OF $H$ Outside of $P$? I BELIevE the aNswER is yes, Even In cases (like the deuTEron) when insuffIcieNT InfOrmATion Is Available from | system weakly bound (1.47 5 MeV) tobreak ing up a s $\ alph a$+d, it is no t obv ious that a two-body L ee-Su zu k i tr a ns forma tion wo u ld t rea tth e r el e va nt Ja cob i coord inate resp ons ib le for the d o mi nant energ y d ependence. T his shoul dbee xplor ed. \
$\b ullet$ I beli eve the c on c lusion s aboutt h eCG e xpansion will app l yt o other effect ive in te r ac t i ons . F or example ,V-low - k [@ach i m] , a so f t potential o btained byi nte gratin gout high-m oment um sta tes, is der ived in a pla ne-wav e basis, where $ T$ isdia gon al.T hu sitsh o uld be an a log ous to o ur C G int erac t i o n , re qui ring a si milar renorma liz atio n wh en em bedde d in to a HO SM sp ace.It would be inter esti ng to tes t t hi s c on clusi o n.
Su mma ry=======
These res ul t s sh ow that the effect iv e in teractio n in t h eHO SM mustha vea ve r y shar p de p en dence on the b i nd in g energ y, defin ed as th e ene r gy o f thebound st ate r e lative to thef irst open cha n ne l . T h is i s t ypically 0to 1 0 MeV for th e b o und s tates o f m o st nuclei (and 2.22 M eV for thedeuteron grou nd state e x p l ored her e).O nc e this energy d epend ence is id e ntified, theset of e ffectivei n teractio n m atr ixele m e nt s can be repr e s ente dquite w ell by a C G e xpa nsi on, a nd the re sults fo rsu cc es siv e LO, NLO, and N NLO c alc ulati o ns imp rovesyst em at i cal ly.
Th i sr e sult s ug gest s t ha t the exp l ici t energ y depende nce of t he B H equat ion is almost e ntirely du eto$QT$ – t hough th is inference, based ont he beha vio r ofmatr ix elemen tsbetwee n s t ates w ith di ffere nt nu m b er of H Oqua nt a, must be t est ed in a cas e where multiple bound st a tes exist.
We a lso pre s e nt eda s i mpl er ede f i nition of the g radients a ss o ci ated withC G e xp ansions , viewi ng th e expans ion as on e aroundamome n t umscale $\si m 1/b$.This defi n ition re moves op erator m ixi ng, m akingN NLO andhigher -o rder f its v er y simple . The expansion then be comesone i n n odal quan tum num bers, wit h th e coeffici ent s o f the ex p ansio n re l at edt o Tal mi i n tegrals,g en era l i ze d for nonlo c a l in terac tio n s.
Wh ileour exploration h e re has been ba sedo n “d ata " obt ai ned from an ex act B H calculat io n of the ef fectivein t eract ion fo r the$av18$p o te n tial,this ra ises theque st i on, issu ch a pote ntia lnecess ary to theS M ? That is, now t hat t h e succ e ssof an N NLO des c ript ion of $Q$ -space cont ributi onsis es tablish ed , coul d o ne start wit h $PHP$ an d det ermineth e co eff icient s fo r sucha po te nti al direct l y f r om d a ta, wit houtkn owle dge of ma t rix elem ent s of $H$ o uts i d e of $ P $? I believe t heanswe r is yes, ev e n in ca s es (l ike th e deut eron) w h enin suffici ent i nformatio n is avai l abl efrom | system_weakly bound_(1.475 MeV) to breaking_up as_$\alpha$+d,_it is_not_obvious that a_two-body Lee-Suzuki transformation_would treat the relevant_Jacobi coordinate responsible_for_the dominant energy dependence. This should be explored.\
$\bullet$ I believe the conclusions about the_CG_expansion will_apply_to_other effective interactions. For example,_V-low-k [@achim], a soft potential_obtained by_integrating out high-momentum states, is derived in a_plane-wave_basis, where $T$_is diagonal. Thus it should be analogous to our_CG interaction, requiring a similar renormalization_when embedded into_a_HO_SM space. It would_be interesting to test this conclusion.
Summary
=======
These_results show that the effective interaction_in the HO SM must have a_very sharp dependence on the binding_energy, defined as the energy_of the_bound state relative to the_first open channel._This is_typically 0 to_10 MeV for the bound states_of most nuclei_(and 2.22 MeV for the deuteron_ground_state explored here)._Once_this_energy dependence_is identified, the_set_of effective_interaction_matrix elements can be represented quite_well_by a CG expansion, and the results_for successive LO, NLO,_and_NNLO calculations improve systematically.
This_result suggests that the explicit_energy dependence of the BH equation_is almost_entirely due_to $QT$ – though this inference, based on the behavior of_matrix elements between states with different_number of HO quanta,_must be_tested_in a case_where_multiple bound_states exist.
We also presented a simple redefinition_of the_gradients associated with CG expansions, viewing_the expansion as one_around_a momentum scale $\sim 1/b$. This_definition removes operator mixing, making NNLO_and higher-order fits very simple._The_expansion_then becomes one in nodal_quantum numbers, with the coefficients of_the expansion related_to Talmi integrals, generalized for nonlocal interactions.
While_our_exploration here has been based on_“data"_obtained from an exact BH calculation_of_the_effective interaction for the $av18$_potential, this raises the question, is_such a potential necessary to the SM? That is,_now that the_success of an NNLO description_of_$Q$-space_contributions is established, could one start with $PHP$ and determine_the coefficients_for such a_potential directly from data, without knowledge of matrix elements of_$H$ outside of $P$? I believe the_answer is yes, even in cases (like the deuteron) when insufficient_information is available from |
^m}\rightarrow 0$ as $m\rightarrow\infty$ since $w(\varepsilon)>0$, so since $A$ is complete, the sum $\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}$ converges in $A$, and hence $a^{p^m}\rightarrow a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}\in A$.\
So let $b:=a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}$, then $b^p=a^p+(\underset{m\geq 0}{\sum}{\varepsilon^{p^m}})^p=a+\varepsilon+\underset{m\geq 1}{\sum}{\varepsilon^{p^m}}=a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}=b$ as required.
\[divided-powers\]
Let $Q=\widehat{Q(\frac{kG}{P})}$, and let $\delta_1,\cdots,\delta_r:kG\to kG$ be derivations such that $\tau\delta_i(P)\neq 0$ for all $i$. Set $N:=\{(a_1,\cdots,a_r)\in Q^r:(a_1\tau\delta_1+\cdots+a_r\tau\delta_r)(P)=0\}$.\
Then $N$ is a $Q$-bisubmodule of $Q^r$, and either $N=0$ or there exist $\alpha_1,\cdots,\alpha_r\in Z(Q)$, not all zero, such that for all $(a_1,\cdots,a_r)\in N$, $\alpha_1a_1+\cdots+\alpha_ra_r=0$.
Since $v$ is a non-commutative valuation, we have that $Q$ is simple and artinian, and the proof that $N$ is a $Q$-bisubmodule of $Q^r$ is similar to the proof of Lemma \[artinian\]. For the second statement, we will proceed using induction on $r$.\
First suppose that $r=1$, then $N$ is a two sided ideal of the simple ring $Q$, so it is either $0$ or $Q$. But if $N=Q$ then $1\in Q$ so $\tau\delta_1(P)=0$ – contradiction | ^m}\rightarrow 0 $ as $ m\rightarrow\infty$ since $ w(\varepsilon)>0 $, so since $ A$ is complete, the sum $ \underset{m\geq 0}{\sum}{\varepsilon^{p^m}}$ converges in $ A$, and hence $ a^{p^m}\rightarrow a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}\in A$ .\
So let $ b:=a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}$, then $ b^p = a^p+(\underset{m\geq 0}{\sum}{\varepsilon^{p^m}})^p = a+\varepsilon+\underset{m\geq 1}{\sum}{\varepsilon^{p^m}}=a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}=b$ as command.
\[divided - powers\ ]
lease $ Q=\widehat{Q(\frac{kG}{P})}$, and let $ \delta_1,\cdots,\delta_r: kG\to kG$ be derivations such that $ \tau\delta_i(P)\neq 0 $ for all $ i$. Set $ N:=\{(a_1,\cdots, a_r)\in Q^r:(a_1\tau\delta_1+\cdots+a_r\tau\delta_r)(P)=0\}$.\
Then $ N$ is a $ Q$-bisubmodule of $ Q^r$, and either $ N=0 $ or there exist $ \alpha_1,\cdots,\alpha_r\in Z(Q)$, not all zero, such that for all $ (a_1,\cdots, a_r)\in N$, $ \alpha_1a_1+\cdots+\alpha_ra_r=0$.
Since $ v$ is a non - commutative evaluation, we have that $ Q$ is simple and artinian, and the proof that $ N$ is a $ Q$-bisubmodule of $ Q^r$ is alike to the proof of Lemma \[artinian\ ]. For the second statement, we will continue using induction on $ r$.\
foremost suppose that $ r=1 $, then $ N$ is a two sided ideal of the simple ring $ Q$, so it is either $ 0 $ or $ Q$. But if $ N = Q$ then $ 1\in Q$ so $ \tau\delta_1(P)=0 $ – contradiction | ^m}\rihhtarrow 0$ as $m\rightarrow\lnfty$ since $w(\vargpwilon)>0$, vo sinde $A$ is zomplete, the sum $\underset{m\gee 0}{\wum}{\vaeepsilon^{p^m}}$ converges iv $A$, and hvnce $a^{p^m}\rughterrow a+\underset{m\jsq 0}{\sum}{\varepsimln^{p^m}}\nn A$.\
So let $b:=a+\undgrset{m\geq 0}{\suk}{\varepsilon^{p^m}}$, dhdn $b^p=a^p+(\underset{m\geq 0}{\sum}{\varepsilon^{p^m}})^p=a+\darepsikoj+\underset{m\geq 1}{\fum}{\vswepsjlon^{p^m}}=a+\underset{m\geq 0}{\sum}{\varepsilon^{p^j}}=b$ as rtquired.
\[divided-powets\]
Let $Q=\widehat{Q(\frac{kG}{P})}$, and let $\delta_1,\cdots,\delta_r:kH\to kG$ be dgdivwrions such tfat $\tau\delua_n(P)\neq 0$ for zll $i$. Set $N:=\{(a_1,\cdots,a_r)\in Q^r:(a_1\tau\delga_1+\cdocs+a_r\tau\deltq_r)(P)=0\}$.\
Thft $N$ is a $Q$-bmsubmoqule of $Q^r$, akc eithar $N=0$ or there exist $\akphe_1,\cdors,\alpha_r\in Z(Q)$, not all zero, such that fot all $(a_1,\cdods,c_r)\in N$, $\alpha_1a_1+\cdots+\alpya_ea_r=0$.
Sitce $e$ is q nun-ckmkufative vamuation, we have that $W$ is simple and artonyqn, and the prkof thwt $N$ is a $Q$-bisubmodule of $Q^r$ is similar tm tge proof of Lemma \[artinuan\]. For the second stwtement, wq will proceed using induction on $r$.\
First suppose dhat $c=1$, gheu $N$ is q hwo sided ideal of the simple ring $Q$, so it is siuheg $0$ or $Q$. But if $N=Q$ then $1\in Q$ xo $\tso\delta_1(P)=0$ – contrxdiction | ^m}\rightarrow 0$ as $m\rightarrow\infty$ since $w(\varepsilon)>0$, so is the sum 0}{\sum}{\varepsilon^{p^m}}$ converges in 0}{\sum}{\varepsilon^{p^m}}\in So let $b:=a+\underset{m\geq then $b^p=a^p+(\underset{m\geq 0}{\sum}{\varepsilon^{p^m}})^p=a+\varepsilon+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}=b$ as required. \[divided-powers\] Let $Q=\widehat{Q(\frac{kG}{P})}$, let $\delta_1,\cdots,\delta_r:kG\to kG$ be derivations such that $\tau\delta_i(P)\neq 0$ for all $i$. Set Q^r:(a_1\tau\delta_1+\cdots+a_r\tau\delta_r)(P)=0\}$.\ Then $N$ is a $Q$-bisubmodule of $Q^r$, and either $N=0$ or there $\alpha_1,\cdots,\alpha_r\in not zero, that for all $(a_1,\cdots,a_r)\in N$, $\alpha_1a_1+\cdots+\alpha_ra_r=0$. Since $v$ is a non-commutative valuation, we have that $Q$ simple and artinian, and the proof that $N$ a $Q$-bisubmodule of $Q^r$ similar to the proof of \[artinian\]. the second we proceed induction on $r$.\ suppose that $r=1$, then $N$ is a two sided ideal of the simple ring $Q$, so it either $0$ But if then Q$ $\tau\delta_1(P)=0$ – contradiction | ^m}\rightarrow 0$ as $m\rightarrow\iNfty$ since $w(\VarepSilOn)>0$, sO sInce $a$ is cOmplete, the sum $\uNDersEt{m\geq 0}{\sum}{\varepsilon^{p^m}}$ cOnverGeS In $A$, aND hEnce $a^{P^m}\rightARrOW A+\unDeRsEt{m\GeQ 0}{\SuM}{\varePsiLon^{p^m}}\in a$.\
So let $b:=a+\unDerSeT{m\geq 0}{\sum}{\varePSiLon^{p^m}}$, then $b^P=a^p+(\Underset{m\geq 0}{\Sum}{\VarepsIlOn^{p^M}})^P=a+\varEpsIlon+\uNderseT{M\geq 1}{\suM}{\varepsilOn^{P^M}}=a+\undeRSet{m\geq 0}{\SUM}{\vArepSilon^{p^m}}=b$ as requireD.
\[DiVIded-powers\]
Let $Q=\WidehaT{Q(\FRaC{Kg}{P})}$, aNd lEt $\delta_1,\cdoTs,\Delta_R:KG\to kG$ bE DeRIVAtiONs such that $\tau\Delta_i(P)\neq 0$ fOR alL $i$. Set $N:=\{(A_1,\cDotS,A_r)\in Q^r:(A_1\tau\dElTA_1+\cdOts+a_r\tau\delTa_r)(P)=0\}$.\
then $N$ is a $Q$-BisubmODule of $Q^R$, And eithEr $N=0$ or tHerE exIst $\aLPhA_1,\cDotS,\aLPha_R\In z(Q)$, nOT alL zero, sucH tHaT for aLl $(a_1,\cDOTS,A_r)\in n$, $\alPha_1a_1+\Cdots+\Alpha_ra_r=0$.
Since $V$ is A non-COmmUtatiVe valUatiOn, We havE that $Q$ Is simPlE and artinian, and The pRoof that $N$ Is a $q$-bIsuBmOdule OF $Q^r$ is sImiLar To the prOof of LeMMa \[aRtINIAn\]. for the second statemEnT, WE wIll proceEd usinG InDuCTion on $r$.\
FIrSt sUppoSE That $r=1$, Then $n$ Is A two sideD ideal OF tHe Simple rInG $Q$, so it Is EitHer $0$ Or $Q$. BuT If $N=Q$ Then $1\in q$ so $\tau\deLta_1(P)=0$ – cONtradiction | ^m}\rightarrow 0$ as $m\ri ghtarrow\i nfty$ si nce $ w(\v arep silon)>0$, sos ince $A$ is complete, thesum $ \u n ders e t{ m\geq 0}{\su m }{ \ v are ps il on^ {p ^ m} }$ co nve rges in $A$, andhen ce $a^{p^m}\ri g ht arrow a+\u nde rset{m\geq 0 }{\ sum}{\ va rep s ilon^ {p^ m}}\i n A$.\ So let $b:=a+\u nd e rset{m \ geq 0}{ \ s um }{\v arepsilon^{p^m}}$ , t h en $b^p=a^p+(\ unders et { m\ g e q 0 }{\ sum}{\vare ps ilon^ { p^m}})^ p =a + \ v are p silon+\unders et{m\geq 1} { \su m}{\va re psi l on^{p^ m}}=a +\ u nde rset{m\geq0}{\ sum}{\var epsilo n ^{p^m}} = b$ as r equire d.
\[ divi d ed -p owe rs \ ]
L et $Q = \wi dehat{Q( \f ra c{kG} {P}) } $ , andlet $\d elta_ 1,\cdots,\del ta_ r:kG \ tokG$ b e der ivat io ns su ch tha t $\t au \delta_i(P)\neq 0$for all $ i$. S et$N :=\{( a _1,\cd ots ,a_ r)\in Q ^r:(a_1 \ tau \d e l t a_ 1+\cdots+a_r\tau\d el t a _r )(P)=0\} $.\
Th e n$N $ is a $Q $- bis ubmo d u le of $Q^ r $, and eit her $N = 0$ o r there e xist $ \a lph a_1 ,\cdo t s,\a lpha_r \in Z(Q) $, no t all zero, suc h that for all $( a _ 1, \ cdot s,a _r)\in N$,$\al p ha_1 a_1+ \ cd ots + \alph a_ra_ r= 0 $.
Since $v$ is a non -c ommuta tivevaluation, we have that $ Q $ is sim plea nd artinian, andthe p roof that$ N$ is a$Q$-b isubmodu le of $Q^ r $ is simi lar to th e p r o of of Lemma \[a r t inia n\ ]. Forthe second st ate men t,we will pro ceed usi ng i nd uc tio n on$ r$.\
Fir st su pp ose that $r=1$, then $N$ i sa tw o sided id e a l of t he sim ple r ing $ Q$,s o i t is ei ther $0$or$ Q$.Bu tif $N=Q $ then $1\inQ$ so $\tau\ de lta _1(P)= 0 $ – contr adiction | ^m}\rightarrow 0$_as $m\rightarrow\infty$_since $w(\varepsilon)>0$, so since_$A$ is_complete,_the sum_$\underset{m\geq_0}{\sum}{\varepsilon^{p^m}}$ converges in_$A$, and hence_$a^{p^m}\rightarrow a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}\in A$.\
So_let $b:=a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}$,_then_$b^p=a^p+(\underset{m\geq 0}{\sum}{\varepsilon^{p^m}})^p=a+\varepsilon+\underset{m\geq 1}{\sum}{\varepsilon^{p^m}}=a+\underset{m\geq 0}{\sum}{\varepsilon^{p^m}}=b$ as required.
\[divided-powers\]
Let $Q=\widehat{Q(\frac{kG}{P})}$, and let $\delta_1,\cdots,\delta_r:kG\to kG$ be derivations such_that_$\tau\delta_i(P)\neq 0$_for_all_$i$. Set $N:=\{(a_1,\cdots,a_r)\in Q^r:(a_1\tau\delta_1+\cdots+a_r\tau\delta_r)(P)=0\}$.\
Then $N$_is a $Q$-bisubmodule of $Q^r$,_and either_$N=0$ or there exist $\alpha_1,\cdots,\alpha_r\in Z(Q)$, not all_zero,_such that for_all $(a_1,\cdots,a_r)\in N$, $\alpha_1a_1+\cdots+\alpha_ra_r=0$.
Since $v$ is a non-commutative valuation,_we have that $Q$ is simple_and artinian, and_the_proof_that $N$ is a_$Q$-bisubmodule of $Q^r$ is similar to_the proof of Lemma \[artinian\]. For_the second statement, we will proceed using_induction on $r$.\
First suppose that $r=1$,_then $N$ is a two_sided ideal_of the simple ring $Q$,_so it is_either $0$_or $Q$. But_if $N=Q$ then $1\in Q$ so_$\tau\delta_1(P)=0$ – contradiction |
Communication
=============
According the above analysis, self-consistent is very important for category representation. However, even when $\mathfrak{A}^{\alpha}$ is self-consistent, $\mathfrak{A}^{\alpha}$ may be something wrong. For example, [@Putnam1963Brains] has stated that madmen sometimes have consistent delusional systems. Why? As the final goal of a category is to help communication, different personal conceptual systems concerned on the category $\mathfrak{A}$ maybe have different representations, which results in misunderstanding, even contradiction.
For a communication between two men $\alpha$ and $\beta$ concerned on the category $\mathfrak{A}$, two simple cases about the category $\mathfrak{A}$ are discussed as follows.
One case is that only one man knows the category $\mathfrak{A}$. Without loss of generality, it is assumed that the man $\alpha$ knows $\mathfrak{A}$, i.e. he knows $\mathfrak{A}^{\alpha}$ and the man $\beta$ has no idea of $\mathfrak{A}$, then the man $\beta$ must learn the knowledge about the category $\mathfrak{A}$ from $\mathfrak{A}^{\alpha}$.
In order to simplify learning, it is supposed that the $\mathfrak{A}^{\beta}$ and $\mathfrak{A}^{\alpha}$ have the same outer name and all the relevant conceptual systems are plain. Under such assumptions, the man $\beta$ needs to learn how to represent $\mathfrak{A}$ in $L^{\beta}=\{\mathfrak{A}^{\beta}\}$. Hence, the above question can be described as follows: Let the input representation be $(O_A^{\alpha},I_A^{\alpha},\underline{A}^{\alpha},Sim_A^{\alpha})$ and the output representation be $(O_A^{\beta},I_A^{\beta},\underline{A}^{\beta},Sim_A^{\beta})$ with respect to a learning algorithm, if a subset of $(O_A^{\alpha},I_A^{\alpha})$ or $O_A^{\alpha}$ is known, try to output $(O_A^{\beta},I_A^{\beta},\underline{A}^{\beta},Sim_A^{\beta})$, which is a standard categorization problem and has been well studied in [@JianYu2014Categorization],[@JianYu2015Categorization].
The other case is that both sides in communication know the category $\mathfrak{A}$. Assume that the representation of the category | Communication
= = = = = = = = = = = = =
According the above analysis, self - consistent is very important for category theatrical performance. However, even when $ \mathfrak{A}^{\alpha}$ is self - reproducible, $ \mathfrak{A}^{\alpha}$ may be something wrong. For example, [ @Putnam1963Brains ] has express that lunatic sometimes have consistent delusional systems. Why? As the concluding finish of a category is to help communication, different personal conceptual systems concerned on the class $ \mathfrak{A}$ maybe have different representations, which result in misunderstanding, even contradiction.
For a communication between two men $ \alpha$ and $ \beta$ concerned on the category $ \mathfrak{A}$, two simple font about the category $ \mathfrak{A}$ are discussed as be.
One case is that only one man knows the class $ \mathfrak{A}$. Without loss of generalization, it is assumed that the man $ \alpha$ knows $ \mathfrak{A}$, i.e. he knows $ \mathfrak{A}^{\alpha}$ and the homo $ \beta$ has no idea of $ \mathfrak{A}$, then the man $ \beta$ must learn the knowledge about the class $ \mathfrak{A}$ from $ \mathfrak{A}^{\alpha}$.
In ordering to simplify learning, it is supposed that the $ \mathfrak{A}^{\beta}$ and $ \mathfrak{A}^{\alpha}$ have the same outer name and all the relevant conceptual systems are plain. Under such assumptions, the man $ \beta$ necessitate to learn how to represent $ \mathfrak{A}$ in $ L^{\beta}=\{\mathfrak{A}^{\beta}\}$. therefore, the above doubt can be identify as follows: Let the stimulation representation be $ (O_A^{\alpha},I_A^{\alpha},\underline{A}^{\alpha},Sim_A^{\alpha})$ and the output representation be $ (O_A^{\beta},I_A^{\beta},\underline{A}^{\beta},Sim_A^{\beta})$ with respect to a eruditeness algorithm, if a subset of $ (O_A^{\alpha},I_A^{\alpha})$ or $ O_A^{\alpha}$ is known, try to output $ (O_A^{\beta},I_A^{\beta},\underline{A}^{\beta},Sim_A^{\beta})$, which is a standard categorization problem and has been well studied in [ @JianYu2014Categorization],[@JianYu2015Categorization ].
The other case is that both sides in communication know the category $ \mathfrak{A}$. Assume that the representation of the class |
Comlunication
=============
According the xbove analysis, self-convistenf is verh important for category repcesebtatiin. However, even when $\mxthfrak{A}^{\appha}$ is welf-ronsistent, $\mathfczk{A}^{\alpha}$ may gc somztiing wrong. For gxample, [@Putndm1963Brains] has sdageb that madmen sometimes have consistqnt delislonal systems. Rhy? Sf ths final goal of a category is to hslp comkunication, digferent personal conceptuap sydtems concerned on the categotg $\mwrhfrak{A}$ maybd have different repreaentations, which results in misjnderxtanding, eceb clttradiction.
Hor a bommunication between two mem $\alpha$ and $\beba$ coicerbed on the category $\methfrak{A}$, two simple sases abogt the category $\matyfeak{A}$ dre giscjwsea aa holmows.
Onf cese is that only one mqn knows the categoty $\nathfrak{A}$. Witgout ljsf of generality, it is assumed that the kan $\alpha$ knows $\mathfrak{A}$, u.e. he knows $\mathfrak{A}^{\wlpha}$ and the man $\beta$ has no idea of $\mathfrak{A}$, then the mdn $\beva$ muwt uwagn the knowledge about the category $\mathfrak{A}$ rrpm $\mathfrak{A}^{\alphc}$.
In order to simlllfu learning, it ks sup'kssd that the $\mathfrwk{A}^{\beta}$ and $\nathfrak{A}^{\wlphs}$ have the same outer name qnd all the gelecant conceptual sydtems are puain. Undet such assumptions, the oan $\geta$ needs ho learn guw to represent $\oatmfrdk{A}$ in $L^{\beta}=\{\mathfrak{A}^{\beta}\}$. Rence, the abore questkon van be described as follows: Let the inpuh reptesentdtion be $(O_W^{\alpha},I_A^{\alpha},\underline{A}^{\alpha},Sim_E^{\elpha})$ and the ogtplt represzntatipn be $(O_A^{\beta},I_W^{\beta},\underline{C}^{\beta},Sim_C^{\beta})$ dith respebt to a lxarning algowithm, if a sutdet of $(O_A^{\alpia},I_A^{\alpha})$ or $I_A^{\alpha}$ is yvown, try to ouyput $(O_A^{\bena},N_A^{\beta},\undwrline{A}^{\beta},Sim_A^{\bete})$, whkdh is a standarb xategorization lroclei wnv has teen well stgdiea iv [@JianHu2014Categorizctkon],[@JoanYu2015Categorization].
Tve ofher case is that nobh sides un commugication know the category $\mathvrak{A}$. Asrume yhaj the representation of the catsgory | Communication ============= According the above analysis, self-consistent important category representation. even when $\mathfrak{A}^{\alpha}$ something For example, [@Putnam1963Brains] stated that madmen have consistent delusional systems. Why? As final goal of a category is to help communication, different personal conceptual systems on the category $\mathfrak{A}$ maybe have different representations, which results in misunderstanding, even For communication two $\alpha$ and $\beta$ concerned on the category $\mathfrak{A}$, two simple cases about the category $\mathfrak{A}$ are as follows. One case is that only one knows the category $\mathfrak{A}$. loss of generality, it is that man $\alpha$ $\mathfrak{A}$, he $\mathfrak{A}^{\alpha}$ and the $\beta$ has no idea of $\mathfrak{A}$, then the man $\beta$ must learn the knowledge about the category from $\mathfrak{A}^{\alpha}$. to simplify it supposed the $\mathfrak{A}^{\beta}$ and the same outer name and all systems are plain. Under such assumptions, the man needs to how to represent $\mathfrak{A}$ in $L^{\beta}=\{\mathfrak{A}^{\beta}\}$. the above question can be described as follows: the input representation be $(O_A^{\alpha},I_A^{\alpha},\underline{A}^{\alpha},Sim_A^{\alpha})$ and the output representation be $(O_A^{\beta},I_A^{\beta},\underline{A}^{\beta},Sim_A^{\beta})$ with respect to a if a subset of or $O_A^{\alpha}$ is try output which a standard problem and has been well studied in [@JianYu2014Categorization],[@JianYu2015Categorization]. The other case that both sides in communication know the category $\mathfrak{A}$. Assume representation the category |
Communication
=============
According the Above analySis, seLf-cOnsIsTent Is veRy important for CAtegOry representation. HowevEr, eveN wHEn $\maTHfRak{A}^{\aLpha}$ is sELf-CONsiStEnT, $\maThFRaK{A}^{\alpHa}$ mAy be somEthing wronG. FoR eXample, [@Putnam1963bRaIns] has statEd tHat madmen somEtiMes havE cOnsIStent DelUsionAl systEMs. Why? AS the final GoAL of a caTEgory is TO HeLp coMmunication, differENt PErsonal conceptUal sysTeMS cONCerNed On the categOrY $\mathFRak{A}$ mayBE hAVE DifFErent represenTations, whicH ResUlts in MiSunDErstanDing, eVeN ConTradiction.
FOr a cOmmunicatIon betWEen two mEN $\alpha$ aNd $\beta$ ConCerNed oN ThE cAteGoRY $\maTHfRak{a}$, Two Simple caSeS aBout tHe caTEGORy $\maThfRak{A}$ Are diScussed as follOws.
one cASe iS that Only oNe maN kNows tHe cateGory $\mAtHfrak{A}$. Without loSs of GeneralitY, it Is AssUmEd thaT The man $\AlpHa$ kNows $\matHfrak{A}$, i.E. He kNoWS $\MAtHfrak{A}^{\alpha}$ and the mAn $\BETa$ Has no ideA of $\matHFrAk{a}$, Then the mAn $\BetA$ musT LEarn tHe knOWlEdge abouT the caTEgOrY $\mathfrAk{a}$ from $\mAtHfrAk{A}^{\Alpha}$.
iN ordEr to siMplify leArninG, It is supposed thAT the $\mathfrak{A}^{\BEtA}$ ANd $\MAthfRak{a}^{\alpha}$ have tHe saME outEr naME aNd aLL the rElevaNt COnCEptual systems are plaIn. under sUch asSumptions, the mAn $\beta$ needS TO Learn how To rePReSEnt $\mathfrak{A}$ in $l^{\beta}=\{\Mathfrak{A}^{\bETa}\}$. Hence, tHe aboVe questiOn can be deSCRibed as fOllOws: let The INPuT representatiON Be $(O_A^{\AlPha},I_A^{\alPha},\UnderliNe{A}^{\AlpHa},SIm_A^{\AlPha})$ and the Output rePrEsEnTaTioN be $(O_A^{\BEta},I_A^{\betA},\uNdeRlIne{a}^{\beta},sIm_A^{\betA})$ with RespEcT tO A leArning aLGoRIThm, iF a SuBset Of $(O_a^{\aLpha},I_a^{\alpHA})$ or $o_A^{\alpha}$ Is known, trY to OUtpuT $(O_a^{\bEta},I_A^{\beTa},\underline{A}^{\bEtA},Sim_A^{\beta})$, wHiCh iS a stanDARd categoRization problem and has beEN well stUdiEd in [@JIanYU2014CategoriZatIon],[@JiaNYu2015cAtegorIzatioN].
The oThEr cASE is thAT BoTh sIdEs in communICAtiOn knoW tHe caTegory $\mAthfrak{A}$. Assume that THe rEpresentation Of tHe caTEGoRy |
Communication
=========== ==
Accord ing t heabo ve ana lysi s, self-consis t entis very important forcateg or y rep r es entat ion. Ho w ev e r , e ve nwhe n$ \m athfr ak{ A}^{\al pha}$ is s elf -c onsistent, $ \ ma thfrak{A}^ {\a lpha}$ may b e s omethi ng wr o ng. F orexamp le, [@ P utnam1 963Brains ]h as sta t ed that m ad mensometimes have co n si s tent delusiona l syst em s .W h y?Asthe finalgo al of a categ o ry i s to help communic ation, diff e ren t pers on alc oncept ual s ys t ems concernedon t he catego ry $\m a thfrak{ A }$ mayb e have di ffe rent re pr ese nt a tio n s, wh i chresultsin m isund erst a n d i ng,eve n co ntrad iction.
Fora c ommu n ica tionbetwe en t wo men$\alph a$ an d$\beta$ concern ed o n the cat ego ry $\ ma thfra k {A}$,two si mple ca ses abo u t t he c a te gory $\mathfrak{A} $a r ediscusse d as f o ll ow s .
One c as e i s th a t only one ma n knowsthe ca t eg or y $\mat hf rak{A} $. Wi tho ut lo s s of gener ality, i t isa ssumed that th e man $\alpha$ kn o w s$ \mat hfr ak{A}$, i.e . he know s $\ m at hfr a k{A}^ {\alp ha } $a nd the man $\beta$ha s no i dea o f $\mathfrak{ A}$, thent h e man $\b eta$ mu s t learn the kn owled ge about t h e catego ry $\ mathfrak {A}$ from $ \mathfra k{A }^{ \al pha } $ .
In order tos i mpli fy learni ng, it issup pos edtha tthe $\mat hfrak{A} ^{ \b et a} $ a nd $\ m athfrak{ A} ^{\ al pha }$ ha v e thesameoute rna m e a nd allt he r elev an tconc ept ua l sys tems are plain. Under su cha ssum pt io ns, the man $\beta$ne eds to lea rn ho w to r e p resent $ \mathfrak{A}$ in $L^{\b e ta}=\{\ mat hfrak {A}^ {\beta}\} $.Hence, th e above quest ion c an be d escri b e dasfo llows: Let t heinput r epre sentati on be $(O_A^{\alph a },I _A^{\alpha},\ und erli n e {A }^{ \ al p ha} ,S i m_A ^ { \alpha})$ and t he outputre p re sentationb e $ (O _A^{\be ta},I_A ^{\be t a},\und erline{A} ^{\beta}, Si m_A^ { \ bet a})$ withrespectto a lear n ing a l go rithm , i f a su bs etof $( O_A^{\ a lph a},I_ A^{\al ph a})$ o r $O_ A^ {\alpha} $ is known, try to outp ut $(O _A^{\ bet a},I_A^{\ bet a },\ underline {A}^ {\beta},Si m_A ^{\ beta} )$, which isa s tan d ard c ateg o rizationp ro ble m an d has beenw e l l s tudie d i n [@Jia nYu2 014Categorization ] ,[@JianYu2015C ateg o r iza tio n ].
Th e other case i s t ha t both sid es in communi cation k no w thecatego ry $\m athfrak { A }$ . Assum e th atthe repre sen ta t ion ofth ec ategor y |
Communication
=============
According the_above analysis,_self-consistent is very important_for category_representation._However, even_when_$\mathfrak{A}^{\alpha}$ is self-consistent,_$\mathfrak{A}^{\alpha}$ may be_something wrong. For example,_[@Putnam1963Brains] has stated_that_madmen sometimes have consistent delusional systems. Why? As the final goal of a category_is_to help_communication,_different_personal conceptual systems concerned on_the category $\mathfrak{A}$ maybe have_different representations,_which results in misunderstanding, even contradiction.
For a communication_between_two men $\alpha$_and $\beta$ concerned on the category $\mathfrak{A}$, two simple_cases about the category $\mathfrak{A}$ are_discussed as follows.
One_case_is_that only one man_knows the category $\mathfrak{A}$. Without loss_of generality, it is assumed that_the man $\alpha$ knows $\mathfrak{A}$, i.e. he_knows $\mathfrak{A}^{\alpha}$ and the man $\beta$_has no idea of $\mathfrak{A}$,_then the_man $\beta$ must learn the_knowledge about the_category $\mathfrak{A}$_from $\mathfrak{A}^{\alpha}$.
In order_to simplify learning, it is supposed_that the $\mathfrak{A}^{\beta}$_and $\mathfrak{A}^{\alpha}$ have the same outer_name_and all the_relevant_conceptual_systems are_plain. Under such_assumptions,_the man_$\beta$_needs to learn how to represent_$\mathfrak{A}$_in $L^{\beta}=\{\mathfrak{A}^{\beta}\}$. Hence, the above question can_be described as follows:_Let_the input representation be_$(O_A^{\alpha},I_A^{\alpha},\underline{A}^{\alpha},Sim_A^{\alpha})$ and the output representation_be $(O_A^{\beta},I_A^{\beta},\underline{A}^{\beta},Sim_A^{\beta})$ with respect to a_learning algorithm,_if a_subset of $(O_A^{\alpha},I_A^{\alpha})$ or $O_A^{\alpha}$ is known, try to output $(O_A^{\beta},I_A^{\beta},\underline{A}^{\beta},Sim_A^{\beta})$,_which is a standard categorization problem_and has been well_studied in_[@JianYu2014Categorization],[@JianYu2015Categorization].
The_other case is_that_both sides_in communication know the category $\mathfrak{A}$. Assume_that the_representation of the category |
_{{\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i}} \overline{J_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i})]_{i=1}^Q$ arranged as a row vector with $\zeta_i>0$. The cost functions $\{\overline{J_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i})\}_{i=1}^Q$ satisfy the diagonal strict convexity property if ${\boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}})$ is strictly decreasing in ${\boldsymbol{w}}$ for some positive vector ${\boldsymbol{\zeta}}$, that is, $({\boldsymbol{g}}(\hat{\boldsymbol{w}},{\boldsymbol{\zeta}})-{\boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}}))^\top(\hat{\boldsymbol{w}}-{\boldsymbol{w}})<0$ for all nonequal ${\boldsymbol{w}}$, $\hat{\boldsymbol{w}}$.
---
abstract: 'In this note, we combine ideas of several previous proofs in order to obtain a quite short proof of Grötzsch theorem.'
author:
- 'Zdeněk Dvořák[^1]'
title: A simplified discharging proof of Grötzsch theorem
---
Grötzsch [@Gro] proved that every planar triangle-free graph is $3$-colorable, using the discharging method. This proof was simplified by Thomassen [@ThoGro] (who also gave a principally different proof [@ThoShortlist]). Dvořák et al. [@DKT] give another variation of the discharging proof. Both of the later arguments were developed in order to obtain more general results (the Thomassen’s proof gives extensions to girth $5$ graphs in the torus and the projective plane, while the proof of Dvořák et al. aims at algorithmic applications), and thus their presentation of the proof of Gr | _ { { \boldsymbol{w}}_{{{\ifmmode { \mathcal{C}}\else $ { \mathcal{C}}$\fi}}_i } } \overline{J_{{{\ifmmode { \mathcal{C}}\else $ { \mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode { \mathcal{C}}\else $ { \mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmode { \mathcal{C}}\else $ { \mathcal{C}}$\fi}}_i})]_{i=1}^Q$ arranged as a row vector with $ \zeta_i>0$. The cost functions $ \{\overline{J_{{{\ifmmode { \mathcal{C}}\else $ { \mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode { \mathcal{C}}\else $ { \mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmode { \mathcal{C}}\else $ { \mathcal{C}}$\fi}}_i})\}_{i=1}^Q$ meet the aslant stern convexity property if $ { \boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}})$ is rigorously decreasing in $ { \boldsymbol{w}}$ for some positive vector $ { \boldsymbol{\zeta}}$, that is, $ ({ \boldsymbol{g}}(\hat{\boldsymbol{w}},{\boldsymbol{\zeta}})-{\boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}}))^\top(\hat{\boldsymbol{w}}-{\boldsymbol{w}})<0 $ for all nonequal $ { \boldsymbol{w}}$, $ \hat{\boldsymbol{w}}$.
---
abstract:' In this note, we blend ideas of several former proof in order to prevail a quite short proof of Grötzsch theorem.'
generator:
-' Zdeněk Dvořák[^1 ]'
claim: A simplified discharging proof of Grötzsch theorem
---
Grötzsch [ @Gro ] proved that every planar triangle - free graph is $ 3$-colorable, use the discharging method. This proof was simplify by Thomassen [ @ThoGro ] (who also gave a chiefly different validation [ @ThoShortlist ]). Dvořák et al. [ @DKT ] give another variation of the discharging proof. Both of the later arguments were develop in order to obtain more general results (the Thomassen ’s proof move over extensions to girth $ 5 $ graphs in the torus and the projective plane, while the proof of Dvořák et al. aims at algorithmic applications), and thus their presentation of the proof of Gr | _{{\bolfsymbol{w}}_{{{\ifmmode {\mathcal{C}}\tlse ${\mathcal{C}}$\fi}}_i}} \overlinx{J_{{{\ifmmose {\mathcxl{C}}\else ${\mathcal{C}}$\fi}}_i}}({\boldsymbop{w}}_{{{\ufmmoee {\mathcal{C}}\else ${\mathcau{C}}$\fi}}_i},{\bolddymbol{w}}_{-{{\idmmove {\mathcal{C}}\else ${\mathcal{C}}$\nn}}_i})]_{i=1}^Q$ zvrangzd as a row vectpr with $\zeda_i>0$. The cost fgnztnons $\{\overline{J_{{{\ifmmode {\mathcal{C}}\else ${\mwthcal{C}}$\gi}}_l}}({\boldsymbol{w}}_{{{\ifimodt {\mwthczl{C}}\else ${\mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmkde {\matical{C}}\else ${\mathcsl{C}}$\fi}}_i})\}_{i=1}^Q$ satisfy the diagojal dtrict convexity pgoperty if ${\voldftmbol{g}}({\boldsyobol{w}},{\boldsjkbol{\zeta}})$ ia strictly decreasing in ${\boldsyobol{w}}$ for some koaihhve vector ${\uoldsyibol{\zeta}}$, thab is, $({\bondsymbok{g}}(\hat{\boldsymbok{w}},{\bmldwymbol{\zeta}})-{\boldsymbol{g}}({\uoldsymbol{w}},{\boldsymboj{\zeta}}))^\top(\hdt{\yoldsymbol{w}}-{\boldsymbol{q}})<0$ dor anl nmneqjql ${\colssbmbkl{w}}$, $\hah{\bomdsymbol{w}}$.
---
zbstract: 'In this note, we combime pceas of sevedal prqvyous proofs in order to obtain a quite vhodt proof of Grötzsch theirem.'
author:
- 'Zdeněk Dvořák[^1]'
jitle: A siiplified discharging proof of Grötzsch theorem
---
Grötzvch [@Grk] privcd tfqt every planar triangle-free graph is $3$-colorable, hsonn the discharginn method. This proog aax simplified bi Thomasseh [@ThoGro] (who also gwve a ptincipqlly difftrent proof [@ThoShortlist]). Dvořák et ql. [@DKT] give auotyer variation of tke dischargiug propf. Boyh of the later argumencs wers developed in order go obtain more gdnegal sesults (ufe Thomassen’s projf gives xxtenxions tu gitth $5$ grwphs in thf tormv and the projectige plcne, wvile the pgoof of Dvořák et al. aims at algocmthmic applicsthonv), and thbs thelr presentation of the proof pf Gr | _{{\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i}} \overline{J_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i})]_{i=1}^Q$ as a vector with $\zeta_i>0$. ${\mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode ${\mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i})\}_{i=1}^Q$ the diagonal strict property if ${\boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}})$ is strictly decreasing ${\boldsymbol{w}}$ for some positive vector ${\boldsymbol{\zeta}}$, that is, $({\boldsymbol{g}}(\hat{\boldsymbol{w}},{\boldsymbol{\zeta}})-{\boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}}))^\top(\hat{\boldsymbol{w}}-{\boldsymbol{w}})<0$ for all nonequal ${\boldsymbol{w}}$, --- abstract: 'In this note, we combine ideas of several previous proofs in to a short of Grötzsch theorem.' author: - 'Zdeněk Dvořák[^1]' title: A simplified discharging proof of Grötzsch theorem --- [@Gro] proved that every planar triangle-free graph is using the discharging method. proof was simplified by Thomassen (who gave a different [@ThoShortlist]). et al. [@DKT] another variation of the discharging proof. Both of the later arguments were developed in order to obtain general results proof gives to $5$ in the torus projective plane, while the proof of aims at algorithmic applications), and thus their presentation the proof Gr | _{{\boldsymbol{w}}_{{{\ifmmode {\mathcal{c}}\else ${\mathcAl{C}}$\fi}}_I}} \ovErlInE{J_{{{\ifMmodE {\mathcal{C}}\else ${\mAThcaL{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmodE {\mathCaL{c}}\elsE ${\MaThcal{c}}$\fi}}_i},{\bolDSyMBOl{w}}_{-{{\IfMmOde {\MaTHcAl{C}}\elSe ${\mAthcal{C}}$\Fi}}_i})]_{i=1}^Q$ arranGed As A row vector wiTH $\zEta_i>0$. The cosT fuNctions $\{\overlIne{j_{{{\ifmmoDe {\MatHCal{C}}\eLse ${\MathcAl{C}}$\fi}}_i}}({\BOldsymBol{w}}_{{{\ifmmoDe {\MAthcal{c}}\Else ${\matHCAl{c}}$\fi}}_i},{\Boldsymbol{w}}_{-{{\ifmmodE {\MaTHcal{C}}\else ${\mathcAl{C}}$\fi}}_i})\}_{I=1}^Q$ SAtISFy tHe dIagonal strIcT convEXity proPErTY IF ${\boLDsymbol{g}}({\boldsYmbol{w}},{\boldsYMboL{\zeta}})$ iS sTriCTly decReasiNg IN ${\boLdsymbol{w}}$ foR somE positive Vector ${\BOldsymbOL{\zeta}}$, thAt is, $({\boLdsYmbOl{g}}(\hAT{\bOlDsyMbOL{w}},{\bOLdSymBOl{\zEta}})-{\boldsYmBoL{g}}({\bolDsymBOL{W}},{\BoldSymBol{\zEta}}))^\toP(\hat{\boldsymboL{w}}-{\bOldsYMboL{w}})<0$ for All noNequAl ${\BoldsYmbol{w}}$, $\Hat{\boLdSymbol{w}}$.
---
abstract: 'in thIs note, we cOmbInE idEaS of seVEral prEviOus Proofs iN order tO ObtAiN A QUiTe short proof of GrötZsCH ThEorem.'
autHor:
- 'ZdeNĚk dvOŘák[^1]'
title: a sImpLifiED DischArgiNG pRoof of GrÖtzsch THeOrEm
---
GrötzScH [@Gro] prOvEd tHat Every PLanaR trianGle-free gRaph iS $3$-Colorable, using THe discharging MEtHOD. THIs prOof Was simplifiEd by tHomaSsen [@tHogro] (WHo alsO gave A pRInCIpally different prooF [@THoShorTlist]). dvořák et al. [@DKT] Give anotheR VARiation oF the DIsCHarging proof. BoTh of tHe later argUMents werE deveLoped in oRder to obtAIN more genEraL reSulTs (tHE thOmassen’s proof GIVes eXtEnsions To gIrth $5$ graPhs In tHe tOruS aNd the projEctive plAnE, wHiLe The Proof OF Dvořák eT aL. aiMs At aLgoriTHmic apPlicaTionS), aNd THus Their prESeNTAtioN oF tHe prOof Of gr | _{{\boldsymbol{w}}_{{{\ifm mode {\mat hcal{ C}} \el se ${\ math cal{C}}$\fi}}_ i }} \ overline{J_{{{\ifmmode {\ma th c al{C } }\ else${\math c al { C }}$ \f i} }_i }} ( {\ bolds ymb ol{w}}_ {{{\ifmmod e { \m athcal{C}}\e l se ${\mathca l{C }}$\fi}}_i}, {\b oldsym bo l{w } }_{-{ {\i fmmod e {\ma t hcal{C }}\else $ {\ m athcal { C}}$\fi } } _i })]_ {i=1}^Q$ arranged as a row vector w ith $\ ze t a_ i > 0$. Th e cost fun ct ions$ \{\over l in e { J _{{ { \ifmmode {\ma thcal{C}}\e l se${\mat hc al{ C }}$\fi }}_i} }( { \bo ldsymbol{w} }_{{ {\ifmmode {\mat h cal{C}} \ else ${ \mathc al{ C}} $\fi } }_ i} ,{\ bo l dsy m bo l{w } }_{ -{{\ifmm od e{\mat hcal { C } } \els e $ {\ma thcal {C}}$\fi}}_i} )\} _{i= 1 }^Q $ sat isfythedi agona l stri ct co nv exity propertyif $ {\boldsym bol {g }}( {\ bolds y mbol{w }}, {\b oldsymb ol{\zet a }}) $i s st rictly decreasingin $ {\ boldsymb ol{w}} $ f or some pos it ive vec t o r ${\ bold s ym bol{\zet a}}$,t ha tis, $({ \b oldsym bo l{g }}( \hat{ \ bold symbol {w}},{\b oldsy m bol{\zeta}})-{ \ boldsymbol{g} } ({ \ b ol d symb ol{ w}},{\bolds ymbo l {\ze ta}} ) )^ \to p (\hat {\bol ds y mb o l{w}}-{\boldsymbol{ w} })<0$for a ll nonequal $ {\boldsymb o l { w}}$, $\ hat{ \ bo l dsymbol{w}}$. ---abstract:' In thisnote, we comb ine ideas o f severa l p rev iou s p r o of s in order to o btai na quite sh ort pro ofofGrö tzs ch theorem. '
author :-'Z de něk Dvoř á k[^1]'
t it le: A si mplif i ed dis charg ingpr oo f of Grötzs c ht h eore m-- -
G röt zs ch [@ Gro] pro ved tha t every p lan a r tr ia ng le-free graph is $3$ -c olorable,us ing the d i s charging method. This proof was simplif ied by T homa ssen [@Th oGr o] (wh o a l so gav e a pr incip al lyd i ffere n t p roo f[@ThoShort l i st] ). Dv oř ák e t al. [ @DKT] give another var iation of the di scha r g in g p r oo f . B ot h of t he later argume nts were d ev e lo ped in ord e r t oobtainmore ge neral results (the Tho massen’spr oofg i ves extension s to gir th $5$ gr a phs i n t he to rus and t he pr oject ive pl a ne, whil e thepr oof of Dvoř ák et al.aims at algorithmic app licati ons), an d thus th eir pre sentation ofthe proofofGr | _{{\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else_${\mathcal{C}}$\fi}}_i}} \overline{J_{{{\ifmmode_{\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmode_{\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i})]_{i=1}^Q$_arranged_as a_row_vector with $\zeta_i>0$._The cost functions_$\{\overline{J_{{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i}}({\boldsymbol{w}}_{{{\ifmmode {\mathcal{C}}\else_${\mathcal{C}}$\fi}}_i},{\boldsymbol{w}}_{-{{\ifmmode {\mathcal{C}}\else ${\mathcal{C}}$\fi}}_i})\}_{i=1}^Q$_satisfy_the diagonal strict convexity property if ${\boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}})$ is strictly decreasing in ${\boldsymbol{w}}$ for some_positive_vector ${\boldsymbol{\zeta}}$,_that_is,_$({\boldsymbol{g}}(\hat{\boldsymbol{w}},{\boldsymbol{\zeta}})-{\boldsymbol{g}}({\boldsymbol{w}},{\boldsymbol{\zeta}}))^\top(\hat{\boldsymbol{w}}-{\boldsymbol{w}})<0$ for all nonequal ${\boldsymbol{w}}$,_$\hat{\boldsymbol{w}}$.
---
abstract: 'In this note,_we combine_ideas of several previous proofs in order to_obtain_a quite short_proof of Grötzsch theorem.'
author:
- 'Zdeněk Dvořák[^1]'
title: A simplified discharging_proof of Grötzsch theorem
---
Grötzsch [@Gro] proved that_every planar triangle-free_graph_is_$3$-colorable, using the discharging_method. This proof was simplified by_Thomassen [@ThoGro] (who also gave a principally_different proof [@ThoShortlist]). Dvořák et al. [@DKT] give another_variation of the discharging proof. Both_of the later arguments were_developed in_order to obtain more general_results (the Thomassen’s_proof gives_extensions to girth_$5$ graphs in the torus and_the projective plane,_while the proof of Dvořák et_al._aims at algorithmic_applications),_and_thus their_presentation of the_proof_of Gr |
) leads to a simple mathematical representation of the final basic equation (if a complex-valued state function is introduced). One may also say that the linearity of the equations is a consequence of this principle of maximal disorder. This is the deeper reason why it has been possible, see [@klein:schroedingers], to derive Schrödinger’s equation from a set of assumptions including linearity.
Besides the Pauli equation we found, as a second important result of our spin calculation, that the following local force is compatible with the statistical constraint: $$\label{eq:LO3DAZI8FF}
\vec{F}^{L}+\vec{F}^{I}=
e \left( \vec{E} + \frac{1}{c} \vec{v} \times \vec{B} \right)
-\vec{\mu} \cdot \frac{\partial}{\partial\vec{x} } \vec{B}
\mbox{.}$$ Here, the velocity field $\tilde{\vec{v}}(x,t)$ and the magnetic moment field $\vec{\mu}(x,t)=-(e/mc)\vec{s}(x,t)$ have been replaced by corresponding particle quantities $\vec{v}(t)$ and $\vec{\mu}(t)$; the dot denotes the inner product between $\vec{\mu}$ and $\vec{B}$. The first force in (\[eq:LO3DAZI8FF\]), the Lorentz force, has been derived here from first principles without any additional assumptions. The same cannot be said about the second force which takes this particular form as a consequence of some additional assumptions concerning the form of the ’internal force’ $\vec{F}^{R}$ \[see (\[eq:NG89NWG2EN\])\]. In particular, the field appearing in $\vec{F}^{R}$ was arbitrary as well as the proportionality constant (g-factor of the electron) and had to be adjusted by hand. It is well-known that in a relativistic treatment the spin term appears automatically if the potentials are introduced. Interestingly, this unity is not restricted to the relativistic regime. Following [@arunsalam:hamiltonians] and [@gould:intrinsic] we report in the next section an alternative (non-relativistic) derivation of spin, which does not contain any arbitrary fields or constants - but is unable to yield the expression for the macroscopic electromagnetic forces.
In the present treatment spin has been introduced as a | ) leads to a simple mathematical representation of the concluding basic equality (if a building complex - valued state function is introduce). One may also say that the one-dimensionality of the equations is a consequence of this principle of maximal disorder. This is the deep reason why it has been possible, experience [ @klein: schroedingers ], to derive Schrödinger ’s equation from a bent of assumption including linearity.
Besides the Pauli equation we found, as a second important consequence of our spin calculation, that the following local force is compatible with the statistical restraint: $ $ \label{eq: LO3DAZI8FF }
\vec{F}^{L}+\vec{F}^{I}=
e \left (\vec{E } + \frac{1}{c } \vec{v } \times \vec{B } \right)
-\vec{\mu } \cdot \frac{\partial}{\partial\vec{x } } \vec{B }
\mbox{.}$$ Here, the velocity plain $ \tilde{\vec{v}}(x, t)$ and the magnetic here and now field $ \vec{\mu}(x, t)=-(e / mc)\vec{s}(x, t)$ have been replaced by corresponding particle quantities $ \vec{v}(t)$ and $ \vec{\mu}(t)$; the acid denotes the inner product between $ \vec{\mu}$ and $ \vec{B}$. The first force in (\[eq: LO3DAZI8FF\ ]), the Lorentz force, has been derived here from first principles without any additional assumptions. The same cannot be said about the second force which takes this particular form as a consequence of some extra premise concerning the mannequin of the ’ internal force ’ $ \vec{F}^{R}$ \[see (\[eq: NG89NWG2EN\])\ ]. In particular, the field appear in $ \vec{F}^{R}$ was arbitrary as well as the proportionality constant (g - agent of the electron) and had to be adjusted by hand. It is well - known that in a relativistic treatment the spin terminus appears automatically if the potentials are introduced. Interestingly, this oneness is not restricted to the relativistic government. Following [ @arunsalam: hamiltonians ] and [ @gould: intrinsic ] we report in the next section an alternative (non - relativistic) derivation of tailspin, which does not contain any arbitrary fields or constants - but is unable to yield the expression for the macroscopic electromagnetic forces.
In the present treatment spin has been bring in as a | ) lewds to a simple mathematlcal representatnin of vhe finzl basic equation (if a complex-valued srate dunction is introduced). One may wlso say thau the linearity oh the eqmctiona is c ronsequence of jhis principne of maximal giroxder. This is the deeper reason why ie has brej possible, see [@hleim:fchrkvdlngers], to derive Schrödinger’s equztion fgom a set of assukptions including linearitj.
Besldes the Pauli equwtion we foohd, ww a second ioportant rtsblt of our apin calculation, that the folloding kocal forcg js wompatible xith tre statistical constsaint: $$\lsbel{eq:LO3DAZI8FF}
\yec{F}^{L}+\tec{F}^{U}=
e \left( \vec{E} + \frac{1}{c} \vxc{v} \times \vec{B} \right)
-\dec{\mu} \cdod \yrac{\partial}{\partial\vec{z} } \vec{B}
\kbox{.}$$ Herd, thd vtloritg fielf $\tmlde{\vec{v}}(x,t)$ znd the magbetic moment field $\fes{\nu}(x,t)=-(e/mc)\vec{s}(x,t)$ have feqn replaced by corresponding particle qlantjties $\vec{v}(t)$ and $\vec{\mu}(t)$; rhe dot denotes the ijner prodtct between $\vec{\mu}$ and $\vec{B}$. The first force in (\[eq:LO3GAZI8FH\]), ghe Ljfwnhz force, has been derived here from first prigdiklex without any cdditional assumltlomf. The same cavnot bz szid about the secojd forcg whicy takes tris larticular form as a conseqyence of somv adeitional assumptious concerniny the gorm pf the ’internal force’ $\vzc{F}^{R}$ \[sse (\[eq:NG89NWG2EN\])\]. In partidjlar, the field aopesrhng in $\vtz{F}^{R}$ was arbitrary as well es thz proporgionslity sonstant (g-vactov of the electron) ajd hab to te adjustef by hand. It is well-known that mi a relativisjic trvatment tke spik term appears wutomatically nf the pjtentkals are ihtroducxd. Interestiggly, this unidi is not restcicted to the relqtivistkz regime. Folloeing [@arunscjam:hqmiltonians] and [@goukd:ivfrinsic] we repoxu ib the next sectoon an aptxrnatyee (non-relatieistkc) arrivagion of spik, wfich does not contain anf argitrary fields or vokstants - vut is ugable to yielc the expression flr thx macruscopoc glectromagnetic forces.
In the prssent trewtmcnt spin has feen introduced cs a | ) leads to a simple mathematical representation final equation (if complex-valued state function say the linearity of equations is a of this principle of maximal disorder. is the deeper reason why it has been possible, see [@klein:schroedingers], to derive equation from a set of assumptions including linearity. Besides the Pauli equation we as second result our spin calculation, that the following local force is compatible with the statistical constraint: $$\label{eq:LO3DAZI8FF} \vec{F}^{L}+\vec{F}^{I}= \left( \vec{E} + \frac{1}{c} \vec{v} \times \vec{B} \right) \cdot \frac{\partial}{\partial\vec{x} } \vec{B} Here, the velocity field $\tilde{\vec{v}}(x,t)$ the moment field have replaced corresponding particle quantities and $\vec{\mu}(t)$; the dot denotes the inner product between $\vec{\mu}$ and $\vec{B}$. The first force in (\[eq:LO3DAZI8FF\]), Lorentz force, derived here first without additional assumptions. The be said about the second force particular form as a consequence of some additional concerning the of the ’internal force’ $\vec{F}^{R}$ \[see In particular, the field appearing in $\vec{F}^{R}$ was as well as the proportionality constant (g-factor of the electron) and had to be adjusted It is well-known that a relativistic treatment spin appears if potentials are Interestingly, this unity is not restricted to the relativistic regime. Following and [@gould:intrinsic] we report in the next section an alternative of which does not any arbitrary fields or - is unable to yield for macroscopic the treatment has been introduced as | ) leads to a simple mathematicaL representAtion Of tHe fInAl baSic eQuation (if a compLEx-vaLued state function is intRoducEd). oNe maY AlSo say That the LInEARitY oF tHe eQuATiOns is A coNsequenCe of this prIncIpLe of maximal dISoRder. This is The Deeper reason Why It has bEeN poSSible, See [@Klein:SchroeDIngers], To derive SChRÖdingeR’S equatiON FrOm a sEt of assumptions inCLuDIng linearity.
BeSides tHe pAuLI EquAtiOn we found, aS a SeconD ImportaNT rESULt oF Our spin calculAtion, that thE FolLowing LoCal FOrce is CompaTiBLe wIth the statiSticAl constraInt: $$\labEL{eq:LO3DAzi8FF}
\vec{F}^{l}+\vec{F}^{I}=
E \leFt( \vEc{E} + \fRAc{1}{C} \vEc{v} \TiMEs \vEC{B} \RigHT)
-\veC{\mu} \cdot \fRaC{\pArtiaL}{\parTIAL\Vec{x} } \Vec{b}
\mboX{.}$$ Here, The velocity fiEld $\TildE{\Vec{V}}(x,t)$ anD the mAgneTiC momeNt fielD $\vec{\mU}(x,T)=-(e/mc)\vec{s}(x,t)$ have bEen rEplaced by CorReSpoNdIng paRTicle qUanTitIes $\vec{v}(T)$ and $\vec{\MU}(t)$; tHe DOT DeNotes the inner produCt BETwEen $\vec{\mu}$ And $\vec{b}$. thE fIRst force In (\[Eq:Lo3DAZi8ff\]), the LOrenTZ fOrce, has bEen derIVeD hEre from FiRst priNcIplEs wIthouT Any aDditioNal assumPtionS. the same cannot bE Said about the sECoND FoRCe whIch Takes this paRticULar fOrm aS A cOnsEQuencE of soMe ADdITional assumptions coNcErning The foRm of the ’internAl force’ $\vec{f}^{r}$ \[SEe (\[eq:NG89NWg2EN\])\]. IN PaRTicular, the fielD appeAring in $\vec{f}^{r}$ was arbiTrary As well as The proporTIOnality cOnsTanT (g-fActOR Of The electron) anD HAd to Be AdjusteD by Hand. It iS weLl-kNowN thAt In a relatiVistic trEaTmEnT tHe sPin teRM appears AuTomAtIcaLly if THe poteNtialS are InTrODucEd. InterEStINGly, tHiS uNity Is nOt RestrIcteD To tHe relatIvistic reGimE. follOwInG [@arunsaLam:hamiltoniaNs] And [@gould:inTrInsIc] we rePORt in the nExt section an alternative (NOn-relatIviStic) dErivAtion of spIn, wHich doEs nOT contaIn any aRbitrArY fiELDs or cONStAntS - bUt is unable TO YieLd the ExPresSion for The macroscopic elecTRomAgnetic forces.
in tHe prESEnT trEAtMEnt SpIN haS BEen introduced as A | ) leads to a simple mathem atical rep resen tat ion o f th e fi nal basic equa t ion(if a complex-valued s tatefu n ctio n i s int roduced ) .O n e m ay a lso s a ythatthe linear ity of the eq ua tions is a c o ns equence of th is principle of maxim al di s order . T his i s thed eeperreason wh yi t hasb een pos s i bl e, s ee [@klein:schroe d in g ers], to deriv e Schr öd i ng e r ’sequ ation from a seto f assum p ti o n s in c luding linear ity.
Besid e s t he Pau li eq u ationwe fo un d , a s a secondimpo rtant res ult of our spi n calcul ation, th atthef ol lo win gl oca l f orc e is compati bl ewiththes t a t isti cal con strai nt: $$\label{ eq: LO3D A ZI8 FF}
\ vec{F }^{L }+ \vec{ F}^{I} =
e \ le ft( \vec{E} + \ frac {1}{c} \v ec{ v} \t im es \v e c{B} \ rig ht)
-\vec{ \mu} \c d ot\f r a c {\ partial}{\partial\ ve c { x} } \vec{ B}
\mb o x{ .} $ $ Here,th e v eloc i t y fie ld $ \ ti lde{\vec {v}}(x , t) $and the m agneti cmom ent fiel d $\v ec{\mu }(x,t)=- (e/mc ) \vec{s}(x,t)$h ave been repl a ce d by corr esp onding part icle quan titi e s$\v e c{v}( t)$ a nd $\ v ec{\mu}(t)$; the do tdenote s the inner produc t between$ \ v ec{\mu}$ and $\ v ec{B}$. The fi rst f orce in (\ [ eq:LO3DA ZI8FF \]), the Lorentzf o rce, has be ender ive d he re from first p rinc ip les wit hou t any a ddi tio nal as su mptions.The same c an no tbesaida bout the s eco nd fo rce w h ich ta kes t hispa rt i cul ar form as a con se qu ence of s ome a ddit i ona l assum ptions co nce r ning t he form o f the ’intern al force’ $\ ve c{F }^{R}$ \ [see (\[ eq:NG89NWG2EN\])\]. Inp articul ar, thefiel d appeari ngin $\v ec{ F }^{R}$ was a rbitr ar y a s wella s t hepr oportional i t y c onsta nt (g- factorof the electron) a n d h ad to be adju ste d by h an d.I ti s w el l -kn o w n that in a rel ativistictr e at ment the s p inte rm appe ars aut omati c ally if the pote ntials ar eintr o d uce d. Interes tingly,this unit y is n o trestr ict ed toth e r elati vistic reg ime.Follow in g [@ar unsal am :hamilto nians] and [@gould:intr insic] we r epo rt in the ne x t s ection an alt ernative ( non -re lativ ist i c) de riva t io n o f spin , wh i ch does n o tcon t a in any arbitr a r y fi eldsorc onstan ts - but is unable to yield the expr essi o n f ort he m ac roscopic elect rom ag n e tic forc es .
In the p resent t re a tment spinhas be en intr o d uc e d as a | ) leads_to a_simple mathematical representation of_the final_basic_equation (if_a_complex-valued state function_is introduced). One_may also say that_the linearity of_the_equations is a consequence of this principle of maximal disorder. This is the deeper_reason_why it_has_been_possible, see [@klein:schroedingers], to derive Schrödinger’s_equation from a set of_assumptions including_linearity.
Besides the Pauli equation we found, as a_second_important result of_our spin calculation, that the following local force is_compatible with the statistical constraint: $$\label{eq:LO3DAZI8FF}
\vec{F}^{L}+\vec{F}^{I}=
e_\left( \vec{E} +_\frac{1}{c}_\vec{v}_\times \vec{B} \right)
-\vec{\mu} \cdot_\frac{\partial}{\partial\vec{x} } \vec{B}
\mbox{.}$$ Here, the velocity_field $\tilde{\vec{v}}(x,t)$ and the magnetic moment_field $\vec{\mu}(x,t)=-(e/mc)\vec{s}(x,t)$ have been replaced by corresponding_particle quantities $\vec{v}(t)$ and $\vec{\mu}(t)$; the_dot denotes the inner product_between $\vec{\mu}$_and $\vec{B}$. The first force_in (\[eq:LO3DAZI8FF\]), the Lorentz_force, has_been derived here_from first principles without any additional_assumptions. The same_cannot be said about the second_force_which takes this_particular_form_as a_consequence of some_additional_assumptions concerning_the_form of the ’internal force’ $\vec{F}^{R}$_\[see (\[eq:NG89NWG2EN\])\]._In particular, the field appearing in $\vec{F}^{R}$_was arbitrary as well_as_the proportionality constant (g-factor_of the electron) and had_to be adjusted by hand. It_is well-known_that in_a relativistic treatment the spin term appears automatically if the potentials_are introduced. Interestingly, this unity is_not restricted to the_relativistic regime._Following [@arunsalam:hamiltonians]_and [@gould:intrinsic] we report_in_the next_section an alternative (non-relativistic) derivation of spin,_which does_not contain any arbitrary fields or_constants - but is_unable_to yield the expression for the_macroscopic electromagnetic forces.
In the present treatment_spin has been introduced as_a |
genceless condition implies that only two components are independent.
A similar procedure applies to the symmetric tensor $h_{ij}$, which can be decomposed as h\_[ij]{}=2C\_[ij]{}+2 \_i\_j E+2\_[(i]{}E\_[j)]{} +\_[ij]{}, \[hij\] with $\overline{E}^{ij}$ transverse and traceless (TT), i.e. $\nabla_i\overline{E}^{ij}=0$ (transverse) and $\overline{E}^{ij}\d_{ij}=0$ (traceless), and $E_i$ transverse. The parentheses around the indices denote symmetrization, namely $2\nabla_{(i}E_{j)}= \nabla_{i}E_{j} + \nabla_{j}E_{i}$. We have thus defined two scalar modes, $C$ and $E$, two vector modes, $E_i$, and two tensor modes, $\E_{ij}$.
In the following, we will be mainly interested in the metric with only [*scalar*]{} perturbations, since scalar modes are the most relevant for cosmology and they can be treated independently of the vector and tensor modes. In matrix notation, the perturbed metric will thus be of the form $$\label{scalar_pert_metric}
g_{\mu\nu} =a^2 \left[
\begin{array}{ccc}
-(1+2A) & & \nabla_iB \\ && \\
\nabla_j B & & \left\{ (1+2C) \d_{ij} + 2\nabla_i\nabla_jE \right\}
\end{array}
\right] \,.$$ After the description of the perturbed geometry, we turn to the perturbations of the matter in the next subsection.
Perturbations of the matter
---------------------------
Quite generally, the perturbed energy-momentum tensor can be written in the form $$\label{Tmunu}
T^\mu_\nu =
\left[
\begin{array}{ccc}
-(\rho+\delta\rho) & & q_j \\ &&\\
-q^i+(\rho+P)B^i & & (P+\delta P)\delta^i_j + \pi^i_j
\end{array}
\right] \,,
\label{pert_T}$$ where $q_i$ is the | genceless condition implies that only two components are autonomous.
A exchangeable procedure applies to the symmetrical tensor $ h_{ij}$, which can be decomposed as h\_[ij]{}=2C\_[ij]{}+2 \_i\_j E+2\_[(i]{}E\_[j) ] { } + \_[ij ] { }, \[hij\ ] with $ \overline{E}^{ij}$ cross and traceless (TT), i.e. $ \nabla_i\overline{E}^{ij}=0 $ (transverse) and $ \overline{E}^{ij}\d_{ij}=0 $ (traceless), and $ E_i$ transverse. The parentheses around the index denote symmetrization, namely $ 2\nabla_{(i}E_{j)}= \nabla_{i}E_{j } + \nabla_{j}E_{i}$. We have therefore defined two scalar modes, $ C$ and $ E$, two vector modes, $ E_i$, and two tensor modes, $ \E_{ij}$.
In the following, we will be chiefly interested in the metric with only [ * scalar * ] { } perturbations, since scalar mood are the most relevant for cosmology and they can be treat independently of the vector and tensor modes. In matrix notation, the perturb metric will thus be of the form $ $ \label{scalar_pert_metric }
g_{\mu\nu } = a^2 \left [
\begin{array}{ccc }
-(1 + 2A) & & \nabla_iB \\ & & \\
\nabla_j B & & \left\ { (1 + 2C) \d_{ij } + 2\nabla_i\nabla_jE \right\ }
\end{array }
\right ] \,.$$ After the description of the flustered geometry, we turn to the perturbations of the topic in the next subsection.
disturbance of the matter
---------------------------
Quite generally, the perturbed energy - momentum tensor can be written in the mannequin $ $ \label{Tmunu }
T^\mu_\nu =
\left [
\begin{array}{ccc }
-(\rho+\delta\rho) & & q_j \\ & & \\
-q^i+(\rho+P)B^i & & (P+\delta P)\delta^i_j + \pi^i_j
\end{array }
\right ] \, ,
\label{pert_T}$$ where $ q_i$ is the | genfeless condition implies that only two eimponeits are indepenaent.
A similar procedure applmes ro tht symmetric tensor $f_{ij}$, which can be eeconposed as h\_[mn]{}=2C\_[ij]{}+2 \_i\_j E+2\_[(i]{}E\_[j)]{} +\_[jm]{}, \[hij\] xith $\overline{E}^{im}$ transversa and tracelesv (GT), i.e. $\nabla_i\overline{E}^{ij}=0$ (transverse) anq $\overlonf{E}^{ij}\d_{ij}=0$ (tracelgss), amq $E_i$ nrcnsverse. The parentheses around the invices denote sykmetrization, namely $2\nabla_{(i}F_{j)}= \nwbla_{i}E_{j} + \nabla_{j}E_{i}$. Ae have thuw desuned two scauar modes, $B$ and $E$, two vector modes, $E_i$, and two tensor modex, $\E_{ij}$.
In thg ropnowing, we wmll be mainly intevvsted it the mrtric with onln [*scanar*]{} perturbations, since vcalar modes are tre most ralzvant for cosmology abd they can be geeaged iidelendenhly of the vedtor and tebsor modes. In matric gitation, the psrturbqd metric will thus be of the form $$\label{sbalad_pert_metric}
g_{\mu\nu} =a^2 \left[
\vegin{array}{ccc}
-(1+2A) & & \nablw_iB \\ && \\
\nabja_j B & & \left\{ (1+2C) \d_{ij} + 2\nabla_i\nabla_jE \right\}
\end{array}
\rhght] \,.$$ Xfttr the awsfription of the perturbed geometry, we turn to fht pvrturbations of tme matter in the nrxh xobsection.
Pertutbations or the matter
---------------------------
Quite heneraljy, thw perturbtd enrrgy-momentum tensor can be qritten in tke dorm $$\label{Tmunu}
T^\mu_\uu =
\left[
\begiu{array}{vcc}
-(\rhp+\delta\rho) & & q_j \\ &&\\
-q^i+(\rho+P)B^n & & (P+\dslta P)\delta^l_j + \pi^i_j
\eha{array}
\right] \,,
\labeu{pegt_T}$$ fhere $q_i$ is the | genceless condition implies that only two components A procedure applies the symmetric tensor as \_i\_j E+2\_[(i]{}E\_[j)]{} +\_[ij]{}, with $\overline{E}^{ij}$ transverse traceless (TT), i.e. $\nabla_i\overline{E}^{ij}=0$ (transverse) and (traceless), and $E_i$ transverse. The parentheses around the indices denote symmetrization, namely $2\nabla_{(i}E_{j)}= + \nabla_{j}E_{i}$. We have thus defined two scalar modes, $C$ and $E$, two modes, and tensor $\E_{ij}$. In the following, we will be mainly interested in the metric with only [*scalar*]{} perturbations, scalar modes are the most relevant for cosmology they can be treated of the vector and tensor In notation, the metric thus of the form g_{\mu\nu} =a^2 \left[ \begin{array}{ccc} -(1+2A) & & \nabla_iB \\ && \\ \nabla_j B & & \left\{ (1+2C) + 2\nabla_i\nabla_jE \right] \,.$$ the of perturbed geometry, we the perturbations of the matter in Perturbations of the matter --------------------------- Quite generally, the energy-momentum tensor be written in the form $$\label{Tmunu} = \left[ \begin{array}{ccc} -(\rho+\delta\rho) & & q_j \\ -q^i+(\rho+P)B^i & & (P+\delta P)\delta^i_j + \pi^i_j \end{array} \right] \,, \label{pert_T}$$ where $q_i$ is the | genceless condition implies That only twO compOneNts ArE indEpenDent.
A similar prOCeduRe applies to the symmetriC tensOr $H_{Ij}$, whICh Can be DecompoSEd AS H\_[ij]{}=2c\_[iJ]{}+2 \_i\_J E+2\_[(i]{}e\_[j)]{} +\_[IJ]{}, \[hIj\] witH $\ovErline{E}^{Ij}$ transverSe aNd Traceless (TT), i.E. $\NaBla_i\overliNe{E}^{Ij}=0$ (transverse) And $\OverliNe{e}^{ij}\D_{Ij}=0$ (traCelEss), anD $E_i$ traNSverse. the parentHeSEs arouND the indICEs DenoTe symmetrization, nAMeLY $2\nabla_{(i}E_{j)}= \nabla_{I}E_{j} + \nabLa_{J}e_{i}$. wE HavE thUs defined tWo ScalaR Modes, $C$ aND $E$, TWO VecTOr modes, $E_i$, and tWo tensor modES, $\E_{iJ}$.
In the FoLloWIng, we wIll be MaINly Interested iN the Metric witH only [*sCAlar*]{} perTUrbatioNs, sincE scAlaR modES aRe The MoST reLEvAnt FOr cOsmology AnD tHey caN be tREATEd inDepEndeNtly oF the vector and TenSor mODes. in matRix noTatiOn, The peRturbeD metrIc Will thus be of the Form $$\Label{scalAr_pErT_meTrIc}
g_{\mu\NU} =a^2 \left[
\BegIn{aRray}{ccc}
-(1+2a) & & \nabla_ib \\ && \\
\NabLa_J b & & \LEfT\{ (1+2C) \d_{ij} + 2\nabla_i\nabla_jE \RiGHT\}
\eNd{array}
\rIght] \,.$$ AfTEr ThE DescriptIoN of The pERTurbeD geoMEtRy, we turn To the pERtUrBations Of The matTeR in The Next sUBsecTion.
PeRturbatiOns of THe matter
---------------------------
Quite gENerally, the perTUrBED eNErgy-MomEntum tensor Can bE WritTen iN ThE foRM $$\labeL{TmunU}
T^\MU_\nU =
\Left[
\begin{array}{ccc}
-(\rhO+\dElta\rhO) & & q_j \\ &&\\
-q^i+(\Rho+P)B^i & & (P+\delta P)\Delta^i_j + \pi^i_J
\END{array}
\riGht] \,,
\lABeL{Pert_T}$$ where $q_i$ is The | genceless condition implie s that onl y two co mpo ne ntsareindependent.
A sim ilar procedure applies to t he symm e tr ic te nsor $h _ {i j } $,wh ic h c an be deco mpo sed ash\_[ij]{}= 2C\ _[ ij]{}+2 \_i\ _ jE+2\_[(i]{ }E\ _[j)]{} +\_[ ij] {}, \[ hi j\] with$\o verli ne{E}^ { ij}$ t ransverse a n d trac e less (T T ) ,i.e. $\nabla_i\overli n e{ E }^{ij}=0$ (tra nsvers e) an d $\o ver line{E}^{i j} \d_{i j }=0$ (t r ac e l e ss) , and $E_i$ tr ansverse. T h e p arenth es esa roundthe i nd i ces denote sym metr ization,namely $2\nabl a _{(i}E_ {j)}=\na bla _{i} E _{ j} +\n a bla _ {j }E_ { i}$ . We hav eth us de fine d t w o sc ala r mo des,$C$ and $E$,two vec t ormodes , $E_ i$,an d two tenso r mod es , $\E_{ij}$.
I n th e followi ng, w e w il l bem ainlyint ere sted in the me t ric w i t h o nly [*scalar*]{} p er t u rb ations,sinces ca la r modes a re th e mo s t rele vant fo r cosmol ogy an d t he y can b etreate dind epe ndent l y of the v ector an d ten s or modes. In m a trix notation , t h e p e rtur bed metric wil l th u s be oft he fo r m $$\ label {s c al a r_pert_metric}
g_{\ mu \nu} = a^2 \ left[
\begin{ array}{ccc } - (1+2A) & & \ n ab l a_iB \\ && \\\nabl a_j B & &\left\{(1+2C ) \d_{ij } + 2\nab l a _i\nabla _jE \r igh t\} \ en d{array}
\rig h t ] \, .$ $ After th e descr ipt ion of th eperturbed geometr y, w etu rnto th e perturb at ion softhe m a tter i n the nex tsu b sec tion.
P er t u rbat io ns ofthe m atter
--- - --- ------- --------- --- -
Qu it egeneral ly, the pertu rb ed energy- mo men tum te n s or can b e written in the form $ $ \label{ Tmu nu}
T ^\mu _\nu =
\ lef t[
\be gin { array} {ccc}-(\rh o+ \de l t a\rho ) && q _j \\ &&\\
- q ^ i+( \rho+ P) B^i& & (P+ \delta P)\delta^i_ j +\pi^i_j
\end{ arr ay}\ r ig ht] \, ,
\l ab e l{p e r t_T}$$ where $q _i$ is the | genceless condition_implies that_only two components are_independent.
A similar_procedure_applies to_the_symmetric tensor $h_{ij}$,_which can be_decomposed as h\_[ij]{}=2C\_[ij]{}+2 \_i\_j_E+2\_[(i]{}E\_[j)]{} +\_[ij]{}, \[hij\]_with_$\overline{E}^{ij}$ transverse and traceless (TT), i.e. $\nabla_i\overline{E}^{ij}=0$ (transverse) and $\overline{E}^{ij}\d_{ij}=0$ (traceless), and $E_i$ transverse._The_parentheses around_the_indices_denote symmetrization, namely $2\nabla_{(i}E_{j)}= \nabla_{i}E_{j}_+ \nabla_{j}E_{i}$. We have thus_defined two_scalar modes, $C$ and $E$, two vector modes,_$E_i$,_and two tensor_modes, $\E_{ij}$.
In the following, we will be mainly interested_in the metric with only [*scalar*]{}_perturbations, since scalar_modes_are_the most relevant for_cosmology and they can be treated_independently of the vector and tensor_modes. In matrix notation, the perturbed metric_will thus be of the form_$$\label{scalar_pert_metric}
g_{\mu\nu} =a^2 \left[
\begin{array}{ccc}
-(1+2A) & &_\nabla_iB \\_&& \\
\nabla_j B & &_ \left\{ (1+2C)_\d_{ij} +_2\nabla_i\nabla_jE \right\}
\end{array}
\right] \,.$$_After the description of the perturbed_geometry, we turn_to the perturbations of the matter_in_the next subsection.
Perturbations_of_the_matter
---------------------------
Quite generally,_the perturbed energy-momentum_tensor_can be_written_in the form $$\label{Tmunu}
T^\mu_\nu =
\left[
\begin{array}{ccc}
-(\rho+\delta\rho)_&_& q_j \\ &&\\
-q^i+(\rho+P)B^i & & (P+\delta_P)\delta^i_j + \pi^i_j
\end{array}
\right] \,,
\label{pert_T}$$_where_$q_i$ is the |
multiplying the second row by the scalar $x^{N-1}+x^{N-2}+\cdots+x+1\in R$ yields the zero element $(0,0) \in R^2$.
Acknowledgment {#acknowledgment.unnumbered}
==============
The authors gratefully acknowledge contributions to the proofs by Pascal Vontobel and Lance Small. The authors would also like to thank Dariush Divsalar for useful discussions.
[^1]: This work was presented in part at the IEEE International Symposium on Information Theory, Austin, Texas, June 2010.
[^2]: The authors are with the Department of Electrical and Computer Engineering, University of California, San Diego (UCSD), La Jolla, CA 92093 USA (e-mail: [email protected], [email protected]).
[^3]: This work was supported in part by the Center for Magnetic Recording Research at UCSD and by the National Science Foundation (NSF) under Grant CCF-0829865 and Grant CCF-1116739.
[^4]: The proof holds no matter which row is chosen for removal; however, the row removal process terminates more rapidly if a dependent row is chosen.
---
abstract: 'The large-scale circulation of planetary atmospheres like that of the Earth is traditionally thought of in a dynamical framework. Here, we apply the statistical mechanics theory of turbulent flows to a simplified model of the global atmosphere, the *quasi-geostrophic* model, leading to non-trivial equilibria. Depending on a few global parameters, the structure of the flow may be either a solid-body rotation (zonal flow) or a dipole. A second order phase transition occurs between these two phases, with associated spontaneous symmetry-breaking in the dipole phase. This model allows us to go beyond the general theory of marginal ensemble equivalence through the notion of Goldstone modes.'
author:
- 'C. Herbert'
- 'B. Dubrulle'
- 'P.H. Chavanis'
- 'D. Paillard'
title: |
Phase transitions and marginal ensemble equivalence for freely evolving flows\
on a rotating sphere
---
Introduction
============
It is well known that two-dimensional (2D) turbulent flows develop large-scale coherent structures [@McWilliams1984], and widely believed that the existence of these structures is related to the existence of | multiplying the second row by the scalar $ x^{N-1}+x^{N-2}+\cdots+x+1\in R$ yields the zero element $ (0,0) \in R^2$.
Acknowledgment { # acknowledgment.unnumbered }
= = = = = = = = = = = = = =
The writer appreciatively acknowledge contributions to the proofs by Pascal Vontobel and Lance Small. The generator would also like to thank Dariush Divsalar for utilitarian discussions.
[ ^1 ]: This work was presented in region at the IEEE International Symposium on Information Theory, Austin, Texas, June 2010.
[ ^2 ]: The authors are with the Department of Electrical and Computer Engineering, University of California, San Diego (UCSD), La Jolla, CA 92093 USA (e - chain mail: [email protected], [email protected]).
[ ^3 ]: This work was supported in part by the Center for Magnetic Recording Research at UCSD and by the National Science Foundation (NSF) under Grant CCF-0829865 and Grant CCF-1116739.
[ ^4 ]: The proof hold no matter which row is chosen for removal; however, the row removal procedure terminates more rapidly if a dependent rowing is chosen.
---
abstract:' The large - scale circulation of planetal atmospheres like that of the Earth is traditionally thought of in a dynamical framework. Here, we apply the statistical automobile mechanic theory of turbulent flows to a simplified model of the global atmosphere, the * quasi - geostrophic * model, leading to non - trivial equilibria. Depending on a few global parameters, the structure of the flow may be either a solid - body rotation (zonal stream) or a dipole. A second order phase transition occur between these two phases, with associated spontaneous symmetry - breaking in the dipole phase. This exemplar allows us to go beyond the general theory of bare ensemble equivalence through the notion of Goldstone modes.'
author:
-' C. Herbert'
-' B. Dubrulle'
-' P.H. Chavanis'
-' D. Paillard'
title: |
Phase transitions and borderline ensemble equivalence for freely evolving flows\
on a rotating sphere
---
initiation
= = = = = = = = = = = =
It is well known that two - dimensional (2D) turbulent stream develop large - scale coherent structures [ @McWilliams1984 ], and widely believe that the existence of these structures is related to the existence of | muptiplying the second row by the scalar $r^{B-1}+x^{N-2}+\cdovs+x+1\in R$ yields ghe zero element $(0,0) \in R^2$.
Acknowpeegmenu {#acknowledgment.unnjmbered}
==============
Thv authors grauefully acknowledjs contrlyutiohd to vhe proofs by Psscal Vontmbel and Lance Soapl. The authors would also like to trank Datiksh Divsalar fjr uxqful discussions.
[^1]: This work was presentsd in pert at the IEEE International Symposium oj Invormation Theory, Akstin, Texas, Junq 2010.
[^2]: The authorr are with the Departmgnt of Electrical and Computer Evginezring, Univetsjtj of Califoriia, Sag Diego (UCSD), La Jolld, CA 92093 UXA (e-mail: [email protected], [email protected]).
[^3]: This xork was supported ig part by tke Center for Magnetix Eecorging Resdqrcf au URSD and bj tie National Science Foyndation (NSF) under Brwbt CCF-0829865 and Grznt CCS-1116739.
[^4]: Ehe proof holds no matter which row is bhossn for removal; however, rhe row removal proceds terminwtes more rapidly if a dependent row is chosen.
---
abvtracv: 'Ghe lwfte-dcale circulation of planetary atmospheres lihs uhan of the Earth is traditionaklj yrought of in x dynamiczl framework. Here, ae appli the wtatisticwl mrchanics theory of turbulenr flows to a wimplified model oy the global atmpsphete, the *quasi-geostrophic* modsl, leading ho non-tribkal equilibria. Ddpekditg on a few global parametqrs, the svructbre of tfe fkow mar be eitheg a solid-body rotation (zlnal ylow) mr a dipolf. A second order phase transition occurs betwgen thvse two pkases, eith associatqd spontaneous symmetxy-breaying in thv dipole 'hase. This mjdel allows uv to go beyonv the genqral theiry of oxrginal ensembke equivalence throygh the notion of Noldsjohe modes.'
author:
- 'E. Yerbert'
- 'B. Dubrukle'
- 'P.H. Cjatanis'
- 'G. Paillard'
tidle: |
Lhase transiunonw and marginal ensemble exuivzlence for freely rvjlving foows\
jn a rotating sphere
---
Introductioj
============
It iv wxll knpwn that two-dimensional (2D) turbuleht flows fevclop large-scaje ciherent struetures [@McWilliams1984], and widely believed thet the existence of thgse structures is relcttd to the exixtencq of | multiplying the second row by the scalar yields zero element \in R^2$. Acknowledgment acknowledge to the proofs Pascal Vontobel and Small. The authors would also like thank Dariush Divsalar for useful discussions. [^1]: This work was presented in part the IEEE International Symposium on Information Theory, Austin, Texas, June 2010. [^2]: The are the of and Computer Engineering, University of California, San Diego (UCSD), La Jolla, CA 92093 USA (e-mail: [email protected], [^3]: This work was supported in part by Center for Magnetic Recording at UCSD and by the Science (NSF) under CCF-0829865 Grant [^4]: The proof no matter which row is chosen for removal; however, the row removal process terminates more rapidly if dependent row --- abstract: large-scale of atmospheres like that Earth is traditionally thought of in Here, we apply the statistical mechanics theory of flows to simplified model of the global atmosphere, *quasi-geostrophic* model, leading to non-trivial equilibria. Depending on few global parameters, the structure of the flow may be either a solid-body rotation (zonal a dipole. A second phase transition occurs these phases, associated symmetry-breaking in dipole phase. This model allows us to go beyond the general of marginal ensemble equivalence through the notion of Goldstone modes.' 'C. - 'B. Dubrulle' 'P.H. Chavanis' - 'D. title: Phase transitions and marginal for evolving rotating --- ============ It is well that two-dimensional (2D) turbulent flows large-scale coherent structures [@McWilliams1984], existence of these structures is related to the of | multiplying the second row by The scalar $x^{n-1}+x^{N-2}+\cdOts+X+1\in r$ yIeldS the Zero element $(0,0) \in R^2$.
aCknoWledgment {#acknowledgmenT.unnuMbERed}
==============
THE aUthorS gratefULlY ACknOwLeDge CoNTrIbutiOns To the prOofs by PascAl VOnTobel and LancE smAll. The authOrs Would also likE to Thank DArIusH divsaLar For usEful diSCussioNs.
[^1]: This worK wAS preseNTed in paRT At The IeEE International SYMpOSium on InformatIon TheOrY, auSTIn, TExaS, June 2010.
[^2]: The auThOrs arE With the dEpARTMenT Of Electrical aNd Computer ENGinEering, unIveRSity of califOrNIa, SAn Diego (UCSD), la JoLla, CA 92093 USA (e-Mail: [email protected], [email protected]).
[^3]: thiS worK WaS sUppOrTEd iN PaRt bY The center foR MAgNetic recoRDING ResEarCh at uCSD aNd by the NationAl SCienCE FoUndatIon (NSf) undEr grant cCF-0829865 and grant cCf-1116739.
[^4]: The proof holds nO matTer which rOw iS cHosEn For reMOval; hoWevEr, tHe row reMoval prOCesS tERMInAtes more rapidly if a DePENdEnt row is Chosen.
---
ABsTrACt: 'The larGe-ScaLe ciRCUlatiOn of PLaNetary atMospheREs LiKe that oF tHe EartH iS trAdiTionaLLy thOught oF in a dynaMical FRamework. Here, we APply the statisTIcAL MeCHaniCs tHeory of turbUlenT FlowS to a SImPliFIed moDel of ThE GlOBal atmosphere, the *quaSi-GeostrOphic* Model, leading tO non-triviaL EQUilibria. depeNDiNG on a few global pArameTers, the strUCture of tHe floW may be eiTher a soliD-BOdy rotatIon (ZonAl fLow) OR A dIpole. A second oRDEr phAsE transiTioN occurs BetWeeN thEse TwO phases, wiTh associAtEd SpOnTanEous sYMmetry-brEaKinG iN thE dipoLE phase. this mOdel AlLoWS us To go beyONd THE genErAl TheoRy oF mArginAl enSEmbLe equivAlence thrOugH The nOtIoN of GoldStone modes.'
autHoR:
- 'C. Herbert'
- 'B. duBruLle'
- 'P.H. CHAVanis'
- 'D. PaIllard'
title: |
Phase transitIOns and mArgInal eNsemBle equivaLenCe for fReeLY evolvIng floWs\
on a RoTatING spheRE
---
inTroDuCtion
============
It is wELL knOwn thAt Two-dImensioNal (2D) turbulent flows DEveLop large-scale CohErenT STrUctUReS [@mcWIlLIamS1984], ANd widely believeD that the exIsTEnCe of these sTRucTuRes is reLated to The exIStence oF | multiplying the second ro w by the s calar $x ^{N -1 }+x^ {N-2 }+\cdots+x+1\i n R$yields the zero elemen t $(0 ,0 ) \in R^ 2$.
Acknowl e dg m e nt{# ac kno wl e dg ment. unn umbered }
======== === == =
The autho r sgratefully ac knowledge co ntr ibutio ns to the p roo fs by Pasca l Vonto bel and L an c e Smal l . The a u t ho rs w ould also like to th a nk Dariush Div salarfo r u s e ful di scussions.
[^1]: This wo r kw a s pr e sented in par t at the IE E E I nterna ti ona l Sympo siumon Inf ormation Th eory , Austin, Texas , June 2 0 10.
[^ 2]: Th e a uth orsa re w ith t h e D e pa rtm e ntof Elect ri ca l and Com p u t e r En gin eeri ng, U niversity ofCal ifor n ia, SanDiego (UC SD ), La Jolla , CA92 093 USA (e-mail : bu tler@ieee .or g, ps ie gel@u c sd.edu ).
[^ 3]: Thi s workw assu p p o rt ed in part by theCe n t er for Mag neticR ec or d ing Rese ar chat U C S D and byt he Nationa l Scie n ce F oundati on (NSF) u nde r G rantC CF-0 829865 and Gra nt CC F -1116739.
[^4 ] : The proof h o ld s no matt erwhich row i s ch o senforr em ova l ; how ever, t h er ow removal processte rminat es mo re rapidly if a depende n t row is c hose n .---
abstract:'Thelarge-scal e circula tionof plane tary atmo s p heres li ketha t o f t h e E arth is tradi t i onal ly though t o f in adyn ami cal fr am ework. He re, we a pp ly t he st atist i cal mech an ics t heo ry of turbul ent f lows t oa si mplifie d m o d el o fth e gl oba latmos pher e , t he *qua si-geostr oph i c* m od el , leadi ng to non-tri vi al equilib ri a.Depend i n g on a f ew global parameters, t h e struc tur e oftheflow maybeeither as olid-b ody ro tatio n(zo n a l flo w ) o r a d ipole. A s e c ond orde rphas e trans ition occurs betwe e n t hese two phas es, wit h as soc i at e d s po n tan e o us symmetry-bre aking in t he di pole phase . Th is modelallowsus to go beyo nd the ge neral the or y of m arg inal ensem ble equi valence t h rough th e not ion of Go ld sto ne mo des.'a uth or:
- 'C. H er bert'- 'B. D ubrulle'
- 'P.H. Chavanis'
- 'D . Pail lard'
ti tle: |
P h ase transiti onsand margin alens emble eq u ivale ncef or fr e ely e volv i ng flows\ o n arotating sp h e r e
- --
I ntr o ductio n
== ==========
It is well known tha t tw o - dim ens i onal ( 2D) turbulentflo ws d evelop l ar ge-scale co herent s tr u cture s [@Mc Willia ms1984] , an d widel y be lie ved thatthe e x istence o ft hese s truc tu res is relat e d to t he existence of | multiplying_the second_row by the scalar_$x^{N-1}+x^{N-2}+\cdots+x+1\in R$_yields_the zero_element_$(0,0) \in R^2$.
Acknowledgment_{#acknowledgment.unnumbered}
==============
The authors gratefully_acknowledge contributions to the_proofs by Pascal_Vontobel_and Lance Small. The authors would also like to thank Dariush Divsalar for useful_discussions.
[^1]:_This work_was_presented_in part at the IEEE_International Symposium on Information Theory,_Austin, Texas,_June 2010.
[^2]: The authors are with the Department_of_Electrical and Computer_Engineering, University of California, San Diego (UCSD), La Jolla,_CA 92093 USA (e-mail: [email protected], [email protected]).
[^3]:_This work was_supported_in_part by the Center_for Magnetic Recording Research at UCSD_and by the National Science Foundation_(NSF) under Grant CCF-0829865 and Grant CCF-1116739.
[^4]:_The proof holds no matter which_row is chosen for removal;_however, the_row removal process terminates more_rapidly if a_dependent row_is chosen.
---
abstract:_'The large-scale circulation of planetary atmospheres_like that of_the Earth is traditionally thought of_in_a dynamical framework._Here,_we_apply the_statistical mechanics theory_of_turbulent flows_to_a simplified model of the global_atmosphere,_the *quasi-geostrophic* model, leading to non-trivial equilibria._Depending on a few_global_parameters, the structure of_the flow may be either_a solid-body rotation (zonal flow) or_a dipole._A second_order phase transition occurs between these two phases, with associated spontaneous_symmetry-breaking in the dipole phase. This_model allows us to_go beyond_the_general theory of_marginal_ensemble equivalence_through the notion of Goldstone modes.'
author:
- 'C._Herbert'
- 'B._Dubrulle'
- 'P.H. Chavanis'
- 'D. Paillard'
title: |
_ Phase_transitions_and marginal ensemble equivalence for freely_evolving flows\
on_a rotating sphere
---
Introduction
============
It is well_known_that_two-dimensional (2D) turbulent flows develop_large-scale coherent structures [@McWilliams1984], and widely_believed that the_existence of these structures is related to_the_existence of |
likely to have come from EAS and to reduce to about 20 Hz the rate of events to be sent to the central station. All ToT triggers are directly promoted T2 whereas T1 threshold triggers are requested to pass a higher threshold of 3.2 $I_{VEM}^{est}$ in coincidence for 3 PMTs. Only T2 triggers are used for the definition of a T3.\
The probability for a station to pass the trigger requirements strongly depends on the integrated signal. We define this probability, P(S), as the ratio of stations that trigger divided by the number of stations for a given integrated signal. We have measured this probability directly from the data by two different methods. The first method is based on the existence of two pairs of detectors separated by 11 m from each other. The double sampling of signals in near locations provides a way to estimate the number of signals that did not cause a trigger. This method allows direct comparison of P(s) for individual stations. In figure \[fig1\] we show that the two pairs of stations used in this study have the same P(s) within uncertainties. The second method uses this result and assumes that a single P(s) can describe the behavior of all the stations in the surface array. For each event a Lateral Distribution Function (LDF) is fitted to all the stations that have signal. Here the LDF used is a parabola on a log-log scale. The systematic due to the use of a LDF form different from the one described in this conference is of a few percent, within the statistical uncertainties of the method [@ba05]. The LDF is assumed to be cylindrically symmetric so for each event there are regions of constant signal. In each event the stations that did and did not trigger in a given constant signal region can be identified and P(s) computed.
![\[fig1\] The points represent the two pairs of stations from method one with statistical error bars. For method 1 all showers with $\theta$ less then 60 degrees and S(1000) greater than 2 VEM have been used. The dotted line is the fit from method 2 using the same large theta and S(1000) bins. The functional form of the fit is $P(S) = s^{N}/(s^{N}+ s_{50\%}^{N}).$](ICRC_lhenry_fig1.eps){width="60.00000%"}
The agreement between both methods is shown in figure 1. Method one | likely to have come from EAS and to reduce to about 20 Hz the pace of event to be sent to the central place. All ToT trigger are directly promoted T2 whereas T1 doorway triggers are requested to evanesce a higher threshold of 3.2 $ I_{VEM}^{est}$ in concurrence for 3 PMTs. Only T2 triggers are used for the definition of a T3.\
The probability for a station to guide the trigger requirements strongly depends on the integrated signal. We specify this probability, P(S), as the ratio of stations that trigger divided by the issue of stations for a given integrated signal. We have measured this probability immediately from the data by two unlike methods. The first method is based on the universe of two pairs of detectors separated by 11 m from each other. The bivalent sample distribution of signal in near locations provides a way to estimate the number of signal that did not induce a gun trigger. This method allows direct comparison of P(s) for individual station. In figure \[fig1\ ] we show that the two pairs of place use in this study have the like P(s) within uncertainties. The second method uses this result and assumes that a single P(s) can trace the behavior of all the stations in the surface array. For each event a Lateral Distribution Function (LDF) is meet to all the station that have signal. Here the LDF used is a parabola on a log - log scale. The systematic due to the use of a LDF form unlike from the one described in this league is of a few percent, within the statistical uncertainties of the method [ @ba05 ]. The LDF is assumed to be cylindrically symmetric then for each event there are region of constant signal. In each consequence the stations that did and did not trigger in a given ceaseless signal region can be identified and P(s) computed.
! [ \[fig1\ ] The points represent the two pair of stations from method one with statistical error bars. For method 1 all showers with $ \theta$ less then 60 degrees and S(1000) greater than 2 VEM have been used. The dashed line is the fit from method acting 2 using the same large theta and S(1000) bin. The functional form of the fit is $ P(S) = s^{N}/(s^{N}+ s_{50\%}^{N}).$](ICRC_lhenry_fig1.eps){width="60.00000% " }
The agreement between both methods is shown in figure 1. Method one | linely to have come from EXS and to reducg ro abont 20 Hz fhe rate of events to be sent to the cwntrao station. All ToT trigeers are firectly pronited T2 whecsas T1 tmxeshomf trngjers are requesjed to pass d higher thresvoud of 3.2 $I_{VEM}^{est}$ in coincidence for 3 PMEs. Only T2 triggers are osed gjr tgv befinition of a T3.\
The probabilitg for a station to psss the trigger requiremenhs shrongly depends on the integrqted wignal. We dewine this krmbability, K(S), as the ratio of stations that trigyer divided bt tjg number of wtatijns for a giyvn intecrated xignal. We have meesurwd this probability dmrectly from the datw by two giyferent methods. The furwt mejhod hs bxwed on tie sxistejce of two pajrs of detextors separated by 11 m drom each othsr. The djuble sampling of signals in near locatpons provides a way to estinate the number of sihnals thae did not cause a trigger. This method allows direwt cojoarnwon ow P(d) for individual stations. In figure \[fig1\] we shjs uhan the two pairs on stations used in tjix study have tfe samz P(a) within uncertainhies. Thg secobd method usex this result and assumes tyat a single P(s) can describe thz behavior oy all jhe ststions in the surface axray. Fkr each evejt a Latedxl Distribution Wunbtiot (LDF) is fitted to all the stations thac have skgnak. Here the LDF uded is a parabola on a llg-log swale. The sjstematic due to the use of a LVH form differgnt frmm the oue desgribed in this sonference is pf a fec percdnt, within the stetistical unsertainties ox the method [@ua05]. The LDS is assymed to ce cylindricalky symmetgie so for wach event there ave reejons of constanc wignal. In each rvevt ehv svatiogv that did atd dkd vpt trkgger in a nivdn cpnstant signal regiot cah be identified anc I(s) computgd.
![\[fig1\] The points reprexent the two pairs of svationr frok mgthod one with statistical errod bars. Fog mcthod 1 all shjwerw with $\theta$ less then 60 degrees and S(1000) greater than 2 VEM have been usgd. The dotted line is uhe fit from kethoq 2 using dhe same large theta and S(1000) bins. The fmnctional form of the rit is $P(S) = d^{N}/(s^{N}+ s_{50\%}^{N}).$](ICRC_lhenry_fig1.eps){width="60.00000%"}
The agreement between both methods is shown ib figuce 1. Method ons | likely to have come from EAS and to 20 Hz rate of events central All ToT triggers directly promoted T2 T1 threshold triggers are requested to a higher threshold of 3.2 $I_{VEM}^{est}$ in coincidence for 3 PMTs. Only T2 are used for the definition of a T3.\ The probability for a station pass trigger strongly on the integrated signal. We define this probability, P(S), as the ratio of stations that trigger by the number of stations for a given signal. We have measured probability directly from the data two methods. The method based the existence of pairs of detectors separated by 11 m from each other. The double sampling of signals in near provides a estimate the of that not cause a method allows direct comparison of P(s) In figure \[fig1\] we show that the two of stations in this study have the same within uncertainties. The second method uses this result assumes that a single P(s) can describe the behavior of all the stations in the For each event a Distribution Function (LDF) fitted all stations have signal. the LDF used is a parabola on a log-log scale. The due to the use of a LDF form different from described this conference is a few percent, within statistical of the method [@ba05]. is to so each there are regions of signal. In each event the that did and did constant signal region can be identified and P(s) ![\[fig1\] The points represent the two pairs stations from method one with statistical error bars. For method 1 all with $\theta$ 60 degrees and S(1000) greater than 2 VEM been used. The dotted is the fit from method 2 using the same theta S(1000) bins. functional form of fit is $P(S) s^{N}/(s^{N}+ s_{50\%}^{N}).$](ICRC_lhenry_fig1.eps){width="60.00000%"} The both methods shown figure | likely to have come from EAS anD to reduce tO abouT 20 Hz The RaTe of EvenTs to be sent to thE CentRal station. All ToT triggeRs are DiREctlY PrOmoteD T2 whereAS T1 THResHoLd TriGgERs Are reQueSted to pAss a higher ThrEsHold of 3.2 $I_{VEM}^{esT}$ In CoincidencE foR 3 PMTs. Only T2 trIggErs are UsEd fOR the dEfiNitioN of a T3.\
THE probaBility for A sTAtion tO Pass the TRIgGer rEquirements strongLY dEPends on the inteGrated SiGNaL. wE deFinE this probaBiLity, P(s), As the raTIo OF STatIOns that triggeR divided by tHE nuMber of StAtiONs for a Given InTEgrAted signal. WE havE measured This prOBabilitY DirectlY from tHe dAta By twO DiFfEreNt MEthODs. the FIrsT method iS bAsEd on tHe exISTENce oF twO paiRs of dEtectors separAteD by 11 m FRom Each oTher. THe doUbLe samPling oF signAlS in near locationS proVides a way To eStImaTe The nuMBer of sIgnAls That did Not causE A trIgGER. thIs method allows direCt COMpArison of p(s) for iNDiViDUal statiOnS. In FiguRE \[Fig1\] we Show THaT the two pAirs of STaTiOns used In This stUdY haVe tHe samE p(s) wiThin unCertaintIes. ThE Second method usES this result anD AsSUMeS That A siNgle P(s) can deScriBE the BehaVIoR of ALl the StatiOnS In THe surface array. For eaCh Event a laterAl DistributioN Function (Ldf) IS fitted tO all THe STations that havE signAl. Here the Ldf used is a ParabOla on a loG-log scale. tHE systemaTic Due To tHe uSE Of A LDF form diffeRENt frOm The one dEscRibed in ThiS coNfeRenCe Is of a few pErcent, wiThIn ThE sTatIsticAL uncertaInTieS oF thE methOD [@ba05]. The lDF is AssuMeD tO Be cYlindriCAlLY SymmEtRiC so fOr eAcH evenT theRE arE regionS of constaNt sIGnal. in EaCh event The stations thAt Did and did nOt TriGger in A GIven consTant signal region can be idENtified And p(s) comPuteD.
![\[fig1\] The poIntS repreSenT The two Pairs oF statIoNs fROM methOD OnE wiTh StatisticaL ERroR bars. foR metHod 1 all sHowers with $\theta$ lesS TheN 60 degrees and S(1000) gReaTer tHAN 2 VeM hAVe BEen UsED. ThE DOtted line is the fIt from methOd 2 USiNg the same lARge ThEta and S(1000) Bins. The FunctIOnal forM of the fit Is $P(S) = s^{N}/(s^{N}+ s_{50\%}^{n}).$](IcRC_lHENry_Fig1.eps){widtH="60.00000%"}
The agreEment betwEEn botH MeThods Is sHown in FiGurE 1. MethOd one | likely to have come fromEAS and to redu cetoab out20 H z the rate ofe vent s to be sent to the ce ntral s t atio n .All T oT trig g er s are d ir ect ly pr omote d T 2 where as T1 thre sho ld triggers ar e r equested t o p ass a higher th reshol dof3 .2 $I _{V EM}^{ est}$i n coin cidence f or 3 PMTs . Only T 2 tr igge rs are used for t h ed efinition of a T3.\Th e p r o bab ili ty for a s ta tiont o passt he t r igg e r requirement s stronglyd epe nds on t hei ntegra ted s ig n al. We definethis probabil ity, P ( S), ast he rati o of s tat ion s th a ttr igg er div i de d b y th e number o fstati onsf o r a gi ven int egrat ed signal. We ha ve m e asu red t his p roba bi litydirect ly fr om the data by tw o di fferent m eth od s.Th e fir s t meth odisbased o n the e x ist en c e of two pairs of dete ct o r sseparate d by 1 1 m f r om eachot her . Th e doubl e sa m pl ing of s ignals in n ear loc at ions p ro vid esa way to e stimat e the nu mbero f signals that did not cause at r ig g er.Thi s method al lows dire ct c o mp ari s on of P(s) f o ri ndividual stations. I n figu re \[ fig1\] we sho w that the t w o pairsof s t at i ons used in th is st udy have t h e same P (s) w ithin un certainti e s . The se con d m eth odu s es this resulta n d as su mes tha t a single P( s)can de sc ribe thebehavior o fal lthe stat i ons in t he su rf ace arra y . Foreacheven taL ate ral Dis t ri b u tion F un ctio n ( LD F) is fit t edto allthe stati ons that h av e signa l. Here the L DF used is a p ara bola o n a log-lo g scale. The systematic due tothe useof a LDF form di fferen t f r om the one d escri be d i n thisc o nf ere nc e is of af e w p ercen t, wit hin the statistical uncer t ain ties of the m eth od [ @ b a0 5]. Th e LD Fi s a s s umed to be cyli ndricallysy m me tric so fo r ea ch eventthere a re re g ions of constant signal.In eac h eve nt the sta tions th at did an d didn ot trig ger in agi ven cons tant s i gna l reg ion ca nbe ide ntifi ed and P(s ) computed.
![\[fig1\] The p oints re present t het wopairs ofstat ions frommet hod onewit h stat isti c al er r or ba rs.F or method 1all s ho wers with $ \ t h eta $ les s t h en 60degr ees and S(1000) g r eater than 2 V EM h a v e b een used .The dotted lin e i st h e fit fr om method 2 u sing the s a me la rge th eta an d S(100 0 ) b i ns. Th e fu nct ional for m o ft he fitis $ P (S) =s^{N }/ (s^{N} + s_{5 0 \%}^ { N }).$](ICRC_lhenr y_fig 1 . eps){ w idt h="60 .0 0000%"}
The agreement between bo th met hods is s hown in f igure1.Me thod one | likely_to have_come from EAS and_to reduce_to_about 20_Hz_the rate of_events to be_sent to the central_station. All ToT_triggers_are directly promoted T2 whereas T1 threshold triggers are requested to pass a higher_threshold_of 3.2_$I_{VEM}^{est}$_in_coincidence for 3 PMTs. Only_T2 triggers are used for_the definition_of a T3.\
The probability for a station to_pass_the trigger requirements_strongly depends on the integrated signal. We define this_probability, P(S), as the ratio of_stations that trigger_divided_by_the number of stations_for a given integrated signal. We_have measured this probability directly from_the data by two different methods. The_first method is based on the_existence of two pairs of_detectors separated_by 11 m from each_other. The double_sampling of_signals in near_locations provides a way to estimate_the number of_signals that did not cause a_trigger._This method allows_direct_comparison_of P(s)_for individual stations._In_figure \[fig1\]_we_show that the two pairs of_stations_used in this study have the same_P(s) within uncertainties. The_second_method uses this result_and assumes that a single_P(s) can describe the behavior of_all the_stations in_the surface array. For each event a Lateral Distribution Function (LDF)_is fitted to all the stations_that have signal. Here_the LDF_used_is a parabola_on_a log-log_scale. The systematic due to the use_of a_LDF form different from the one_described in this conference_is_of a few percent, within the_statistical uncertainties of the method [@ba05]._The LDF is assumed to_be_cylindrically_symmetric so for each event_there are regions of constant signal._In each event_the stations that did and did not_trigger_in a given constant signal region_can_be identified and P(s) computed.
![\[fig1\] The_points_represent_the two pairs of stations_from method one with statistical error_bars. For method 1 all showers with $\theta$ less_then 60 degrees_and S(1000) greater than 2_VEM_have_been used. The dotted line is the fit from method_2 using_the same large_theta and S(1000) bins. The functional form of the fit_is $P(S) = s^{N}/(s^{N}+ s_{50\%}^{N}).$](ICRC_lhenry_fig1.eps){width="60.00000%"}
The agreement between_both methods is shown in figure 1. Method one |
in chemical engineering in 2013 from the University of New Hampshire, Durham, NH, USA. She received the M.S. degree in electrical and computer engineering in 2017 from Oregon State University, Corvallis, OR, USA, where she is currently working toward the Ph.D. degree in electrical and computer engineering.
She is currently a Graduate Fellow at Idaho National Laboratory, Idaho Falls, ID, USA. Her research interests included power system reliability, dynamic modeling, and rural electrification.
[Eduardo Cotilla-Sanchez]{} (S’08-M-12-SM-19) received the M.S. and Ph.D. degrees in electrical engineering from the University of Vermont, Burlington, VT, USA, in 2009 and 2012, respectively.
He is currently an Associate Professor in the School of Electrical Engineering and Computer Science, Oregon State University, Corvallis, OR, USA. His primary field of research is electrical infrastructure resilience and protection, in particular, the study of cascading outages.
Prof. Cotilla-Sanchez is the Vice-Chair of the IEEE Working Group on Cascading Failures and President of the Society of Hispanic Professional Engineers Oregon Chapter.
[^1]: This material is based upon work supported by the National Science Foundation Graduate Research Fellowship under Grant No. 1314109-DGE.
[^2]: The authors are with the School of Electrical Engineering and Computer Science, Oregon State University, Corvallis, OR 97331 USA e-mail: ([email protected]; [email protected]).
---
abstract: 'We study the suppression of the small-scale power spectrum due to the decay of charged matter to dark matter prior to recombination. Prior to decay, the charged particles couple to the photon-baryon fluid and participate in its acoustic oscillations. However, after decaying to neutral dark matter the photon-baryon fluid is coupled only gravitationally to the newly-created dark matter. This generically leads to suppression of power on length scales that enter the horizon prior to decay. For decay times of $\sim$$3.5$ years this leads to suppression of power on subgalactic scales, bringing the observed number of Galactic substructures in line with observation. Decay times of a few years are possible if the dark matter is purely gravitationally interacting, such as the gravitino in supersymmetric models | in chemical engineering in 2013 from the University of New Hampshire, Durham, NH, USA. She received the M.S. degree in electrical and calculator technology in 2017 from Oregon State University, Corvallis, OR, USA, where she is presently working toward the Ph.D. degree in electric and computer engineering.
She is currently a Graduate Fellow at Idaho National Laboratory, Idaho Falls, ID, USA. Her inquiry interest included ability system reliability, active modeling, and rural electrification.
[ Eduardo Cotilla - Sanchez ] { } (S’08 - M-12 - SM-19) received the M.S. and Ph.D. degrees in electric engineering from the University of Vermont, Burlington, VT, USA, in 2009 and 2012, respectively.
He is currently an Associate Professor in the School of Electrical Engineering and Computer Science, Oregon State University, Corvallis, OR, USA. His basal field of research is electric infrastructure resilience and protection, in particular, the study of cascade outages.
Prof. Cotilla - Sanchez is the Vice - Chair of the IEEE Working Group on Cascading Failures and President of the Society of Hispanic Professional Engineers Oregon Chapter.
[ ^1 ]: This fabric is based upon study supported by the National Science Foundation Graduate Research Fellowship under Grant No. 1314109 - DGE.
[ ^2 ]: The generator are with the School of Electrical Engineering and Computer Science, Oregon State University, Corvallis, OR 97331 USA e - mail: ([email protected]; [email protected]).
---
abstract:' We study the inhibition of the small - scale power spectrum due to the decay of charged matter to dark matter prior to recombination. Prior to decay, the charge particles couple to the photon - baryon fluid and participate in its acoustic oscillations. However, after decaying to neutral dark matter the photon - baryon fluid is coupled only gravitationally to the newly - created black matter. This generically leads to inhibition of exponent on distance scales that enter the horizon prior to decay. For decay time of $ \sim$$3.5 $ years this leads to suppression of power on subgalactic scale, bringing the observed number of Galactic substructures in line with observation. Decay time of a few years are possible if the dark matter is purely gravitationally interacting, such as the gravitino in supersymmetric models | in chemical engineering in 2013 from the Univgrwity oh New Hzmpshire, Durham, NH, USA. She received vhe N.S. detree in electrical and computer engineeeing un 2017 from Ocsgon State Unjyersicy, Corvallis, OR, OSA, where sha is currently wurning toward the Ph.D. degree in electwical amd computer engigeerpnd.
She ps currently a Graduate Fellow zt Idahm National Lanoratory, Idaho Falls, ID, USW. Heg research intereshs included powqe system relkability, djuamic modeljng, and rural electrification.
[Edjardo Cotilla-Sabcyez]{} (V’08-M-12-SM-19) receivxd the M.S. and Ph.D. dengees in electrocal engineerikg frmm rhe University of Verkont, Burlington, VT, USA, in 2009 dnb 2012, respectively.
He is xuerentny at Asricixte Pcofsssor ln vhe School kf Electricql Engineering and Voiiiter Science, Oregog Ftate University, Corvallis, OR, USA. His pgimady field of research is electrical infrastrufture resylience and protection, in particular, the study of cascedkng omtagdw.
Pgof. Cotilla-Sanchez is the Vice-Chair of the IEEQ Wprling Group on Gascading Failures ajd Kresident of tfe Socnstg of Hispanic Proffssionaj Enguneers Ortgon Vhapter.
[^1]: This material is bawed upon worh supported by the Uational Sciznce Fpundayion Graduate Research Yellowahip under Hrant No. 1314109-DFD.
[^2]: The authors ard wpth dhe School of Electrical Eggineerinj and Computdr Svience, Oregon Stwte Ukhversity, Corvallis, OR 97331 OSA e-mdil: ([email protected]; [email protected]).
---
abstract: 'Wg sdudj the sup'ressipn of the smajl-scale power xpectrui due to the debay of cherged matter to dark mattat prior to rerombinatijn. Peior to decxh, the charged larticles couple to the photon-baryon nluid znd participate in its acoustic oxciulaeilnx. Rmwever, after decxyivb to veutral darj matyer the photon-baryon flujd is coupled only gvavitatiobally to the newly-crested dark matter. Tjis gxnericelly lradf to suppression of power on lsngth scapes that enter ehe morieon prior tp decay. For decay times of $\sim$$3.5$ years thms leads to suppression of power on subgalccuic scales, brmnging the obsesved number of Galacric substructures in line with observafion. Dacay himes of a few years are possible if the dark matter is purely gravitatiobally mneeracting, shch ss thv gxavntino ig su'ecsymmetric models | in chemical engineering in 2013 from the New Durham, NH, She received the computer in 2017 from State University, Corvallis, USA, where she is currently working the Ph.D. degree in electrical and computer engineering. She is currently a Graduate at Idaho National Laboratory, Idaho Falls, ID, USA. Her research interests included power reliability, modeling, rural [Eduardo Cotilla-Sanchez]{} (S’08-M-12-SM-19) received the M.S. and Ph.D. degrees in electrical engineering from the University of Burlington, VT, USA, in 2009 and 2012, respectively. is currently an Associate in the School of Electrical and Science, Oregon University, OR, His primary field research is electrical infrastructure resilience and protection, in particular, the study of cascading outages. Prof. Cotilla-Sanchez is Vice-Chair of Working Group Cascading and of the Society Professional Engineers Oregon Chapter. [^1]: This upon work supported by the National Science Foundation Research Fellowship Grant No. 1314109-DGE. [^2]: The authors with the School of Electrical Engineering and Computer Oregon State University, Corvallis, OR 97331 USA e-mail: ([email protected]; [email protected]). --- abstract: 'We study the the small-scale power spectrum to the decay charged to matter to recombination. to decay, the charged particles couple to the photon-baryon fluid and in its acoustic oscillations. However, after decaying to neutral dark photon-baryon is coupled only to the newly-created dark This leads to suppression of length that prior decay. decay times of $\sim$$3.5$ this leads to suppression of on subgalactic scales, bringing substructures in line with observation. Decay times of few years are possible if the dark is purely gravitationally interacting, such as the gravitino in supersymmetric models | in chemical engineering in 2013 frOm the UniveRsity Of NEw HAmPshiRe, DuRham, NH, USA. She reCEiveD the M.S. degree in electricAl and CoMPuteR EnGineeRing in 2017 fROm oREgoN STaTe UNiVErSity, COrvAllis, OR, uSA, where shE is CuRrently workiNG tOward the Ph.d. deGree in electrIcaL and coMpUteR EnginEerIng.
ShE is curREntly a graduate FElLOw at IdAHo NatioNAL LAborAtory, Idaho Falls, ID, usA. hEr research inteRests iNcLUdED PowEr sYstem reliaBiLity, dYNamic moDElING, And RUral electrifiCation.
[EduarDO CoTilla-SAnCheZ]{} (s’08-M-12-SM-19) reCeiveD tHE M.S. And Ph.D. degreEs in ElectricaL enginEEring frOM the UniVersitY of verMont, bUrLiNgtOn, vt, USa, In 2009 And 2012, REspEctively.
he Is CurreNtly AN aSSociAte profEssor In the School of eleCtriCAl ENgineEring And COmPuter sciencE, OregOn state University, corvAllis, OR, USa. HiS pRimArY fielD Of reseArcH is ElectriCal infrAStrUcTURE rEsilience and protecTiON, In ParticulAr, the sTUdY oF CascadinG oUtaGes.
PROF. CotiLla-SANcHez is the vice-ChAIr Of The IEEE woRking GRoUp oN CaScadiNG FaiLures aNd PresidEnt of THe Society of HisPAnic ProfessioNAl eNGiNEers oreGon Chapter.
[^1]: THis mATeriAl is BAsEd uPOn worK suppOrTEd BY the National Science foUndatiOn GraDuate Research fellowship UNDEr Grant NO. 1314109-DGE.
[^2]: tHe AUthors are with tHe SchOol of ElectRIcal EngiNeeriNg and ComPuter ScieNCE, Oregon STatE UnIveRsiTY, coRvallis, OR 97331 USA e-MAIl: ([email protected]; [email protected]).
---
abstrAct: 'We stuDy ThE sUpPreSsion OF the smalL-sCalE pOweR specTRum due To the DecaY oF cHArgEd matteR To DARk maTtEr PrioR to ReCombiNatiON. PrIor to deCay, the chaRgeD PartIcLeS couple To the photon-baRyOn fluid and PaRtiCipate IN Its acousTic oscillations. However, aFTer decaYinG to neUtraL dark mattEr tHe photOn-bARyon flUid is cOupleD oNly GRAvitaTIOnAllY tO the newly-cREAteD dark MaTter. this genErically leads to supPResSion of power on LenGth sCALeS thAT eNTer ThE HorIZOn prior to decay. FOr decay timEs OF $\sIm$$3.5$ years thiS LeaDs To supprEssion oF poweR On subgaLactic scaLes, bringiNg The oBSErvEd number of galactic SubstructURes in LInE with ObsErvatiOn. decAy timEs of a fEW yeArs arE possiBlE if the Dark mAtTer is purEly gravitationally interActing, Such aS thE gravitinO in SUpeRsymmetriC modEls | in chemical engineering i n 2013 fro m the Un ive rs ityof N ew Hampshire,D urha m, NH, USA. She receiv ed th eM .S.d eg ree i n elect r ic a l an dco mpu te r e ngine eri ng in 2 017 from O reg on State Unive r si ty, Corval lis , OR, USA, w her e sheis cu r rentl y w orkin g towa r d thePh.D. deg re e in el e ctrical a nd com puter engineering .
S he is currentl y a Gr ad u at e Fel low at IdahoNa tiona l Labora t or y , Ida h o Falls, ID,USA. Her re s ear ch int er est s inclu ded p ow e r s ystem relia bili ty, dynam ic mod e ling, a n d rural elect rif ica tion .
[E dua rd o Co t il la- S anc hez]{} ( S’ 08 -M-12 -SM- 1 9 ) rece ive d th e M.S . and Ph.D. d egr eesi n e lectr icalengi ne ering fromthe U ni versity of Verm ont, Burlingt on, V T,US A, in 2009 a nd201 2, resp ectivel y .
He i s c urrently an Associ at e Pr ofessorin the Sc ho o l of Ele ct ric al E n g ineer inga nd Compute r Scie n ce ,OregonSt ate Un iv ers ity , Cor v alli s, OR, USA. Hi s pri m ary field of r e search is ele c tr i c al infr ast ructure res ilie n ce a nd p r ot ect i on, i n par ti c ul a r, the study of cas ca ding o utage s.
Prof. Cot illa-Sanch e z is the V ice- C ha i r of the IEEEWorki ng Group o n Cascadi ng Fa ilures a nd Presid e n t of the So cie tyofH i sp anic Professi o n al E ng ineersOre gon Cha pte r.
[^ 1]: T his mater ial is b as ed u po n w ork s u pportedby th eNat ional Scienc e Fou ndat io nG rad uate Re s ea r c h Fe ll ow ship un de r Gra nt N o . 1 314109- DGE.
[^2 ]:T he a ut ho rs arewith the Scho ol of Electr ic alEngine e r ing andComputer Science, Orego n StateUni versi ty,Corvallis , O R 9733 1 U S A e-ma il: (c icili op @or e g onsta t e .e du; e cs@oregons t a te. edu).
---abstrac t: 'We study the s u ppr ession of the sm all- s c al e p o we r sp ec t rum d ue to the decay of charge dm at ter to dar k ma tt er prio r to re combi n ation.Prior todecay, th echar g e d p articles c ouple to the phot o n-bar y on flui d a nd par ti cip ate i n itsa cou sticoscill at ions.Howev er , afterdecaying to neutral dar k matt er th e p hoton-bar yon flu id is cou pled only grav ita tio nally to the n ewly - cr eat e d dar k ma t ter. This ge ner i c al ly leads to s u ppr essio n o f power onlength scales tha t enter the hor izon p rio r t o dec ay . For decay ti mes o f $\sim$$3 .5 $ years thi s leadsto suppr ession of po wer ons u bg a lactic sca les , bringin g t he observe dnu m ber of Gal ac tic su bstruc t ures i n line with obse rvati o n . Dec a y t imesof a fewy ears are possi ble if thedark m atte r ispurelygr avitat ion al ly interac t ing, such as t he grav it inoinsupers ymme t r ic mo dels | in_chemical engineering_in 2013 from the_University of_New_Hampshire, Durham,_NH,_USA. She received_the M.S. degree_in electrical and computer_engineering in 2017_from_Oregon State University, Corvallis, OR, USA, where she is currently working toward the Ph.D._degree_in electrical_and_computer_engineering.
She is currently a Graduate_Fellow at Idaho National Laboratory,_Idaho Falls,_ID, USA. Her research interests included power system_reliability,_dynamic modeling, and_rural electrification.
[Eduardo Cotilla-Sanchez]{} (S’08-M-12-SM-19) received the M.S. and Ph.D. degrees in_electrical engineering from the University of_Vermont, Burlington, VT,_USA,_in_2009 and 2012, respectively.
He_is currently an Associate Professor in_the School of Electrical Engineering and_Computer Science, Oregon State University, Corvallis, OR,_USA. His primary field of research_is electrical infrastructure resilience and_protection, in_particular, the study of cascading_outages.
Prof. Cotilla-Sanchez is the_Vice-Chair of_the IEEE Working_Group on Cascading Failures and President_of the Society_of Hispanic Professional Engineers Oregon Chapter.
[^1]:_This_material is based_upon_work_supported by_the National Science_Foundation_Graduate Research_Fellowship_under Grant No. 1314109-DGE.
[^2]: The authors are_with_the School of Electrical Engineering and Computer_Science, Oregon State University,_Corvallis,_OR 97331 USA e-mail:_([email protected]; [email protected]).
---
abstract: 'We study_the suppression of the small-scale power_spectrum due_to the_decay of charged matter to dark matter prior to recombination. Prior_to decay, the charged particles couple_to the photon-baryon fluid_and participate_in_its acoustic oscillations._However,_after decaying_to neutral dark matter the photon-baryon fluid_is coupled_only gravitationally to the newly-created dark_matter. This generically leads_to_suppression of power on length scales_that enter the horizon prior to_decay. For decay times of_$\sim$$3.5$_years_this leads to suppression of_power on subgalactic scales, bringing the_observed number of_Galactic substructures in line with observation. Decay_times_of a few years are possible_if_the dark matter is purely gravitationally_interacting,_such_as the gravitino in supersymmetric_models |
= 2^{-18} \lambda^{-2}$.
\[lem:stability\] Let $\lambda {\geqslant}3$ and $n,k \in {\mathbb{N}}$, with $k {\geqslant}(\log n)^4$ and $k {\geqslant}2^{480} \lambda^{20}$. We have $$|\Lambda \setminus \Lambda^*| {\leqslant}\frac{n^2}{k} \cdot |{\mathcal{I}}| + \exp\bigg( - \frac{\delta k}{2^{10} \lambda} \bigg) {\lambda k / 2 \choose k}.$$
To prove Lemma \[lem:stability\], we will successively refine $\Lambda \setminus \Lambda^*$, at each step showing that some subset with a particular property is small. The first step in the proof of Lemma \[lem:stability\] is the following stability lemma, which is an almost immediate consequence of Theorem \[thm:M:stability\].
\[lem:stability:first\] Let $\lambda {\geqslant}3$ and $n,k \in {\mathbb{N}}$, with $k {\geqslant}(\log n)^4$ and $k {\geqslant}2^{480} \lambda^{4}$. There are at most $$\exp\bigg( - \frac{\delta k}{2^9 \lambda} \bigg) {\lambda k / 2 \choose k}$$ sets $A \in \Lambda$ such that $$| A \setminus P | > \delta k$$ for every arithmetic progression $P$ of size $\lambda k/2$.
Note first that if $k {\geqslant}(\log n)^4$ and $k {\geqslant}2^{480} \lambda^{4}$ then $\frac{k}{(\log n)^{3}} {\geqslant}k^{1/4} {\geqslant}2^{120} \lambda$. Therefore, applying Theorem \[thm:M:stability\] with $\gamma = 2^{-9} \lambda^{-1} \delta$, it follows that for all but at most $$\exp\bigg( - \frac{\delta k}{2^9\lambda} \bigg) {\lambda k / 2 \choose k}$$ sets $A \in \Lambda$, there exists $T \subset A$, with $|T| {\leqslant}(2^9 + 2^7 \lambda) \gamma k {\leqslant}\delta k$, such that $A \setminus T$ is contained in an arithmetic | = 2^{-18 } \lambda^{-2}$.
\[lem: stability\ ] Let $ \lambda { \geqslant}3 $ and $ n, k \in { \mathbb{N}}$, with $ k { \geqslant}(\log n)^4 $ and $ k { \geqslant}2^{480 } \lambda^{20}$. We have $ $ |\Lambda \setminus \Lambda^*| { \leqslant}\frac{n^2}{k } \cdot |{\mathcal{I}}| + \exp\bigg (- \frac{\delta k}{2^{10 } \lambda } \bigg) { \lambda k / 2 \choose k}.$$
To prove Lemma \[lem: stability\ ], we will successively polish $ \Lambda \setminus \Lambda^*$, at each gradation showing that some subset with a particular property is little. The first step in the validation of Lemma \[lem: stability\ ] is the following stability lemma, which is an almost immediate consequence of Theorem \[thm: M: stability\ ].
\[lem: constancy: first\ ] lease $ \lambda { \geqslant}3 $ and $ n, k \in { \mathbb{N}}$, with $ k { \geqslant}(\log n)^4 $ and $ k { \geqslant}2^{480 } \lambda^{4}$. There be at most $ $ \exp\bigg (- \frac{\delta k}{2 ^ 9 \lambda } \bigg) { \lambda k / 2 \choose k}$$ sets $ A \in \Lambda$ such that $ $ | A \setminus P | > \delta k$$ for every arithmetical progression $ P$ of size $ \lambda k/2$.
Note first that if $ k { \geqslant}(\log n)^4 $ and $ k { \geqslant}2^{480 } \lambda^{4}$ then $ \frac{k}{(\log n)^{3 } } { \geqslant}k^{1/4 } { \geqslant}2^{120 } \lambda$. Therefore, applying Theorem \[thm: M: stability\ ] with $ \gamma = 2^{-9 } \lambda^{-1 } \delta$, it succeed that for all but at most $ $ \exp\bigg (- \frac{\delta k}{2 ^ 9\lambda } \bigg) { \lambda k / 2 \choose k}$$ sets $ A \in \Lambda$, there exist $ T \subset A$, with $ |T| { \leqslant}(2 ^ 9 + 2 ^ 7 \lambda) \gamma k { \leqslant}\delta k$, such that $ A \setminus T$ is contained in an arithmetic | = 2^{-18} \pambda^{-2}$.
\[lem:stability\] Let $\lxmbda {\geqslant}3$ cbd $n,k \mn {\mathgb{N}}$, with $k {\geqslant}(\log n)^4$ and $k {\geqslent}2^{480} \oambdq^{20}$. We have $$|\Lambda \setmivus \Lambdw^*| {\leqslabt}\frec{n^2}{k} \cdot |{\mathcal{I}}| + \exp\blyg( - \fdwc{\denva k}{2^{10} \lambda} \bign) {\lambda k / 2 \choose k}.$$
To psoxe Lemma \[lem:stability\], we will successidely regije $\Lambda \setmynus \Jambsa^*$, at each step showing that some aubset xith a particulsr property is small. The flrst step in the proof of Lemma \[len:stafulity\] is the following stability lgmma, which is an almost immediatd conxequence od Rheltem \[thm:M:stabioity\].
\[lvm:stability:fivxt\] Let $\lambda {\geqslant}3$ and $k,k \in {\marhbb{N}}$, with $k {\geqslant}(\nog n)^4$ and $k {\geqslagt}2^{480} \lambda^{4}$. Tkere are at most $$\exp\bugt( - \frdc{\denta y}{2^9 \laobdz} \uigf) {\lamhda k / 2 \chooss k}$$ sets $A \un \Lambda$ such that $$| W \setminus P | > \delta k$$ for every arithmetic progression $P$ of vizs $\lambda k/2$.
Note first thqt if $k {\geqslant}(\log n)^4$ and $k {\gezslant}2^{480} \lambda^{4}$ then $\frac{k}{(\log n)^{3}} {\geqslant}k^{1/4} {\geqslant}2^{120} \nambde$. Ghexcnore, qpolying Theorem \[thm:M:stability\] with $\gamma = 2^{-9} \lambqz^{-1} \cekta$, it follows that for akl bij at most $$\exp\bkgg( - \fxzc{\selta k}{2^9\lambda} \bigg) {\lambda k / 2 \xhoose k}$$ fets $A \in \Lambda$, there exists $T \subset A$, winh $|T| {\leqslant}(2^9 + 2^7 \lambda) \gamma k {\ledslamt}\delya k$, such that $A \setminbs T$ ia contained in an arjghmetic | = 2^{-18} \lambda^{-2}$. \[lem:stability\] Let $\lambda {\geqslant}3$ \in with $k n)^4$ and $k \setminus {\leqslant}\frac{n^2}{k} \cdot |{\mathcal{I}}| \exp\bigg( - \frac{\delta \lambda} \bigg) {\lambda k / 2 k}.$$ To prove Lemma \[lem:stability\], we will successively refine $\Lambda \setminus \Lambda^*$, at step showing that some subset with a particular property is small. The first in proof Lemma is the following stability lemma, which is an almost immediate consequence of Theorem \[thm:M:stability\]. \[lem:stability:first\] Let {\geqslant}3$ and $n,k \in {\mathbb{N}}$, with $k {\geqslant}(\log and $k {\geqslant}2^{480} \lambda^{4}$. are at most $$\exp\bigg( - k}{2^9 \bigg) {\lambda / \choose sets $A \in such that $$| A \setminus P | > \delta k$$ for every arithmetic progression $P$ of size k/2$. Note if $k n)^4$ $k \lambda^{4}$ then $\frac{k}{(\log {\geqslant}2^{120} \lambda$. Therefore, applying Theorem \[thm:M:stability\] 2^{-9} \lambda^{-1} \delta$, it follows that for all at most - \frac{\delta k}{2^9\lambda} \bigg) {\lambda k 2 \choose k}$$ sets $A \in \Lambda$, there $T \subset A$, with $|T| {\leqslant}(2^9 + 2^7 \lambda) \gamma k {\leqslant}\delta k$, such that T$ is contained in arithmetic | = 2^{-18} \lambda^{-2}$.
\[lem:stability\] Let $\lambDa {\geqslant}3$ And $n,k \In {\mAthBb{n}}$, witH $k {\geQslant}(\log n)^4$ and $k {\GEqslAnt}2^{480} \lambda^{20}$. We have $$|\Lambda \sEtminUs \lAmbdA^*| {\LeQslanT}\frac{n^2}{k} \CDoT |{\MAthCaL{I}}| + \Exp\BiGG( - \fRac{\deLta K}{2^{10} \lambda} \Bigg) {\lambda K / 2 \chOoSe k}.$$
To prove LeMMa \[Lem:stabiliTy\], wE will successIveLy refiNe $\lamBDa \setMinUs \LamBda^*$, at eACh step Showing thAt SOme subSEt with a PARtIculAr property is small. tHe FIrst step in the pRoof of leMMa \[LEM:stAbiLity\] is the fOlLowinG StabiliTY lEMMA, whICh is an almost iMmediate conSEquEnce of thEorEM \[thm:M:sTabilItY\].
\[Lem:Stability:fiRst\] LEt $\lambda {\gEqslanT}3$ And $n,k \in {\MAthbb{N}}$, wIth $k {\geQslAnt}(\Log n)^4$ ANd $K {\gEqsLaNT}2^{480} \laMBdA^{4}$. ThERe aRe at most $$\ExP\bIgg( - \frAc{\deLTA K}{2^9 \LambDa} \bIgg) {\lAmbda K / 2 \choose k}$$ sets $A \In \LAmbdA$ SucH that $$| a \setmInus p | > \dElta k$$ For eveRy ariThMetic progressioN $P$ of Size $\lambdA k/2$.
NOtE fiRsT that IF $k {\geqsLanT}(\loG n)^4$ and $k {\gEqslant}2^{480} \LAmbDa^{4}$ THEN $\fRac{k}{(\log n)^{3}} {\geqslant}k^{1/4} {\gEqSLAnT}2^{120} \lambda$. THerefoRE, aPpLYing TheoReM \[thM:M:stABIlity\] With $\GAmMa = 2^{-9} \lambda^{-1} \Delta$, iT FoLlOws that FoR all buT aT moSt $$\eXp\bigG( - \Frac{\Delta k}{2^9\Lambda} \biGg) {\lamBDa k / 2 \choose k}$$ sets $a \In \Lambda$, there EXiSTS $T \SUbseT A$, wIth $|T| {\leqslanT}(2^9 + 2^7 \lamBDa) \gaMma k {\LEqSlaNT}\deltA k$, sucH tHAt $a \Setminus T$ is containeD iN an ariThmetIc | = 2^{-18} \lambda^{-2}$.
\[lem:sta bilit y\] Le t$\la mbda {\geqslant}3$ and$n,k \in {\mathbb{N}}$ , wit h$ k {\ g eq slant }(\logn )^ 4 $ an d$k {\ ge q sl ant}2 ^{4 80} \la mbda^{20}$ . W ehave $$|\Lam b da \setminus \L ambda^*| {\l eqs lant}\ fr ac{ n ^2}{k } \ cdot|{\mat h cal{I} }| + \exp \b i gg( -\ frac{\d e l ta k}{ 2^{10} \lambda} \ b ig g ) {\lambda k / 2 \ch oo s ek } .$$
T o prove Le mm a \[l e m:stabi l it y \ ] , w e will success ively refin e $\ Lambda \ set m inus \ Lambd a^ * $,at each ste p sh owing tha t some subsetw ith a p articu lar pr oper t yis sm al l . T h efir s t s tep in t he p roofof L e m m a \[l em: stab ility \] is the fol low ings tab ility lemm a, w hi ch is an al mostim mediate consequ ence of Theor em\[ thm :M :stab i lity\] .
\[l em:stab ility:f i rst \] L e t$\lambda {\geqslan t} 3 $ a nd $n,k\in {\ m at hb b {N}}$, w it h $ k {\ g e qslan t}(\ l og n)^4$ a nd $k{ \g eq slant}2 ^{ 480} \ la mbd a^{ 4}$.T here are a t most $ $\exp \ bigg( - \frac{ \ delta k}{2^9\ la m b da } \bi gg) {\lambdak /2 \ch oose k} $$s ets $ A \in \ L am b da$ such that $$| A \ setmin us P| > \delta k$ $ for ever y a rithmeti c pr o gr e ssion $P$ of s ize $ \lambda k/ 2 $.
Note firs t that i f $k {\ge q s lant}(\l ogn)^ 4$and $ k{\geqslant}2^ { 4 80}\l ambda^{ 4}$ then $ \fr ac{ k}{ (\l og n)^{3}}{\geqsla nt }k ^{ 1/ 4}{\geq s lant}2^{ 12 0}\l amb da$.T herefo re, a pply in gT heo rem \[t h m: M : stab il it y\]wit h$\gam ma = 2^{ -9} \la mbda^{-1} \d e lta$ ,it follow s that for al lbut at mos t$$\ exp\bi g g ( - \fra c{\delta k}{2^9\lambda} \bigg){\l ambda k / 2 \choos e k }$$ se ts$ A \in\Lambd a$, t he ree x ists$ T \ sub se t A$, with $ |T| {\le qs lant }(2^9 + 2^7 \lambda) \gam m a k {\leqslant}\ del ta k $ , s uch th a t $ A\ set m i nus T$ is conta ined in an a r it hmetic | =_2^{-18} \lambda^{-2}$.
\[lem:stability\]_Let $\lambda {\geqslant}3$ and_$n,k \in_{\mathbb{N}}$,_with $k_{\geqslant}(\log_n)^4$ and $k_{\geqslant}2^{480} \lambda^{20}$. We_have $$|\Lambda \setminus \Lambda^*|_{\leqslant}\frac{n^2}{k} \cdot |{\mathcal{I}}|_+_\exp\bigg( - \frac{\delta k}{2^{10} \lambda} \bigg) {\lambda k / 2 \choose k}.$$
To prove Lemma \[lem:stability\],_we_will successively_refine_$\Lambda_\setminus \Lambda^*$, at each step_showing that some subset with_a particular_property is small. The first step in the_proof_of Lemma \[lem:stability\] is_the following stability lemma, which is an almost immediate_consequence of Theorem \[thm:M:stability\].
\[lem:stability:first\] Let $\lambda {\geqslant}3$_and $n,k \in_{\mathbb{N}}$,_with_$k {\geqslant}(\log n)^4$ and_$k {\geqslant}2^{480} \lambda^{4}$. There are at_most $$\exp\bigg( - \frac{\delta k}{2^9 \lambda}_\bigg) {\lambda k / 2 \choose_k}$$ sets $A \in \Lambda$ such_that $$| A \setminus P_| >_\delta k$$ for every arithmetic_progression $P$ of_size $\lambda_k/2$.
Note first that_if $k {\geqslant}(\log n)^4$ and $k_{\geqslant}2^{480} \lambda^{4}$ then_$\frac{k}{(\log n)^{3}} {\geqslant}k^{1/4} {\geqslant}2^{120} \lambda$. Therefore,_applying_Theorem \[thm:M:stability\] with $\gamma_=_2^{-9}_\lambda^{-1} \delta$,_it follows that_for_all but_at_most $$\exp\bigg( - \frac{\delta k}{2^9\lambda} \bigg)_{\lambda_k / 2 \choose k}$$ sets $A_\in \Lambda$, there exists_$T_\subset A$, with $|T|_{\leqslant}(2^9 + 2^7 \lambda) \gamma_k {\leqslant}\delta k$, such that $A_\setminus T$_is contained_in an arithmetic |
10.1140/epjd/e2007-00251-1) [****, ()](\doibase 10.1103/PhysRevB.87.035139) [****, ()](\doibase 10.1007/s00339-012-7130-0) [****, ()](\doibase
http://dx.doi.org/10.1016/j.nimb.2006.11.031) [****, ()](\doibase
http://dx.doi.org/10.1016/j.nima.2007.04.063) [****, ()](\doibase
10.1103/PhysRevLett.107.165003) [****, ()](\doibase http://dx.doi.org/10.1016/j.nima.2011.01.116) [****, ()](\doibase 10.1063/1.2737359) [****, ()](\doibase 10.1016/0031-8914(67)90062-6) [****, ()](\doibase 10.1103/PhysRevB.27.7466) [****, ()](\doibase 10.1016/j.ijheatmasstransfer.2004.09.015) [****, ()](\doibase 10.1016/j.cpc.2011.10.002) [****, ()](\doibase
http://dx.doi.org/10.1016/j.nimb.2013.10.087) [**](http://www.gnu.org/software/gsl), ed. (, ) [****, ()](http://stacks.iop.org/0022-3727/40/i=21/a=047) [****, ()](\doibase 10.1103/PhysRevB.26.2133) [****, ()](\doibase
http://dx.doi.org/10.1016/S0301-0104(99)00301-8) [****, ()](\doibase 10.1103/PhysRev.106.882) @noop [**]{}, ed. (, , ) [****, ()](\doibase 10.1088/0034-4885/76/3/036502) [****, ()](http://stacks.iop.org/0295-5075/56/i=1/a=138) [****, ()](\doibase 10.1002/ctpp.201010005) [****, ()](\doibase
10.1007/s00339-012-7189-7) [****, ()](\doibase 10.1063/1.1510565) [****, ()](\doibase 10.1063/1.36664 | 10.1140 / epjd / e2007 - 00251 - 1) [ * * * *, () ] (\doibase 10.1103 / PhysRevB.87.035139) [ * * * *, () ] (\doibase 10.1007 / s00339 - 012 - 7130 - 0) [ * * * *, () ] (\doibase
http://dx.doi.org/10.1016/j.nimb.2006.11.031) [ * * * *, () ] (\doibase
http://dx.doi.org/10.1016/j.nima.2007.04.063) [ * * * *, () ] (\doibase
10.1103 / PhysRevLett.107.165003) [ * * * *, () ] (\doibase http://dx.doi.org/10.1016/j.nima.2011.01.116) [ * * * *, () ] (\doibase 10.1063/1.2737359) [ * * * *, () ] (\doibase 10.1016/0031 - 8914(67)90062 - 6) [ * * * *, () ] (\doibase 10.1103 / PhysRevB.27.7466) [ * * * *, () ] (\doibase 10.1016 / j.ijheatmasstransfer.2004.09.015) [ * * * *, () ] (\doibase 10.1016 / j.cpc.2011.10.002) [ * * * *, () ] (\doibase
http://dx.doi.org/10.1016/j.nimb.2013.10.087) [ * * ] (http://www.gnu.org/software/gsl), ed. (, ) [ * * * *, () ] (http://stacks.iop.org/0022-3727/40/i=21/a=047) [ * * * *, () ] (\doibase 10.1103 / PhysRevB.26.2133) [ * * * *, () ] (\doibase
http://dx.doi.org/10.1016/S0301-0104(99)00301-8) [ * * * *, () ] (\doibase 10.1103 / PhysRev.106.882) @noop [ * * ] { }, ed. (, , ) [ * * * *, () ] (\doibase 10.1088/0034 - 4885/76/3/036502) [ * * * *, () ] (http://stacks.iop.org/0295-5075/56/i=1/a=138) [ * * * *, () ] (\doibase 10.1002 / ctpp.201010005) [ * * * *, () ] (\doibase
10.1007 / s00339 - 012 - 7189 - 7) [ * * * *, () ] (\doibase 10.1063/1.1510565) [ * * * *, () ] (\doibase 10.1063/1.36664 | 10.1140/epjf/e2007-00251-1) [****, ()](\doibase 10.1103/PhysRevB.87.035139) [****, ()](\dolbase 10.1007/s00339-012-7130-0) [****, ()](\doibase
http://de.doi.org/10.1016/n.nimb.2006.11.031) [****, ()](\duibase
http://dx.doi.org/10.1016/j.nima.2007.04.063) [****, ()](\dlivase
10.1103/PhysRevLett.107.165003) [****, ()](\doibase fttp://dx.doi.lrg/10.1016/j.nima.2011.01.116) [****, ()](\dombase 10.1063/1.2737359) [****, ()](\doibase 10.1016/0031-8914(67)90062-6) [****, ()](\doibasc 10.1103/PhyaVevB.27.7466) [****, ()](\voibase 10.1016/j.ijheatkasstransfar.2004.09.015) [****, ()](\doibase 10.1016/j.c[c.2011.10.002) [****, ()](\boibase
http://dx.doi.org/10.1016/j.nimb.2013.10.087) [**](http://www.gnt.org/sofywwre/gsl), ed. (, ) [****, ()](http://ftaclf.iop.kgg/0022-3727/40/l=21/a=047) [****, ()](\doibase 10.1103/PhysRevB.26.2133) [****, ()](\doibase
hftp://dx.dop.org/10.1016/S0301-0104(99)00301-8) [****, ()](\doibase 10.1103/PhusRev.106.882) @noop [**]{}, ed. (, , ) [****, ()](\doibase 10.1088/0034-4885/76/3/036502) [****, ()](hhtp://shacks.iop.org/0295-5075/56/i=1/a=138) [****, ()](\doihase 10.1002/ctpp.201010005) [****, ()](\eoibwwe
10.1007/s00339-012-7189-7) [****, ()](\doibare 10.1063/1.1510565) [****, ()](\doibase 10.1063/1.36664 | 10.1140/epjd/e2007-00251-1) [****, ()](\doibase 10.1103/PhysRevB.87.035139) [****, ()](\doibase 10.1007/s00339-012-7130-0) http://dx.doi.org/10.1016/j.nimb.2006.11.031) ()](\doibase http://dx.doi.org/10.1016/j.nima.2007.04.063) ()](\doibase 10.1103/PhysRevLett.107.165003) [****, [****, 10.1016/0031-8914(67)90062-6) [****, ()](\doibase [****, ()](\doibase 10.1016/j.ijheatmasstransfer.2004.09.015) ()](\doibase 10.1016/j.cpc.2011.10.002) [****, ()](\doibase http://dx.doi.org/10.1016/j.nimb.2013.10.087) [**](http://www.gnu.org/software/gsl), (, ) [****, ()](http://stacks.iop.org/0022-3727/40/i=21/a=047) [****, ()](\doibase 10.1103/PhysRevB.26.2133) [****, ()](\doibase http://dx.doi.org/10.1016/S0301-0104(99)00301-8) [****, ()](\doibase 10.1103/PhysRev.106.882) [**]{}, ed. (, , ) [****, ()](\doibase 10.1088/0034-4885/76/3/036502) [****, ()](http://stacks.iop.org/0295-5075/56/i=1/a=138) [****, ()](\doibase 10.1002/ctpp.201010005) ()](\doibase [****, 10.1063/1.1510565) ()](\doibase 10.1063/1.36664 | 10.1140/epjd/e2007-00251-1) [****, ()](\doibase 10.1103/PhysRevB.87.035139) [****, ()](\doibaSe 10.1007/s00339-012-7130-0) [****, ()](\doibase
Http://dX.doI.orG/10.1016/j.Nimb.2006.11.031) [****, ()](\DoibAse
http://dx.doi.orG/10.1016/J.nimA.2007.04.063) [****, ()](\doibase
10.1103/PhysRevLett.107.165003) [****, ()](\doiBase hTtP://Dx.doI.OrG/10.1016/j.nimA.2011.01.116) [****, ()](\doibasE 10.1063/1.2737359) [****, ()](\DoIBAse 10.1016/0031-8914(67)90062-6) [****, ()](\DoIbAse 10.1103/phYSREvB.27.7466) [****, ()](\doIbaSe 10.1016/j.ijheAtmasstranSfeR.2004.09.015) [****, ()](\dOibase 10.1016/j.cpc.2011.10.002) [****, ()](\doIBaSe
http://dx.doI.orG/10.1016/j.nimb.2013.10.087) [**](http://wwW.gnU.org/soFtWarE/Gsl), ed. (, ) [****, ()](HttP://stacKs.iop.oRG/0022-3727/40/i=21/a=047) [****, ()](\doiBase 10.1103/PhysREvb.26.2133) [****, ()](\DoibasE
Http://dx.dOI.OrG/10.1016/S0301-0104(99)00301-8) [****, ()](\doIbase 10.1103/PhysRev.106.882) @noop [**]{}, eD. (, , ) [****, ()](\DoIBase 10.1088/0034-4885/76/3/036502) [****, ()](http://stacks.Iop.org/0295-5075/56/I=1/a=138) [****, ()](\DOiBASe 10.1002/cTpp.201010005) [****, ()](\Doibase
10.1007/s00339-012-7189-7) [****, ()](\doIbAse 10.1063/1.1510565) [****, ()](\doIBase 10.1063/1.36664 | 10.1140/epjd/e2007-00251-1 ) [****, ()](\ doi bas e10.1 103/ PhysRevB.87.03 5 139) [****, ()](\doibase10.10 07 / s003 3 9- 012-7 130-0)[ ** * * , () ]( \do ib a se
ht tp: //dx.do i.org/10.1 016 /j .nimb.2006.1 1 .0 31) [****, ( )](\doibase h ttp:// dx .do i .org/ 10. 1016/ j.nima . 2007.0 4.063) [* ** * , ()] ( \doibas e 10.1 103/PhysRevLett.1 0 7. 1 65003) [****, ()](\ do i ba s e ht tp: //dx.doi.o rg /10.1 0 16/j.ni m a. 2 0 1 1.0 1 .116) [****,()](\doibas e 10 .1063/ 1. 273 7 359) [ ****, ( ) ](\ doibase 10. 1016 /0031-891 4(67)9 0 062-6)[ ****, ( )](\do iba se10.1 1 03 /P hys Re v B.2 7 .7 466 ) [* ***, () ]( \d oibas e 10 . 1 0 1 6/j. ijh eatm asstr ansfer.2004.0 9.0 15)[ *** *, ( )](\d oiba se 10.1 016/j. cpc.2 01 1.10.002) [**** , ( )](\doiba se htt p: //dx. d oi.org /10 .10 16/j.ni mb.2013 . 10. 08 7 ) [* *](http://www.gnu. or g / so ftware/g sl), e d. ( , ) [**** , () ](ht t p ://st acks . io p.org/00 22-372 7 /4 0/ i=21/a= 04 7) [** ** , ( )]( \doib a se 1 0.1103 /PhysRev B.26. 2 133) [****, () ] (\doibase
h t tp : / /d x .doi .or g/10.1016/S 0301 - 0104 (99) 0 03 01- 8 ) [** **, ( )] ( \d o ibase 10.1103/PhysR ev .106.8 82) @ noop [**]{}, ed. (, ,) [ ****, ( )](\ d oi b ase 10.1088/00 34-48 85/76/3/03 6 502) [** **, ()](http ://stacks . i op.org/0 295 -50 75/ 56/ i = 1/ a=138) [****, ( )](\ do ibase 1 0.1 002/ctp p.2 010 100 05) [ ****, () ](\doiba se
1 0. 100 7/s00 3 39-012-7 18 9-7 )[** **, ( )](\do ibase 10. 10 63 / 1.1 510565) [* * * *, () ]( \doi bas e10.10 63/1 . 366 64 | 10.1140/epjd/e2007-00251-1) [****, _()](\doibase 10.1103/PhysRevB.87.035139)_[****, ()](\doibase 10.1007/s00339-012-7130-0) [****, _()](\doibase
_http://dx.doi.org/10.1016/j.nimb.2006.11.031)_[****, ()](\doibase
__http://dx.doi.org/10.1016/j.nima.2007.04.063) [****, ()](\doibase
_ 10.1103/PhysRevLett.107.165003) [****, _()](\doibase http://dx.doi.org/10.1016/j.nima.2011.01.116) [****, ()](\doibase_10.1063/1.2737359) [****, ()](\doibase_10.1016/0031-8914(67)90062-6)_[****, ()](\doibase 10.1103/PhysRevB.27.7466) [****, ()](\doibase 10.1016/j.ijheatmasstransfer.2004.09.015) [****, ()](\doibase 10.1016/j.cpc.2011.10.002) [****, ()](\doibase
http://dx.doi.org/10.1016/j.nimb.2013.10.087) [**](http://www.gnu.org/software/gsl), ed. (, )_[****, _()](http://stacks.iop.org/0022-3727/40/i=21/a=047) [****,_()](\doibase_10.1103/PhysRevB.26.2133)_[****, ()](\doibase
http://dx.doi.org/10.1016/S0301-0104(99)00301-8) [****,_()](\doibase 10.1103/PhysRev.106.882) @noop [**]{}, ed. (, , ) [****, _()](\doibase 10.1088/0034-4885/76/3/036502)_[****, ()](http://stacks.iop.org/0295-5075/56/i=1/a=138) [****, ()](\doibase 10.1002/ctpp.201010005) [****, ()](\doibase
_10.1007/s00339-012-7189-7)_[****, ()](\doibase 10.1063/1.1510565)_[****, ()](\doibase 10.1063/1.36664 |
rho_{a_1}(W)}$ such that ${\alpha}_1(u_1) = {\alpha}_1(u_2)$ and if $P_{\rho_{a_1}(W)}(\operatorname{\textsf{fact}}, u_1, u_2) = s_1 s_2 s_3$, then $$\label{in7}
\min(|s_2|, |s_1|+|s_3|) \ge \tfrac 13 | {\partial}\rho_{a_1}({\Delta})| = \tfrac 13 |W|_{\bar a_1} \.$$ It follows from definitions and Lemma \[b1\] that there are some vertices $v_1, v_2 \in P_W$ such that $\beta(v_j) = u_j$, $j=1,2$, the vertices ${\alpha}(v_1), {\alpha}(v_2)$ belong to ${\partial}\Gamma$, where $\Gamma$ is an $a_i$-band in ${\Delta}$, $i >1$, and the vertices ${\alpha}(v_1), {\alpha}(v_2)$ can be connected along ${\partial}\Gamma$ by a simple reduced path $r$ such that ${\varphi}(r) \equiv a_1^{\ell}$ for some integer $\ell$, where $\ell = 0$ if $i >2$. Clearly, $r$ is a special arc of ${\Delta}$. Denote $P_{W}(\operatorname{\textsf{fact}}, v_1, v_2) = p_1 p_2 p_3$. Then it follows from definitions that $| {\varphi}(p_i) |_{\bar a_1} = |s_i|$, $i=1,2,3$, hence the inequality yields that $$\min( | {\varphi}(p_2) |_{\bar a_1}, | {\varphi}(p_1) |_{\bar a_1} + | {\varphi}(p_3) |_{\bar a_1} ) \ge \tfrac 13 | {\partial}\rho_{a_1}({\Delta})| = \tfrac 13 |W|_{\bar a_1} ,$$ as required.
Furthermore, since $r$ is simple, reduced and $r_- \ne r_+$ unless $|r| =0$, the path $r$ contains every $a_1$-edge $ | rho_{a_1}(W)}$ such that $ { \alpha}_1(u_1) = { \alpha}_1(u_2)$ and if $ P_{\rho_{a_1}(W)}(\operatorname{\textsf{fact } }, u_1, u_2) = s_1 s_2 s_3 $, then $ $ \label{in7 }
\min(|s_2|, |s_1|+|s_3|) \ge \tfrac 13 | { \partial}\rho_{a_1}({\Delta})| = \tfrac 13 |W|_{\bar a_1 } \.$$ It follows from definitions and Lemma \[b1\ ] that there are some vertices $ v_1, v_2 \in P_W$ such that $ \beta(v_j) = u_j$, $ j=1,2 $, the vertex $ { \alpha}(v_1), { \alpha}(v_2)$ belong to to $ { \partial}\Gamma$, where $ \Gamma$ is an $ a_i$-band in $ { \Delta}$, $ i > 1 $, and the vertices $ { \alpha}(v_1), { \alpha}(v_2)$ can be connected along $ { \partial}\Gamma$ by a childlike reduce path $ r$ such that $ { \varphi}(r) \equiv a_1^{\ell}$ for some integer $ \ell$, where $ \ell = 0 $ if $ i > 2$. Clearly, $ r$ is a particular bow of $ { \Delta}$. Denote $ P_{W}(\operatorname{\textsf{fact } }, v_1, v_2) = p_1 p_2 p_3$. Then it follows from definition that $ | { \varphi}(p_i) |_{\bar a_1 } = |s_i|$, $ i=1,2,3 $, hence the inequality yields that $ $ \min ( | { \varphi}(p_2) |_{\bar a_1 }, | { \varphi}(p_1) |_{\bar a_1 } + | { \varphi}(p_3) |_{\bar a_1 }) \ge \tfrac 13 | { \partial}\rho_{a_1}({\Delta})| = \tfrac 13 |W|_{\bar a_1 }, $ $ as want.
Furthermore, since $ r$ is simple, reduced and $ r_- \ne r_+$ unless $ |r| = 0 $, the way $ r$ contains every $ a_1$-edge $ | rho_{w_1}(W)}$ such that ${\alpha}_1(u_1) = {\alpma}_1(u_2)$ and if $P_{\rho_{a_1}(C)}(\iperatmrname{\fextsf{fazt}}, u_1, u_2) = s_1 s_2 s_3$, then $$\label{in7}
\mmn(|s_2|, |s_1|+|s_3|) \gt \tfrac 13 | {\partial}\rhu_{a_1}({\Delta})| = \tfrac 13 |W|_{\bac a_1} \.$$ It follows hdom defluitiohd anb Oemma \[b1\] that thgre are some vertices $v_1, v_2 \hn P_C$ such that $\beta(v_j) = u_j$, $j=1,2$, the verticef ${\alpha}(f_1), {\wlpha}(v_2)$ belong jo ${\pagtyal}\Gzmma$, where $\Gamma$ is an $a_i$-band in ${\Dslta}$, $i >1$, and the vertoces ${\alpha}(v_1), {\alpha}(v_2)$ can be fonnfcted along ${\partiap}\Gamma$ by a sim[oe reduced pxth $r$ such that ${\varphi}(t) \equiv a_1^{\ell}$ for some integer $\elu$, whexe $\ell = 0$ if $i >2$. Cpgarly, $r$ is a specpal arc of ${\Delta}$. Denode $P_{W}(\oprratorname{\textxf{fect}}, c_1, v_2) = p_1 p_2 p_3$. Then it fmllows from definijions that $| {\rarphi}(p_i) |_{\bar a_1} = |s_i|$, $i=1,2,3$, yebce tve itequxoith yjekda that $$\mii( | {\varphi}(p_2) |_{\bar a_1}, | {\vqrphi}(p_1) |_{\bar a_1} + | {\varkhi}([_3) |_{\bar a_1} ) \ge \ffrac 13 | {\[artial}\rho_{a_1}({\Delta})| = \tfrac 13 |W|_{\bar a_1} ,$$ as rtquirsd.
Furthermore, since $r$ iw simple, reduced and $t_- \ne r_+$ unlqss $|r| =0$, the path $r$ contains every $a_1$-edge $ | rho_{a_1}(W)}$ such that ${\alpha}_1(u_1) = {\alpha}_1(u_2)$ and u_1, = s_1 s_3$, then $$\label{in7} | = \tfrac 13 a_1} \.$$ It from definitions and Lemma \[b1\] that are some vertices $v_1, v_2 \in P_W$ such that $\beta(v_j) = u_j$, $j=1,2$, vertices ${\alpha}(v_1), {\alpha}(v_2)$ belong to ${\partial}\Gamma$, where $\Gamma$ is an $a_i$-band in ${\Delta}$, >1$, the ${\alpha}(v_1), can be connected along ${\partial}\Gamma$ by a simple reduced path $r$ such that ${\varphi}(r) \equiv a_1^{\ell}$ some integer $\ell$, where $\ell = 0$ if >2$. Clearly, $r$ is special arc of ${\Delta}$. Denote v_1, = p_1 p_3$. it from definitions that {\varphi}(p_i) |_{\bar a_1} = |s_i|$, $i=1,2,3$, hence the inequality yields that $$\min( | {\varphi}(p_2) |_{\bar a_1}, | |_{\bar a_1} {\varphi}(p_3) |_{\bar ) \tfrac | {\partial}\rho_{a_1}({\Delta})| = |W|_{\bar a_1} ,$$ as required. Furthermore, simple, reduced and $r_- \ne r_+$ unless $|r| the path contains every $a_1$-edge $ | rho_{a_1}(W)}$ such that ${\alpha}_1(u_1) = {\alpha}_1(u_2)$ And if $P_{\rho_{a_1}(w)}(\operAtoRnaMe{\TextSf{faCt}}, u_1, u_2) = s_1 s_2 s_3$, then $$\labEL{in7}
\mIn(|s_2|, |s_1|+|s_3|) \ge \tfrac 13 | {\partial}\rho_{A_1}({\DeltA})| = \tFRac 13 |W|_{\BAr A_1} \.$$ It foLlows frOM dEFIniTiOnS anD LEMmA \[b1\] thaT thEre are sOme verticeS $v_1, v_2 \In p_W$ such that $\beTA(v_J) = u_j$, $j=1,2$, the verTicEs ${\alpha}(v_1), {\alphA}(v_2)$ bElong tO ${\pArtIAl}\GamMa$, wHere $\GAmma$ is AN $a_i$-banD in ${\Delta}$, $i >1$, AnD The verTIces ${\alpHA}(V_1), {\aLpha}(V_2)$ can be connected alONg ${\PArtial}\Gamma$ by a Simple ReDUcED PatH $r$ sUch that ${\varPhI}(r) \equIV a_1^{\ell}$ foR SoME INteGEr $\ell$, where $\ell = 0$ If $i >2$. Clearly, $r$ IS a sPecial ArC of ${\dElta}$. DeNote $P_{w}(\oPEraTorname{\textSf{faCt}}, v_1, v_2) = p_1 p_2 p_3$. ThEn it foLLows froM DefinitIons thAt $| {\vArpHi}(p_i) |_{\BAr A_1} = |s_I|$, $i=1,2,3$, hEnCE thE InEquALitY yields tHaT $$\mIn( | {\varPhi}(p_2) |_{\BAR A_1}, | {\VarpHi}(p_1) |_{\Bar a_1} + | {\VarphI}(p_3) |_{\bar a_1} ) \ge \tfrac 13 | {\ParTial}\RHo_{a_1}({\delta})| = \Tfrac 13 |w|_{\bar A_1} ,$$ aS requIred.
FuRtherMoRe, since $r$ is simplE, redUced and $r_- \nE r_+$ uNlEss $|R| =0$, tHe patH $R$ contaIns EveRy $a_1$-edge $ | rho_{a_1}(W)}$ such that $ {\alpha}_1 (u_1) ={\a lp ha}_ 1(u_ 2)$ and if $P_ { \rho _{a_1}(W)}(\operatorna me{\t ex t sf{f a ct }}, u _1, u_2 ) = s_1 s _2 s_ 3$ , t hen $ $\l abel{in 7}
\min(|s _2| , |s_1|+|s_3| ) \ ge \tfrac13| {\partial} \rh o_{a_1 }( {\D e lta}) | = \tf rac 13 |W|_{ \bar a_1} \ . $$ Itf ollowsf r om def initions and Lemm a \ [ b1\] that ther e areso m ev e rti ces $v_1, v_2 \ in P_ W $ sucht ha t $ \be t a(v_j) = u_j$ , $j=1,2$,t hevertic es ${ \ alpha} (v_1) ,{ \al pha}(v_2)$belo ng to ${\ partia l }\Gamma $ , where $\Gam ma$ is an$ a_ i$ -ba nd in$ {\ Del t a}$ , $i >1$ ,an d the ver t i c e s ${ \al pha} (v_1) , {\alpha}(v_ 2)$ can beconne ctedalon g${\pa rtial} \Gamm a$ by a simple re duce d path $r $ s uc h t ha t ${\ v arphi} (r) \e quiv a_ 1^{\ell } $ f or s o me integer $\ell$, w he r e $ \ell = 0 $ if $ i > 2$ . Clearly ,$r$ isa speci al a r cof ${\De lta}$. De no te $P_{ W} (\oper at orn ame {\tex t sf{f act}}, v_1, v_ 2) = p_1 p_2 p_3$.T hen it follow s f r o md efin iti ons that $| {\v a rphi }(p_ i )|_{ \ bar a _1} = | s _i | $, $i=1,2,3$, hence t he ine quali ty yields tha t $$\min(| {\varphi }(p_ 2 )| _{\bar a_1}, | {\ varphi}(p_ 1 ) |_{\ba r a_1 } + | { \varphi}( p _ 3) |_{\b ara_1 } ) \ g e\tfrac 13 | { \ p arti al }\rho_{ a_1 }({\Del ta} )|= \tf ra c 13 |W| _{\bar a _1 },$ $asrequi r ed.
Fur th erm or e,since $r$ is simp le,re du c edand $r_ - \ n e r_+ $un less $| r| =0$, the pat h $r$ c ontains e ver y $a_ 1$ -e dge $ | rho_{a_1}(W)}$ such_that ${\alpha}_1(u_1)_= {\alpha}_1(u_2)$ and if_$P_{\rho_{a_1}(W)}(\operatorname{\textsf{fact}}, u_1,_u_2)_ =_s_1_s_2 s_3$, then_$$\label{in7}
\min(|s_2|, |s_1|+|s_3|)_\ge \tfrac 13 |_{\partial}\rho_{a_1}({\Delta})| = _\tfrac_13 |W|_{\bar a_1} \.$$ It follows from definitions and Lemma \[b1\] that there are_some_vertices $v_1,_v_2_\in_P_W$ such that $\beta(v_j) =_u_j$, $j=1,2$, the vertices ${\alpha}(v_1),_{\alpha}(v_2)$ belong_to ${\partial}\Gamma$, where $\Gamma$ is an $a_i$-band in_${\Delta}$,_$i >1$, and_the vertices ${\alpha}(v_1), {\alpha}(v_2)$ can be connected along ${\partial}\Gamma$_by a simple reduced path $r$_such that ${\varphi}(r)_\equiv_a_1^{\ell}$_for some integer $\ell$,_where $\ell = 0$ if $i_>2$. Clearly, $r$ is a special_arc of ${\Delta}$. Denote $P_{W}(\operatorname{\textsf{fact}}, v_1, v_2)_ = p_1 p_2 p_3$. Then_it follows from definitions that_$| {\varphi}(p_i)_|_{\bar a_1} = |s_i|$, $i=1,2,3$,_hence the inequality_yields that_$$\min( |_{\varphi}(p_2) |_{\bar a_1}, |_{\varphi}(p_1) |_{\bar a_1}_+ | {\varphi}(p_3) |_{\bar a_1}_)_ \ge_\tfrac_13_| {\partial}\rho_{a_1}({\Delta})|_= \tfrac_13_ |W|_{\bar_a_1}_,$$ as required.
Furthermore, since $r$ is_simple,_reduced and $r_- \ne r_+$ unless $|r|_=0$, the path $r$_contains_every $a_1$-edge $ |
exp(a|\xi|_p^{2}).$$
See [@kuo], Theorem 8.6.
Limit theorems for Volterra processes {#volterra}
=====================================
One-dimensional case
--------------------
Consider a Volterra process $B=( B_t )_{t\geq 0}$ of the form $$\label{star}
B_t=\int_0^t K(t,s)\,\mathrm{d}W_s,$$ where $K(t,s)$ satisfies $\int_0^tK(t,s)^2 \,\mathrm{d}s<\infty$ for all $t>0$ and $W$ is the Brownian motion defined on the white noise probability space introduced in the last section. Note that the $S$-transform of the random variable $B_t$ is given by $$\label{e1aa}
S(B_t)(\xi)=\int_0^{t}K(t,s)\xi(s)\,\mathrm{d}s$$ for any $\xi\in \mathcal{S}({\mathbb{R}})$. We introduce the following assumptions on the kernel $K$:
1. $K$ is continuously differentiable on $\{0<s<t<\infty \}$ and for any $ t>0$, we have $$\int_0^t \bigg|\frac{\partial K}{\partial t}(t,s) \bigg|(t-s)\,\mathrm{d}s <\infty;$$
2. $k(t)=\int_0^{t}K(t,s)\,\mathrm{d}s $ is continuously differentiable on $(0,\infty)$.
Consider the operator $K_+$ defined by $$K_+\xi(t)=k'(t)\xi(t)+\int_0^t \frac{\partial K}{\partial t}(t,r)\bigl(\xi(r)-\xi(t)\bigr)\,\mathrm{d}r,$$ where $t>0$ and $\xi\in \mathcal{S}({\mathbb{R}})$. From Theorem \[F\], it follows that the linear mapping $\xi\rightarrow K_+\xi(t)$ is the $S$-transform of a Hida distribution. More precisely, according to [@nualartwhite], define the function $$\label{C}
C(t)=|k'(t)|+\int_0^t \bigg|\frac{\partial K | exp(a|\xi|_p^{2}).$$
See [ @kuo ], Theorem 8.6.
Limit theorems for Volterra processes { # volterra }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
One - dimensional case
--------------------
Consider a Volterra procedure $ B= (B_t) _ { t\geq 0}$ of the human body $ $ \label{star }
B_t=\int_0^t K(t, s)\,\mathrm{d}W_s,$$ where $ K(t, s)$ satisfies $ \int_0^tK(t, s)^2 \,\mathrm{d}s<\infty$ for all $ t>0 $ and $ W$ is the Brownian motion defined on the white noise probability distance introduced in the last part. Note that the $ S$-transform of the random variable $ B_t$ is contribute by $ $ \label{e1aa }
S(B_t)(\xi)=\int_0^{t}K(t, s)\xi(s)\,\mathrm{d}s$$ for any $ \xi\in \mathcal{S}({\mathbb{R}})$. We introduce the following assumption on the kernel $ K$:
1. $ K$ is continuously differentiable on $ \{0 < s < t<\infty \}$ and for any $ t>0 $, we get $ $ \int_0^t \bigg|\frac{\partial K}{\partial t}(t, s) \bigg|(t - s)\,\mathrm{d}s < \infty;$$
2. $ k(t)=\int_0^{t}K(t, s)\,\mathrm{d}s $ is continuously differentiable on $ (0,\infty)$.
Consider the operator $ K_+$ defined by $ $ K_+\xi(t)=k'(t)\xi(t)+\int_0^t \frac{\partial K}{\partial t}(t, r)\bigl(\xi(r)-\xi(t)\bigr)\,\mathrm{d}r,$$ where $ t>0 $ and $ \xi\in \mathcal{S}({\mathbb{R}})$. From Theorem \[F\ ], it follows that the linear function $ \xi\rightarrow K_+\xi(t)$ is the $ S$-transform of a Hida distribution. More precisely, according to [ @nualartwhite ], define the function $ $ \label{C }
C(t)=|k'(t)|+\int_0^t \bigg|\frac{\partial potassium | exp(w|\xi|_p^{2}).$$
See [@kuo], Theorem 8.6.
Limiu theorems for Volterra 'rocessss {#voltefra}
=====================================
One-dimensional case
--------------------
Considxr a Volttgra process $B=( B_t )_{t\geq 0}$ of the vorm $$\labwl{ster}
B_t=\int_0^t K(t,s)\,\mathcj{d}W_s,$$ whcxe $K(t,a)$ satnshies $\int_0^tK(t,s)^2 \,\majhrm{d}s<\infty$ xor all $t>0$ and $F$ ks the Brownian motion defined on the white molse probabiliti spabe intdoduced in the last section. Note tgat the $S$-transform og the random variable $B_t$ id gigen by $$\label{e1aa}
S(B_t)(\di)=\int_0^{t}K(t,s)\xi(w)\,\matrem{d}s$$ for any $\xi\in \mathcal{S}({\mathbb{R}})$. We introduce the following assjmptipns on the kwrnfn $K$:
1. $K$ is cintinlously differcmtiabla on $\{0<s<t<\onfty \}$ and for anb $ t>0$, we have $$\int_0^t \bigj|\frac{\partial K}{\partiaj t}(t,s) \bigc|(t-a)\,\mathrm{d}s <\infty;$$
2. $j(t)=\unt_0^{t}K(j,s)\,\matvrm{d}r $ ir cknvinhously dihferentiabls on $(0,\infty)$.
Cinsider the operatot $H_+$ defined by $$K_+\si(t)=k'(t)\xy(t)+\ynt_0^t \frac{\partial K}{\partial t}(t,r)\bigl(\xi(r)-\xi(t)\tigd)\,\mathrm{d}r,$$ where $t>0$ and $\xu\in \mathcal{S}({\mathbb{R}})$. Vrom Theowem \[F\], it follows that the linear mapping $\xi\rightarsow K_+\ei(g)$ iw bhe $R$-rrwnsform of a Hida distribution. More precisely, zcvogding to [@nualartwmite], define the fumchipg $$\label{C}
C(t)=|k'(t)|+\ivt_0^t \biyf|\fdac{\partial K | exp(a|\xi|_p^{2}).$$ See [@kuo], Theorem 8.6. Limit theorems processes ===================================== One-dimensional -------------------- Consider a 0}$ the form $$\label{star} K(t,s)\,\mathrm{d}W_s,$$ where $K(t,s)$ $\int_0^tK(t,s)^2 \,\mathrm{d}s<\infty$ for all $t>0$ and is the Brownian motion defined on the white noise probability space introduced in last section. Note that the $S$-transform of the random variable $B_t$ is given $$\label{e1aa} for $\xi\in We introduce the following assumptions on the kernel $K$: 1. $K$ is continuously differentiable on $\{0<s<t<\infty and for any $ t>0$, we have $$\int_0^t K}{\partial t}(t,s) \bigg|(t-s)\,\mathrm{d}s <\infty;$$ $k(t)=\int_0^{t}K(t,s)\,\mathrm{d}s $ is continuously differentiable $(0,\infty)$. the operator defined $$K_+\xi(t)=k'(t)\xi(t)+\int_0^t K}{\partial t}(t,r)\bigl(\xi(r)-\xi(t)\bigr)\,\mathrm{d}r,$$ where and $\xi\in \mathcal{S}({\mathbb{R}})$. From Theorem \[F\], it follows that the linear mapping $\xi\rightarrow K_+\xi(t)$ is the $S$-transform a Hida precisely, according [@nualartwhite], the $$\label{C} C(t)=|k'(t)|+\int_0^t \bigg|\frac{\partial | exp(a|\xi|_p^{2}).$$
See [@kuo], Theorem 8.6.
Limit Theorems foR VoltErrA prOcEsseS {#volTerra}
=====================================
One-dimensIOnal Case
--------------------
Consider a Volterra pRocesS $B=( b_T )_{t\geQ 0}$ Of The foRm $$\label{STaR}
b_T=\inT_0^t k(t,S)\,\maThRM{d}w_s,$$ wheRe $K(T,s)$ satisFies $\int_0^tK(t,S)^2 \,\maThRm{d}s<\infty$ for ALl $T>0$ and $W$ is the broWnian motion dEfiNed on tHe WhiTE noisE prObabiLity spACe intrOduced in tHe LAst secTIon. Note THAt The $S$-Transform of the ranDOm VAriable $B_t$ is givEn by $$\laBeL{E1aA}
s(b_t)(\xI)=\inT_0^{t}K(t,s)\xi(s)\,\maThRm{d}s$$ fOR any $\xi\iN \MaTHCAl{S}({\MAthbb{R}})$. We introDuce the follOWinG assumPtIonS On the kErnel $k$:
1. $K$ IS coNtinuously dIffeRentiable On $\{0<s<t<\inFTy \}$ and foR Any $ t>0$, we hAve $$\int_0^T \biGg|\fRac{\pARtIaL K}{\pArTIal T}(T,s) \BigG|(T-s)\,\mAthrm{d}s <\iNfTy;$$
2. $K(t)=\int_0^{T}K(t,s)\,\MATHRm{d}s $ Is cOntiNuousLy differentiaBle On $(0,\inFTy)$.
COnsidEr the OperAtOr $K_+$ deFined bY $$K_+\xi(t)=K'(t)\Xi(t)+\int_0^t \frac{\partIal K}{\Partial t}(t,R)\biGl(\Xi(r)-\Xi(T)\bigr)\,\MAthrm{d}R,$$ whEre $T>0$ and $\xi\iN \mathcaL{s}({\maThBB{r}})$. frOm Theorem \[F\], it followS tHAT tHe linear MappinG $\Xi\RiGHtarrow K_+\Xi(T)$ is The $S$-TRAnsfoRm of A hiDa distriBution. mOrE pReciselY, aCcordiNg To [@nUalArtwhITe], deFine thE functioN $$\labeL{c}
C(t)=|k'(t)|+\int_0^t \bigg|\fRAc{\partial K | exp(a|\xi|_p^{2}).$$
See[@kuo], Th eorem 8. 6.
L imit the orems for Volt e rraprocesses {#volterra}===== == = ==== = == ===== ======= = == = = === ==
One -d i me nsion alcase
-- ---------- --- -- ---
Conside r a Volterrapro cess $B=( B_ t ) _{t\ge q0}$ of th e f orm $ $\labe l {star}
B_t=\int _0 ^ t K(t, s )\,\mat h r m{ d}W_ s,$$ where $K(t,s ) $s atisfies $\int _0^tK( t, s )^ 2 \,\ mat hrm{d}s<\i nf ty$ f o r all $ t >0 $ a nd$ W$ is the Bro wnian motio n de finedon th e white nois ep rob ability spa ce i ntroduced in th e last s e ction.Note t hat th e $S $ -t ra nsf or m of th e r a ndo m variab le $ B_t$is g i v e n by$$\ labe l{e1a a}
S(B_t)(\xi )=\ int_ 0 ^{t }K(t, s)\xi (s)\ ,\ mathr m{d}s$ $ for a ny $\xi\in \ma thca l{S}({\ma thb b{ R}} )$ . Wei ntrodu cethe follow ing ass u mpt io n s on the kernel $K$:
1. $K $ is con tinuou s ly d i fferenti ab leon $ \ { 0<s<t <\in f ty \}$ and for a n y$t>0$, w ehave $ $\ int _0^ t \big g|\fra c{\parti al K} { \partial t}(t, s ) \bigg|(t-s) \ ,\ m a th r m{d} s < \infty;$$
2. $ k(t) =\in t _0 ^{t } K(t,s )\,\m at h rm { d}s $ is continuous ly diffe renti able on $(0,\ infty)$.
C o n sider th e op e ra t or $K_+$ defin ed by $$K_+\xi( t )=k'(t)\ xi(t) +\int_0^ t \frac{\ p a rtial K} {\p art ial t} ( t ,r )\bigl(\xi(r) - \ xi(t )\ bigr)\, \ma thrm{d} r,$ $ w her e $ t> 0$ and $\ xi\in \ ma th ca l{ S}( {\mat h bb{R}})$ .Fro mThe orem\ [F\],it fo llow sth a t t he line a rm a ppin g$\ xi\r igh ta rrowK_+\ x i(t )$ is t he $S$-tr ans f ormof a Hida d istribution.Mo re precise ly , a ccordi n g to [@nu alartwhite], define the functio n $ $\lab el{C }
C(t)=|k '(t )|+\in t_0 ^ t \big g|\fra c{\pa rt ial K | exp(a|\xi|_p^{2}).$$
See [@kuo],_Theorem 8.6.
Limit_theorems for Volterra processes_{#volterra}
=====================================
One-dimensional case
--------------------
Consider_a_Volterra process_$B=(_B_t )_{t\geq 0}$_of the form_$$\label{star}
B_t=\int_0^t K(t,s)\,\mathrm{d}W_s,$$ where $K(t,s)$_satisfies $\int_0^tK(t,s)^2 \,\mathrm{d}s<\infty$_for_all $t>0$ and $W$ is the Brownian motion defined on the white noise probability_space_introduced in_the_last_section. Note that the $S$-transform_of the random variable $B_t$_is given_by $$\label{e1aa}
S(B_t)(\xi)=\int_0^{t}K(t,s)\xi(s)\,\mathrm{d}s$$ for any $\xi\in \mathcal{S}({\mathbb{R}})$. We_introduce_the following assumptions_on the kernel $K$:
1. $K$ is continuously differentiable_on $\{0<s<t<\infty \}$ and for any_$ t>0$, we_have_$$\int_0^t_ _\bigg|\frac{\partial K}{\partial t}(t,s) \bigg|(t-s)\,\mathrm{d}s <\infty;$$
2. _$k(t)=\int_0^{t}K(t,s)\,\mathrm{d}s $ is continuously differentiable on_$(0,\infty)$.
Consider the operator $K_+$ defined by $$K_+\xi(t)=k'(t)\xi(t)+\int_0^t_\frac{\partial K}{\partial t}(t,r)\bigl(\xi(r)-\xi(t)\bigr)\,\mathrm{d}r,$$ where $t>0$ and_$\xi\in \mathcal{S}({\mathbb{R}})$. From Theorem \[F\],_it follows_that the linear mapping $\xi\rightarrow_K_+\xi(t)$ is the_$S$-transform of_a Hida distribution._More precisely, according to [@nualartwhite], define_the function $$\label{C}
C(t)=|k'(t)|+\int_0^t_\bigg|\frac{\partial K |
within the supergravity inspired constrained minimal supersymmetric model. We perform a global statistical $\chi^2$ analysis of these data and show that the allowed region of parameter space is bounded from below by the Higgs limit, which depends on the trilinear coupling and from above by the anomalous magnetic moment $a_\mu$. The newest [$b \to X_s \gamma~ $]{}data deviate 1.7 $\sigma$ from recent SM calculations and prefer a similar parameter region as the 2.6 $\sigma$ deviation from $a_\mu$.]{}'
---
-1cm
IEKP-KA/2001-14\
[hep-ph/0106311]{}
[**A global fit to the anomalous magnetic moment, [$b \to X_s \gamma~ $]{} and Higgs limits in the constrained MSSM**]{}\
[**W. de Boer, M. Huber, C. Sander**]{}\
[*Institut für Experimentelle Kernphysik, University of Karlsruhe\
Postfach 6980, D-76128 Karlsruhe, Germany*]{}\
[**D.I. Kazakov**]{}\
[*Bogoliubov Laboratory of Theoretical Physics, Joint Institute for Nuclear Research,\
141 980 Dubna, Moscow Region, Russian Federation*]{}
Introduction
============
Recently a new measurement of the anomalous magnetic moment of the muon became available, which suggests a possible 2.6 standard deviation from the Standard Model (SM) expectation[@BNL]: $\Delta a_\mu=a_\mu^{exp}-a_\mu^{th}=(43\pm 16)\cdot 10^{-10}$. The theoretical prediction depends on the uncertainties in the vacuum polarization and the light-by-light scattering, see e.g. the discussion in [@Y]. However, even with a conservative estimate of the theoretical errors, one has a positive difference $\Delta a_\mu$ of the order of the weak contribution to the anomalous magnetic moment, which opens a window for “new physics”. The most popular explanation is given in the framework of SUSY theories [@CM]-[@baer], since the contribution of superpartners to the anomalous magnetic moment of the muon is of the order of the weak contribution and allows to explain the desired difference $\Delta a_\mu$. It requires the Higgs mixing parameter to be positive[@CN] and the sparticles contributing to the chargino-sneutrino $(\ | within the supergravity inspired constrained minimal supersymmetric model. We perform a ball-shaped statistical $ \chi^2 $ psychoanalysis of these datum and show that the allowed region of argument space is bounded from below by the Higgs terminus ad quem, which count on the trilinear coupling and from above by the anomalous charismatic here and now $ a_\mu$. The newest [ $ b \to X_s \gamma~ $ ] { } datum deviate 1.7 $ \sigma$ from recent SM calculations and prefer a exchangeable parameter region as the 2.6 $ \sigma$ deviation from $ a_\mu$. ] { }'
---
-1 cm
IEKP - KA/2001 - 14\
[ hep - ph/0106311 ] { }
[ * * A global fit to the anomalous charismatic moment, [ $ b \to X_s \gamma~ $ ] { } and Higgs limits in the constrained MSSM**]{}\
[ * * W. de Boer, M. Huber, C. Sander**]{}\
[ * Institut für Experimentelle Kernphysik, University of Karlsruhe\
Postfach 6980, D-76128 Karlsruhe, Germany*]{}\
[ * * D.I. Kazakov**]{}\
[ * Bogoliubov Laboratory of Theoretical Physics, Joint Institute for Nuclear Research,\
141 980 Dubna, Moscow Region, Russian Federation * ] { }
Introduction
= = = = = = = = = = = =
Recently a new measurement of the anomalous charismatic moment of the muon became available, which suggests a possible 2.6 standard deviation from the Standard Model (SM) expectation[@BNL ]: $ \Delta a_\mu = a_\mu^{exp}-a_\mu^{th}=(43\pm 16)\cdot 10^{-10}$. The theoretical prediction depend on the uncertainties in the vacuum polarization and the light - by - light scattering, witness e.g. the discussion in [ @Y ]. However, even with a conservative estimate of the theoretical errors, one has a positivist deviation $ \Delta a_\mu$ of the order of the weak contribution to the anomalous magnetic moment, which opens a window for “ newfangled purgative ”. The about popular explanation is given in the framework of SUSY theory [ @CM]-[@baer ], since the contribution of superpartners to the anomalous magnetic moment of the muon is of the order of the unaccented contribution and admit to explain the desired remainder $ \Delta a_\mu$. It requires the Higgs mixing parameter to be positive[@CN ] and the sparticles contributing to the chargino - sneutrino $ (\ | wihhin the supergravity inrpired constraiuwd minmmal sulersymmegric model. We perform a globel sratisupcal $\chi^2$ analysis of ghese datw and shiw tiat the allowed csgion on parzletex wpace is boundgd from belof by the Higgs lkmnt, which depends on the trilinear cotpling snf from above bi the wnomzlous magnetic moment $a_\mu$. The neweat [$b \tm X_s \gamma~ $]{}dsta deviate 1.7 $\sigma$ from refent SM calculations ajd prefer a simyoar parametef region as the 2.6 $\sigma$ deviation from $a_\mu$.]{}'
---
-1cm
IEKP-KA/2001-14\
[hep-pf/0106311]{}
[**A glpbal fit ti rhe dnomalous megnetib moment, [$b \to X_s \gakma~ $]{} anc Higgs limits in thw constrained MSSM**]{}\
[**W. dx Boer, M. Huber, C. Sanqer**]{}\
[*Institgt für Experimentellw Jernpvysin, Unkcerritg pf Karlsguhx\
Postfach 6980, S-76128 Karlsruhe, Germany*]{}\
[**D.I. Kazakov**]{}\
[*Bpgjoiubov Laborafory os Eheoretical Physics, Joint Institute for Nudlear Research,\
141 980 Dubna, Miscow Region, Russian Vederatiog*]{}
Introduction
============
Recently a new measurement of the anmmalons maykctic nolent of the muon became available, which suggeffs a possible 2.6 stakdard deviation frpm tng Standard Modgl (SM) erledtation[@BNL]: $\Delta a_\lu=a_\mu^{exk}-a_\mu^{th}=(43\pm 16)\cdot 10^{-10}$. Uhe tneoretical prediction depenes on the unbertqinties in the vacbum polarizacion amd thr light-by-light scatteriug, see e.g. the disfussion ih [@H]. However, even wkth a conservative estimate of ehe theorxticak errorr, ong has a positive fiffevance $\Delta a_\mu$ of hhe otder ox the weak contribution to the anomalous magnetic momenj, wvicv opens c windpw for “new phrsics”. The most populax explxnation is given mn the framerork of SUSY djeories [@CM]-[@baec], since tre cintrubution uf superpartnets to the anomalous magnetic moment on the juon is of the irdtr of the weak cpntfibttpon agg allows to axplxin yhe ddsired diffcrevce $\Celta a_\mu$. It requirev ths Higgs mixing parsmcter to bg positivq[@CN] and the slarticles contribuuing tm tie chatgigo-sneutrino $(\ | within the supergravity inspired constrained minimal supersymmetric perform global statistical analysis of these allowed of parameter space bounded from below the Higgs limit, which depends on trilinear coupling and from above by the anomalous magnetic moment $a_\mu$. The newest \to X_s \gamma~ $]{}data deviate 1.7 $\sigma$ from recent SM calculations and prefer similar region the $\sigma$ deviation from $a_\mu$.]{}' --- -1cm IEKP-KA/2001-14\ [hep-ph/0106311]{} [**A global fit to the anomalous magnetic moment, \to X_s \gamma~ $]{} and Higgs limits in constrained MSSM**]{}\ [**W. de M. Huber, C. Sander**]{}\ [*Institut Experimentelle University of Postfach D-76128 Germany*]{}\ [**D.I. Kazakov**]{}\ Laboratory of Theoretical Physics, Joint Institute for Nuclear Research,\ 141 980 Dubna, Moscow Region, Russian Federation*]{} Introduction Recently a of the magnetic of muon became available, a possible 2.6 standard deviation from (SM) expectation[@BNL]: $\Delta a_\mu=a_\mu^{exp}-a_\mu^{th}=(43\pm 16)\cdot 10^{-10}$. The theoretical depends on uncertainties in the vacuum polarization and light-by-light scattering, see e.g. the discussion in [@Y]. even with a conservative estimate of the theoretical errors, one has a positive difference $\Delta the order of the contribution to the magnetic which a for “new The most popular explanation is given in the framework of SUSY [@CM]-[@baer], since the contribution of superpartners to the anomalous magnetic the is of the of the weak contribution allows explain the desired difference It the to positive[@CN] the sparticles contributing to chargino-sneutrino $(\ | within the supergravity inspIred constrAined MinImaL sUperSymmEtric model. We peRForm A global statistical $\chi^2$ aNalysIs OF theSE dAta anD show thAT tHE AllOwEd RegIoN Of ParamEteR space iS bounded frOm bElOw by the Higgs LImIt, which depEndS on the trilinEar CoupliNg And FRom abOve By the AnomalOUs magnEtic momenT $a_\MU$. The neWEst [$b \to X_S \GAmMa~ $]{}daTa deviate 1.7 $\sigma$ froM ReCEnt SM calculatiOns and PrEFeR A SimIlaR parameter ReGion aS The 2.6 $\sigmA$ DeVIATioN From $a_\mu$.]{}'
---
-1cm
IEKP-kA/2001-14\
[hep-ph/0106311]{}
[**A gloBAl fIt to thE aNomALous maGnetiC mOMenT, [$b \to X_s \gamma~ $]{} And HIggs limitS in the COnstraiNEd MSSM**]{}\
[**W. De Boer, m. HuBer, c. SanDEr**]{}\
[*inStiTuT Für eXpEriMEntElle KernPhYsIk, UniVersITY OF KarLsrUhe\
POstfaCh 6980, D-76128 Karlsruhe, GErmAny*]{}\
[**D.i. kazAkov**]{}\
[*BOgoliUbov laBoratOry of THeoreTiCal Physics, Joint instItute for NUclEaR ReSeArch,\
141 980 DUBna, MosCow regIon, RussIan FedeRAtiOn*]{}
iNTRoDuction
============
Recently a neW mEASuRement of The anoMAlOuS Magnetic MoMenT of tHE Muon bEcamE AvAilable, wHich suGGeStS a possiBlE 2.6 standArD deViaTion fROm thE StandArd Model (sM) expECtation[@BNL]: $\DeltA A_\mu=a_\mu^{exp}-a_\mu^{tH}=(43\Pm 16)\CDOt 10^{-10}$. tHe thEorEtical prediCtioN DepeNds oN ThE unCErtaiNties In THe VAcuum polarization anD tHe lighT-by-liGht scattering, See e.g. the diSCUSsion in [@Y]. howeVEr, EVen with a conserVativE estimate oF The theorEticaL errors, oNe has a posITIve diffeRenCe $\DEltA a_\mU$ OF tHe order of the wEAK conTrIbution To tHe anomaLouS maGneTic MoMent, which Opens a wiNdOw FoR “nEw pHysicS”. the most pOpUlaR eXplAnatiON is givEn in tHe frAmEwORk oF SUSY thEOrIES [@CM]-[@bAeR], sInce The CoNtribUtioN Of sUperparTners to thE anOMaloUs MaGnetic mOment of the muoN iS of the ordeR oF thE weak cONTributioN and allows to explain the dESired diFfeRence $\deltA a_\mu$. It reqUirEs the HIggS Mixing ParameTer to Be PosITIve[@CN] AND tHe sPaRticles conTRIbuTing tO tHe chArgino-sNeutrino $(\ | within the supergravity i nspired co nstra ine d m in imal sup ersymmetric mo d el.We perform a global st atist ic a l $\ c hi ^2$ a nalysis of t hes eda taan d s how t hat the al lowed regi onof parameter s p ac e is bound edfrom below b y t he Hig gs li m it, w hic h dep ends o n the t rilinearco u plinga nd from a bo ve b y the anomalous m a gn e tic moment $a_ \mu$.Th e n e w est [$ b \to X_ s\gamm a ~ $]{}d a ta d e via t e 1.7 $\sigma $ from rece n t S M calc ul ati o ns and pref er a s imilar para mete r regionas the 2.6 $\s i gma$ de viatio n f rom $a_ \ mu $. ]{} '- ---1 cmIEK P-KA/200 1- 14 \
[he p-ph / 0 1 0 6311 ]{}
[* *A gl obal fit to t heanom a lou s mag netic mom en t, [$ b \to X_s \ gamma~ $]{} and Hig gs limits in t heco nstra i ned MS SM* *]{ }\
[**W . de Bo e r,M. H u be r, C. Sander**]{}\
[ * I ns titut fü r Expe r im en t elle Ker np hys ik,U n ivers ityo fKarlsruh e\
Pos t fa ch 6980,D- 76128Ka rls ruh e, Ge r many *]{}\[**D.I.Kazak o v**]{}\
[*Bogo l iubov Laborat o ry o fT heor eti cal Physics , Jo i nt I nsti t ut e f o r Nuc learRe s ea r ch,\
141 980 Dubna, M oscowRegio n, Russian Fe deration*] { }
Introdu ctio n
= = ==========
Re centl y a new me a surement of t he anoma lous magn e t ic momen t o f t hemuo n be came availabl e , whi ch sugges tsa possi ble 2. 6 s tan da rd deviat ion from t he S ta nda rd Mo d el (SM)ex pec ta tio n[@BN L ]: $\D eltaa_\m u= a_ \ mu^ {exp}-a _ \m u ^ {th} =( 43 \pm16) \c dot 1 0^{- 1 0}$ . The t heoretica l p r edic ti on depend s on the unce rt ainties in t hevacuum p olarizat ion and the light-by-li g ht scat ter ing,seee.g. thedis cussio n i n [@Y]. Howev er, e ve n w i t h a c o n se rva ti ve estimat e ofthe t he oret ical er rors, one has a po s iti ve difference $\ Delt a a_ \mu $ o f th eo rde r of the weak con tributionto th e anomalou s ma gn etic mo ment, w hicho pens awindow fo r “new ph ys ics” . The most popu lar expl anation i s give n i n the fr amewor kofSUSYtheori e s [ @CM]- [@baer ], since theco ntributi on of superpartners tothe an omalo usmagneticmom e ntof the mu on i s of the o rde r o f the we a k con trib u ti ona nd al lows to explai n t hed e si red differe n c e $\ Delta a_ \ mu$. I t re quires the Higgsm ixing paramete r to b e p osi t ive[ @C N] and the spa rti cl e s contrib ut ing to thechargino -s n eutri no $(\ | within_the supergravity_inspired constrained minimal supersymmetric_model. We_perform_a global_statistical_$\chi^2$ analysis of_these data and_show that the allowed_region of parameter_space_is bounded from below by the Higgs limit, which depends on the trilinear coupling_and_from above_by_the_anomalous magnetic moment $a_\mu$. The_newest [$b \to _X_s \gamma~_$]{}data deviate 1.7 $\sigma$ from recent SM calculations_and_prefer a similar_parameter region as the 2.6 $\sigma$ deviation from $a_\mu$.]{}'
---
-1cm
IEKP-KA/2001-14\
[hep-ph/0106311]{}
[**A_global fit to the anomalous magnetic_moment, [$b _\to__X_s \gamma~ $]{} and_Higgs limits in the constrained MSSM**]{}\
[**W._de Boer, M. Huber, C. Sander**]{}\
[*Institut_für Experimentelle Kernphysik, University of Karlsruhe\
Postfach 6980,_D-76128 Karlsruhe, Germany*]{}\
[**D.I. Kazakov**]{}\
[*Bogoliubov Laboratory of_Theoretical Physics, Joint Institute for_Nuclear Research,\
141_980 Dubna, Moscow Region, Russian_Federation*]{}
Introduction
============
Recently a new_measurement of_the anomalous magnetic_moment of the muon became available,_which suggests a_possible 2.6 standard deviation from the_Standard_Model (SM) expectation[@BNL]:_$\Delta_a_\mu=a_\mu^{exp}-a_\mu^{th}=(43\pm_16)\cdot 10^{-10}$._The theoretical prediction_depends_on the_uncertainties_in the vacuum polarization and the_light-by-light_scattering, see e.g. the discussion in [@Y]. However,_even with a conservative_estimate_of the theoretical errors,_one has a positive difference_$\Delta a_\mu$ of the order of_the weak_contribution to_the anomalous magnetic moment, which opens a window for “new physics”._The most popular explanation is given_in the framework of_SUSY theories [@CM]-[@baer],_since_the contribution of_superpartners_to the_anomalous magnetic moment of the muon is_of the_order of the weak contribution and_allows to explain the_desired_difference $\Delta a_\mu$. It requires the_Higgs mixing parameter to be positive[@CN]_and the sparticles contributing to_the_chargino-sneutrino_$(\ |
=1,2$. Assume that the numbers $ b_1(3), b_2(3)$ satisfy the condition $b_1(3) b_2(3) \ge 0$. We attach ${\Delta}_{b_1}$ to ${\Delta}_{b_2}$ by identification of the vertices $(s_1)_+$ and $(s_2)_-$ and obtain thereby a disk diagram ${\Delta}'_b$ such that $ {\partial}{\Delta}'_b = s_1 s_2 (r_1 r_2)^{-1}$ and $|{\Delta}'_b(2) | = |{\Delta}_{b_1}(2) | + |{\Delta}_{b_2}(2) |$, see Fig. 6.10(a). Note that ${\varphi}( a(c) ) \equiv {\varphi}( s_1 s_2 )$, $r_1 r_2$ is a simple path and ${\varphi}(r_1 r_2 ) \equiv a_1^{b_1(3) +b_2(3)}$. Hence, ${\Delta}'_b$ is a desired disk diagram ${\Delta}_{c}$ for $c$.
(-20,-2) arc (0:180:2); (-24,-2) node (v1) – (-20,-2); (-23,-2) –(-22,-2); (-22,0) –(-21.95,0); (-24,-2) \[fill = black\] circle (.05); (-20,-2) \[fill = black\] circle (.05); at (-22,-1) [$\Delta_{b_2}$]{}; at (-22,0.5) [$s_2$]{}; at (-22,-2.7) [$r_2$]{}; (-24,-2) arc (0:180:2); (-28,-2) node (v1) – (-24,-2); (-27,-2) –(-26,-2); (-26,0) –(-25.95,0); (-28,-2) \[fill = black\] circle (.05); (-24,-2) \[fill = black\] circle (.05); at (-26,-1) [$\Delta_{b_1}$]{}; at (-26,0.5) [$s_1$]{}; at (-24,0) [$\Delta'_b$]{}; at (-26,-2.7) [$r_1$]{}; at (-24,-4 | = 1,2$. Assume that the numbers $ b_1(3), b_2(3)$ satisfy the condition $ b_1(3) b_2(3) \ge 0$. We bind $ { \Delta}_{b_1}$ to $ { \Delta}_{b_2}$ by designation of the vertices $ (s_1)_+$ and $ (s_2)_-$ and obtain thereby a disk diagram $ { \Delta}'_b$ such that $ { \partial}{\Delta}'_b = s_1 s_2 (r_1 r_2)^{-1}$ and $ |{\Delta}'_b(2) | = |{\Delta}_{b_1}(2) | + |{\Delta}_{b_2}(2) |$, witness Fig. 6.10(a). Note that $ { \varphi } (a(c) ) \equiv { \varphi } ( s_1 s_2) $, $ r_1 r_2 $ is a simple way and $ { \varphi}(r_1 r_2) \equiv a_1^{b_1(3) + b_2(3)}$. Hence, $ { \Delta}'_b$ is a desired disk diagram $ { \Delta}_{c}$ for $ c$.
(-20,-2) discharge (0:180:2); (-24,-2) node (v1) – (-20,-2); (-23,-2) – (-22,-2); (-22,0) – (-21.95,0); (-24,-2) \[fill = black\ ] circle (.05); (-20,-2) \[fill = black\ ] circle (.05); at (-22,-1) [ $ \Delta_{b_2}$ ] { }; at (-22,0.5) [ $ s_2 $ ] { }; at (-22,-2.7) [ $ r_2 $ ] { }; (-24,-2) arc (0:180:2); (-28,-2) node (v1) – (-24,-2); (-27,-2) – (-26,-2); (-26,0) – (-25.95,0); (-28,-2) \[fill = black\ ] circle (.05); (-24,-2) \[fill = black\ ] circle (.05); at (-26,-1) [ $ \Delta_{b_1}$ ] { }; at (-26,0.5) [ $ s_1 $ ] { }; at (-24,0) [ $ \Delta'_b$ ] { }; at (-26,-2.7) [ $ r_1 $ ] { }; at (-24,-4 | =1,2$. Asdume that the numbers $ b_1(3), b_2(3)$ satisfy the einditimn $b_1(3) b_2(3) \ge 0$. We xttach ${\Delta}_{b_1}$ to ${\Delta}_{b_2}$ by iventuficaupon of the vertices $(s_1)_+$ and $(s_2)_-$ anf obtain theceby a disk diagczm ${\Delta}'_b$ sucg thac $ {\partial}{\Delta}'_b = s_1 s_2 (r_1 r_2)^{-1}$ and $|{\Delta}'_b(2) | = |{\Ddlca}_{b_1}(2) | + |{\Delta}_{b_2}(2) |$, see Fig. 6.10(a). Note that ${\var[hi}( a(c) ) \eeuiv {\varphi}( f_1 s_2 )$, $w_1 r_2$ js a simple path and ${\varphi}(r_1 r_2 ) \eqhiv a_1^{b_1(3) +b_2(3)}$. Hence, ${\Delts}'_b$ is a desired disk diagrwm ${\Dflta}_{c}$ for $c$.
(-20,-2) arc (0:180:2); (-24,-2) jode (v1) – (-20,-2); (-23,-2) –(-22,-2); (-22,0) –(-21.95,0); (-24,-2) \[fyol = black\] cifcle (.05); (-20,-2) \[fill = black\] cirdle (.05); at (-22,-1) [$\Delta_{b_2}$]{}; at (-22,0.5) [$s_2$]{}; at (-22,-2.7) [$r_2$]{}; (-24,-2) arz (0:180:2); (-28,-2) npde (v1) – (-24,-2); (-27,-2) –(-26,-2); (-26,0) –(-25.95,0); (-28,-2) \[fipn = black\] ciccle (.05); (-24,-2) \[fill = black\] circle (.05); at (-26,-1) [$\Dekta_{b_1}$]{}; at (-26,0.5) [$s_1$]{}; at (-24,0) [$\Denta'_v$]{}; at (-26,-2.7) [$r_1$]{}; at (-24,-4 | =1,2$. Assume that the numbers $ b_1(3), the $b_1(3) b_2(3) 0$. We attach of vertices $(s_1)_+$ and and obtain thereby disk diagram ${\Delta}'_b$ such that $ = s_1 s_2 (r_1 r_2)^{-1}$ and $|{\Delta}'_b(2) | = |{\Delta}_{b_1}(2) | + |{\Delta}_{b_2}(2) see Fig. 6.10(a). Note that ${\varphi}( a(c) ) \equiv {\varphi}( s_1 s_2 )$, r_2$ a path ${\varphi}(r_1 r_2 ) \equiv a_1^{b_1(3) +b_2(3)}$. Hence, ${\Delta}'_b$ is a desired disk diagram ${\Delta}_{c}$ for $c$. arc (0:180:2); (-24,-2) node (v1) – (-20,-2); (-23,-2) (-22,0) –(-21.95,0); (-24,-2) \[fill black\] circle (.05); (-20,-2) \[fill black\] (.05); at [$\Delta_{b_2}$]{}; (-22,0.5) at (-22,-2.7) [$r_2$]{}; arc (0:180:2); (-28,-2) node (v1) – (-24,-2); (-27,-2) –(-26,-2); (-26,0) –(-25.95,0); (-28,-2) \[fill = black\] circle (.05); \[fill = (.05); at [$\Delta_{b_1}$]{}; (-26,0.5) at (-24,0) [$\Delta'_b$]{}; [$r_1$]{}; at (-24,-4 | =1,2$. Assume that the numbers $ b_1(3), b_2(3)$ satIsfy the conDitioN $b_1(3) b_2(3) \Ge 0$. WE aTtacH ${\DelTa}_{b_1}$ to ${\Delta}_{b_2}$ by iDEntiFication of the vertices $(s_1)_+$ And $(s_2)_-$ aNd OBtaiN ThEreby A disk diAGrAM ${\delTa}'_B$ sUch ThAT $ {\pArtiaL}{\DeLta}'_b = s_1 s_2 (r_1 R_2)^{-1}$ and $|{\Delta}'_b(2) | = |{\delTa}_{B_1}(2) | + |{\Delta}_{b_2}(2) |$, see FiG. 6.10(A). NOte that ${\varPhi}( A(c) ) \equiv {\varphI}( s_1 s_2 )$, $R_1 r_2$ is a sImPle PAth anD ${\vaRphi}(r_1 R_2 ) \equiv A_1^{B_1(3) +b_2(3)}$. HencE, ${\Delta}'_b$ is A dESired dISk diagrAM ${\deLta}_{c}$ For $c$.
(-20,-2) arc (0:180:2); (-24,-2) node (v1) – (-20,-2); (-23,-2) –(-22,-2); (-22,0) –(-21.95,0); (-24,-2) \[fill = bLAcK\] Circle (.05); (-20,-2) \[fill = blacK\] circlE (.05); aT (-22,-1) [$\deLTA_{b_2}$]{}; aT (-22,0.5) [$s_2$]{}; aT (-22,-2.7) [$r_2$]{}; (-24,-2) arc (0:180:2); (-28,-2) node (v1) – (-24,-2); (-27,-2) –(-26,-2); (-26,0) –(-25.95,0); (-28,-2) \[FiLl = blaCK\] circle (.05); (-24,-2) \[FIlL = BLAck\] CIrcle (.05); at (-26,-1) [$\Delta_{b_1}$]{}; At (-26,0.5) [$s_1$]{}; at (-24,0) [$\Delta'_b$]{}; AT (-26,-2.7) [$r_1$]{}; aT (-24,-4 | =1,2$. Assume that the num bers $ b_1 (3),b_2 (3) $sati sfythe condition$ b_1( 3) b_2(3) \ge 0$. We a ttach $ { \Del t a} _{b_1 }$ to $ { \D e l ta} _{ b_ 2}$ b y i denti fic ation o f the vert ice s$(s_1)_+$ an d $ (s_2)_-$ a ndobtain there bya disk d iag r am ${ \De lta}' _b$ su c h that $ {\part ia l }{\Del t a}'_b = s _1 s_2 (r_1 r_2)^{-1}$ an d $|{\Delta}'_b (2) |=|{ \ D elt a}_ {b_1}(2) | + |{\D e lta}_{b _ 2} ( 2 ) |$ , see Fig. 6.1 0(a). Notet hat ${\va rp hi} ( a(c)) \eq ui v {\ varphi}( s_1s_2 )$, $ r_1 r_ 2 $ is as imple p ath an d $ {\v arph i }( r_ 1 r _2 ) \ e qu iva_1 ^{b_1(3) + b_ 2(3)} $. H e n c e , ${ \De lta} '_b$is a desireddis k di a gra m ${\ Delta }_{c }$ for$c$.
(-20, -2 ) arc (0:180:2) ; (- 24,-2) no de(v 1)–(-20, - 2); (- 23, -2) –(-22, -2); (- 2 2,0 )– ( - 21 .95,0); (-24,-2) \ [f i l l= black\ ] circ l e(. 0 5); (-20 ,- 2)\[fi l l = bl ack\ ] c ircle (. 05); a t ( -2 2,-1) [ $\ Delta_ {b _2} $]{ }; at (-22 ,0.5)[$s_2$]{ }; at (-22,-2.7) [$r _ 2$]{}; (-24,- 2 )a r c( 0:18 0:2 ); (-28,-2) nod e (v1 ) –( -2 4,- 2 ); (- 27,-2 )– (- 2 6,-2); (-26,0) –(-2 5. 95,0); (-28 ,-2) \[fill = black\] c i r c le (.05) ; (- 2 4, - 2) \[fill = bl ack\] circle (. 0 5); at ( -26,- 1) [$\De lta_{b_1} $ ] {}; at ( -26 ,0. 5)[$s _ 1 $] {}; at (-24,0 ) [$\D el ta'_b$] {}; at (-2 6,- 2.7 ) [ $r_ 1$ ]{}; at ( -24,-4 | =1,2$. Assume_that the_numbers $ b_1(3), b_2(3)$_satisfy the_condition_$b_1(3) b_2(3)_\ge_0$. We attach_${\Delta}_{b_1}$ to ${\Delta}_{b_2}$_by identification of the_vertices $(s_1)_+$ and_$(s_2)_-$_and obtain thereby a disk diagram ${\Delta}'_b$ such that $ {\partial}{\Delta}'_b = s_1 s_2__(r_1 r_2)^{-1}$_and_$|{\Delta}'_b(2)_| = |{\Delta}_{b_1}(2) |_+ |{\Delta}_{b_2}(2) |$, see Fig. 6.10(a)._Note that_${\varphi}( a(c) ) \equiv {\varphi}( s_1_s_2_)$, $r_1 r_2$_is a simple path and ${\varphi}(r_1 r_2 ) \equiv_ a_1^{b_1(3) +b_2(3)}$. Hence, ${\Delta}'_b$ is_a desired disk_diagram_${\Delta}_{c}$_for $c$.
(-20,-2) arc (0:180:2);_(-24,-2) node (v1) – (-20,-2); (-23,-2)_–(-22,-2); (-22,0) –(-21.95,0); (-24,-2) \[fill =_black\] circle (.05); (-20,-2) \[fill = black\]_circle (.05); at (-22,-1) [$\Delta_{b_2}$]{}; at_(-22,0.5) [$s_2$]{}; at (-22,-2.7) [$r_2$]{};_(-24,-2) arc_(0:180:2); (-28,-2) node (v1) –_(-24,-2); (-27,-2) –(-26,-2);_(-26,0) –(-25.95,0);_(-28,-2) \[fill =_black\] circle (.05); (-24,-2) \[fill =_black\] circle (.05);_at (-26,-1) [$\Delta_{b_1}$]{}; at (-26,0.5) [$s_1$]{};_at_(-24,0) [$\Delta'_b$]{}; at_(-26,-2.7)_[$r_1$]{};_at (-24,-4 |
.*]{}, hep-ex/0408065.
D. Atwood, M. Gronau and A. Soni, Phys. Rev. Lett. [**79**]{}, 185 (1997). D. Atwood, T. Gershon, M. Hazumi and A. Soni, hep-ph/0410036.
S. Kurokawa and E. Kikutani [*et al.*]{}, Nucl. Instr. and Meth. A [**499**]{}, 1 (2003).
Belle Collaboration, A. Abashian [*et al.*]{}, Nucl. Instr. and Meth. A [**479**]{}, 117 (2002).
Y. Ushiroda (Belle SVD2 Group), Nucl. Instr. and Meth. A [**511**]{}, 6 (2003).
Belle Collaboration, K. Abe [*et al.*]{}, Phys. Rev. Lett. [**87**]{}, 101801 (2001). ARGUS Collaboration, H. Albrecht *et al.*, Phys. Lett. B [**241**]{}, 278 (1990).
Belle Collaboration, P. Koppenburg [*et al.*]{}, Phys. Rev. Lett. [**93**]{}, 061803 (2004). Belle Collaboration, K. Abe [*et al.*]{}, Phys. Rev. Lett. [**91**]{}, 261801 (2003). Possible $\fcp$ states include $\ks\piz\gamma$ and $\kstarz\gamma$ hereafter.
H. Kakuno, K. Hara [*et al.*]{}, Nucl. Instr. and Meth. A [**533**]{}, 516 (2004). H. Tajima [*et al.*]{}, Nucl. Instr. and Meth. A [**533**]{}, 307 (2004). S. Eidelman [*et al.*]{}, Phys. Lett. B [**592**]{}, 1 (2004).
O. Long, M. Baak, R. N. Cahn and D. Kirkby, Phys. Rev. D [**68**]{}, 034010 (2003).
$l$ $r$ interval $\epsilon_l$
----- --------------- ------------------ ----------------- ------------------ -----------------
1 0.000 – 0.250 $0.397\pm 0.015$ $0.464\pm0.007$ $ | . * ] { }, hep - ex/0408065.
D. Atwood, M. Gronau and A. Soni, Phys. Rev. Lett. [ * * 79 * * ] { }, 185 (1997). D. Atwood, T. Gershon, M. Hazumi and A. Soni, hep - ph/0410036.
S. Kurokawa and E. Kikutani [ * et al. * ] { }, Nucl. Instr. and Meth. A [ * * 499 * * ] { }, 1 (2003).
Belle Collaboration, A. Abashian [ * et al. * ] { }, Nucl. Instr. and Meth. A [ * * 479 * * ] { }, 117 (2002).
Y. Ushiroda (Belle SVD2 Group), Nucl. Instr. and Meth. A [ * * 511 * * ] { }, 6 (2003).
Belle Collaboration, K. Abe [ * et al. * ] { }, Phys. Rev. Lett. [ * * 87 * * ] { }, 101801 (2001). ARGUS Collaboration, H. Albrecht * et al. *, Phys. Lett. B [ * * 241 * * ] { }, 278 (1990).
Belle Collaboration, P. Koppenburg [ * et al. * ] { }, Phys. Rev. Lett. [ * * 93 * * ] { }, 061803 (2004). Belle Collaboration, K. Abe [ * et al. * ] { }, Phys. Rev. Lett. [ * * 91 * * ] { }, 261801 (2003). Possible $ \fcp$ states include $ \ks\piz\gamma$ and $ \kstarz\gamma$ hereafter.
H. Kakuno, K. Hara [ * et al. * ] { }, Nucl. Instr. and Meth. A [ * * 533 * * ] { }, 516 (2004). H. Tajima [ * et al. * ] { }, Nucl. Instr. and Meth. A [ * * 533 * * ] { }, 307 (2004). S. Eidelman [ * et al. * ] { }, Phys. Lett. boron [ * * 592 * * ] { }, 1 (2004).
O. Long, M. Baak, R. N. Cahn and D. Kirkby, Phys. Rev. D [ * * 68 * * ] { }, 034010 (2003).
$ l$ $ r$ time interval $ \epsilon_l$
----- --------------- ------------------ ----------------- ------------------ -----------------
1 0.000 – 0.250 $ 0.397\pm 0.015 $ $ 0.464\pm0.007 $ $ | .*]{}, heo-ex/0408065.
D. Atwood, M. Gronau and A. Roni, Phys. Rev. Letj. [**79**]{}, 185 (1997). D. Atxood, T. Gsrshon, M. Fazumi and A. Soni, hep-ph/0410036.
S. Kurokewa qnd E. Jikutani [*et al.*]{}, Nucl. Inrtr. and Mvth. A [**499**]{}, 1 (2003).
Bwlle Xollaboratmkn, A. Abashian [*sb al.*]{}, Uurl. Instr. and Mejh. A [**479**]{}, 117 (2002).
Y. Ushhroda (Belle SVG2 Erlup), Nucl. Instr. and Meth. A [**511**]{}, 6 (2003).
Belle Cjllaborstlon, K. Abe [*et al.*]{}, Phyx. Wev. Lsnt. [**87**]{}, 101801 (2001). ARGUS Collaboration, H. Albrscht *et al.*, Phys. Lett. B [**241**]{}, 278 (1990).
Belle Collaboration, P. Noppfnburg [*et al.*]{}, Phys. Rfv. Lett. [**93**]{}, 061803 (2004). Bgmle Xollaboratiov, K. Abe [*et al.*]{}, Phys. Rev. Lejt. [**91**]{}, 261801 (2003). Possible $\fcp$ states includd $\ks\pnz\gamma$ and $\kwtage\gamma$ hereahter.
H. Kwkuno, K. Hara [*cn al.*]{}, Nuwl. Instt. and Meth. A [**533**]{}, 516 (2004). H. Vajina [*et al.*]{}, Nucl. Instr. aid Meth. A [**533**]{}, 307 (2004). S. Eidelmwn [*et al.*]{}, [hvs. Lett. B [**592**]{}, 1 (2004).
O. Long, M. Baqk, R. N. Cavn atd D. Yurkcy, Khyx. Rsv. D [**68**]{}, 034010 (2003).
$m$ $r$ intedval $\epsilon_l$
----- --------------- ------------------ ----------------- ------------------ -----------------
1 0.000 – 0.250 $0.397\pm 0.015$ $0.464\pm0.007$ $ | .*]{}, hep-ex/0408065. D. Atwood, M. Gronau and Phys. Lett. [**79**]{}, (1997). D. Atwood, A. hep-ph/0410036. S. Kurokawa E. Kikutani [*et Nucl. Instr. and Meth. A [**499**]{}, (2003). Belle Collaboration, A. Abashian [*et al.*]{}, Nucl. Instr. and Meth. A [**479**]{}, (2002). Y. Ushiroda (Belle SVD2 Group), Nucl. Instr. and Meth. A [**511**]{}, 6 Belle K. [*et Phys. Rev. Lett. [**87**]{}, 101801 (2001). ARGUS Collaboration, H. Albrecht *et al.*, Phys. Lett. B [**241**]{}, (1990). Belle Collaboration, P. Koppenburg [*et al.*]{}, Phys. Lett. [**93**]{}, 061803 (2004). Collaboration, K. Abe [*et al.*]{}, Rev. [**91**]{}, 261801 Possible states $\ks\piz\gamma$ and $\kstarz\gamma$ H. Kakuno, K. Hara [*et al.*]{}, Nucl. Instr. and Meth. A [**533**]{}, 516 (2004). H. Tajima [*et Nucl. Instr. A [**533**]{}, (2004). Eidelman al.*]{}, Phys. Lett. 1 (2004). O. Long, M. Baak, and D. Kirkby, Phys. Rev. D [**68**]{}, 034010 $l$ $r$ $\epsilon_l$ ----- --------------- ------------------ ----------------- ------------------ 1 0.000 – 0.250 $0.397\pm 0.015$ $0.464\pm0.007$ $ | .*]{}, hep-ex/0408065.
D. Atwood, M. Gronau and A. SoNi, Phys. Rev. LEtt. [**79**]{}, 185 (1997). D. ATwoOd, T. geRshoN, M. HaZumi and A. Soni, heP-Ph/0410036.
S. KUrokawa and E. Kikutani [*et aL.*]{}, Nucl. inSTr. anD meTh. A [**499**]{}, 1 (2003).
BeLle CollABoRATioN, A. abAshIaN [*Et Al.*]{}, NucL. InStr. and MEth. A [**479**]{}, 117 (2002).
Y. UshirOda (beLle SVD2 Group), NUCl. instr. and MeTh. A [**511**]{}, 6 (2003).
belle CollaboRatIon, K. AbE [*eT al.*]{}, pHys. ReV. LeTt. [**87**]{}, 101801 (2001). ARGuS CollABoratiOn, H. AlbrecHt *ET al.*, PhyS. lett. B [**241**]{}, 278 (1990).
BeLLE COllaBoration, P. KoppenbuRG [*eT Al.*]{}, Phys. Rev. Lett. [**93**]{}, 061803 (2004). BElle CoLlABoRATioN, K. ABe [*et al.*]{}, Phys. reV. Lett. [**91**]{}, 261801 (2003). pOssible $\FCp$ STATes INclude $\ks\piz\gaMma$ and $\kstarZ\GamMa$ hereAfTer.
h. kakuno, k. Hara [*Et AL.*]{}, NuCl. Instr. and MEth. A [**533**]{}, 516 (2004). h. Tajima [*et Al.*]{}, Nucl. iNstr. and mEth. A [**533**]{}, 307 (2004). S. EiDelman [*Et aL.*]{}, PhYs. LeTT. B [**592**]{}, 1 (2004).
o. LOng, m. BAAk, R. n. caHn aND D. KIrkby, PhyS. REv. d [**68**]{}, 034010 (2003).
$l$ $r$ inTervAL $\EPSiloN_l$
----- --------------- ------------------ ----------------- ------------------ -----------------
1 0.000 – 0.250 $0.397\pM 0.015$ $0.464\pm0.007$ $ | .*]{}, hep-ex/0408065.
D. Atwood, M . Gro nau an dA. S oni, Phys. Rev. Le t t. [**79**]{}, 185 (1997) . D.At w ood, T. Gers hon, M. Ha z u mian dA.So n i, hep- ph/ 0410036 .
S. Kuro kaw aand E. Kikut a ni [*et al.* ]{} , Nucl. Inst r.and Me th . A [**49 9** ]{},1 (200 3 ).
Be lle Colla bo r ation, A. Abas h i an [*e t al.*]{}, Nucl.I ns t r. and Meth. A [**47 9* * ]{ } , 11 7 ( 2002).
Y. U shiro d a (Bell e S V D 2 Gr o up), Nucl. In str. and Me t h.A [**5 11 **] { }, 6 ( 2003) .Bel le Collabor atio n, K. Abe [*eta l.*]{}, Phys. R ev. Le tt. [ **87 * *] {} , 1 01 8 01( 20 01) . AR GUS Coll ab or ation , H. A l b rech t * et a l.*,Phys. Lett. B [* *241 * *]{ }, 27 8 (19 90).
Belle Colla borat io n, P. Koppenbur g [* et al.*]{ },Ph ys. R ev. L e tt. [ **9 3** ]{}, 06 1803 (2 0 04) .B e l le Collaboration, K. A b e [ *et al.* ]{}, P h ys .R ev. Lett . [* *91* * ] {}, 2 6180 1 ( 2003). P ossibl e $ \f cp$ sta te s incl ud e $ \ks \piz\ g amma $ and$\kstarz \gamm a $ hereafter.
H . Kakuno, K.H ar a [* e t al .*] {}, Nucl. I nstr . and Met h .A [ * *533* *]{}, 5 1 6( 2004). H. Tajima [* et al.*] {}, N ucl. Instr. a nd Meth. A [ * *533**]{ }, 3 0 7( 2004). S. Eide lman[*et al.*] { }, Phys. Lett . B [**5 92**]{},1 (2004).
O. Lo ng, M. B aa k, R. N. Cahn a nd D .Kirkby, Ph ys. Rev . D [* *68 **] {} , 034010(2003).
$l $ $ r$ in t erval $\ ep sil on_l$ ---- - - -- ----- ---- - --------- --------- --- ---- -- -- ------- - ----------- -- ----- ---- -- --- ------ - -
1 0.000 – 0.250 $0.39 7 \pm 0.0 15$ $0. 464\ pm0.007$ $ | .*]{}, hep-ex/0408065.
D. Atwood,_M. Gronau and_A. Soni, Phys. Rev. Lett. [**79**]{}, 185_(1997). D. Atwood,_T. Gershon,_M. Hazumi and_A. Soni,_hep-ph/0410036.
S. Kurokawa and E. Kikutani_[*et al.*]{}, Nucl._Instr. and Meth. A_[**499**]{}, 1 (2003).
Belle_Collaboration,_A. Abashian [*et al.*]{}, Nucl. Instr. and Meth. A [**479**]{}, 117 (2002).
Y. Ushiroda (Belle SVD2_Group),_Nucl. Instr._and_Meth._A [**511**]{}, 6 (2003).
Belle Collaboration,_K. Abe [*et al.*]{}, Phys. Rev. Lett. [**87**]{},_101801 (2001)._ARGUS Collaboration, H. Albrecht *et al.*, Phys. Lett. B_[**241**]{},_278 (1990).
Belle Collaboration,_P. Koppenburg [*et al.*]{}, Phys. Rev. Lett. [**93**]{}, 061803 (2004). Belle Collaboration,_K. Abe [*et al.*]{}, Phys. Rev. Lett. [**91**]{}, 261801_(2003). Possible $\fcp$_states_include_$\ks\piz\gamma$ and $\kstarz\gamma$ hereafter.
H. Kakuno,_K. Hara [*et al.*]{}, Nucl. Instr. and_Meth. A [**533**]{}, 516 (2004). H. Tajima_[*et al.*]{}, Nucl. Instr. and Meth. A_[**533**]{}, 307 (2004). S. Eidelman [*et al.*]{},_Phys. Lett. B [**592**]{}, 1_(2004).
O. Long, M. Baak,_R. N. Cahn and D. Kirkby, Phys. Rev. D [**68**]{},_034010 (2003).
_ $l$_ $r$_interval _$\epsilon_l$ _ __ ___ _ __ __ __ _
__----- --------------- ------------------ -----------------_------------------ -----------------
_1 0.000 –_0.250 _ $0.397\pm_0.015$ $0.464\pm0.007$ $ |
$ vertices of the first type of $\Gamma$ and hitting the origin: by the previous arguments, we know that $p\leq (n-1)+1=n$, because the vertex with an edge to $\infty$ has an additional edge, which may or may not hit the origin. The polynomial degree of the differential operator associated to $\Gamma$ equals $n-(2n-1-p)=-n+1+p$, which must be greater or equal than $0$: this forces immediately $p\geq n-1$. This fact, combined with the previous condition on $p$, yields that either $p=n$ or $p=n-1$.
In the case $p=n$, the corresponding differential operator has polynomial degree $1$, and from each vertex of the first type departs exactly one edge to the origin: this implies that the graph $\Gamma$ must have a vertex of the form
\
\
Of course, the vertex from which departs the arrow to $\infty$ may be also isolated, [*i.e.*]{} no edge has it as the endpoint: in this case, as $n\geq 2$, standard dimensional arguments imply that the corresponding weight is trivial. If it is not isolated, the property $[\mathfrak k_0,\mathfrak p_0]\subset \mathfrak p_0$ implies that the two consecutive edges correspond to propagators of type $(+,+)$, and Lemma \[l-vanish-4\] yields triviality of the corresponding weight.
We now consider the case $p=n-1$: the corresponding differential operator has polynomial degree $0$, [*i.e.*]{} it is a translation-invariant differential operator. We first assume that the vertex from which departs an edge to $\infty$ has the other edge hitting the origin: the fact that the differential operator has constant coefficients implies that this vertex cannot be isolated, otherwise Lemma \[l-vanish-4\] would imply triviality of its weight. Therefore, $\Gamma$ must be a disjoint union of wheel-like graphs, one of which must look like as follows:
\
\
Lemma \[l-vanish-4\] implies that the two edges meeting at the vertex with the edge to $\infty$ must have distinct colors, otherwise the corresponding weight vanishes. Therefore, the relations $[\mathfrak k_0,\mathfrak k_0]\subseteq \mathfrak k_0$, $[\mathfrak k_0,\mathfrak p_0 | $ vertices of the first type of $ \Gamma$ and hitting the beginning: by the former arguments, we know that $ p\leq (n-1)+1 = n$, because the vertex with an boundary to $ \infty$ has an extra edge, which may or may not hit the lineage. The polynomial academic degree of the differential operator associated to $ \Gamma$ equals $ n-(2n-1 - p)=-n+1+p$, which must be greater or equal than $ 0 $: this force immediately $ p\geq n-1$. This fact, combined with the previous condition on $ p$, yields that either $ p = n$ or $ p = n-1$.
In the event $ p = n$, the corresponding differential operator receive polynomial degree $ 1 $, and from each vertex of the first type departs precisely one edge to the origin: this implies that the graph $ \Gamma$ must have a vertex of the form
\
\
Of naturally, the vertex from which departs the arrow to $ \infty$ may be also isolated, [ * i.e. * ] { } no edge has it as the endpoint: in this shell, as $ n\geq 2 $, standard dimensional arguments imply that the corresponding weight is trivial. If it is not isolated, the property $ [ \mathfrak k_0,\mathfrak p_0]\subset \mathfrak p_0 $ implies that the two consecutive edges match to propagators of type $ (+, +) $, and Lemma \[l - vanish-4\ ] concede pettiness of the corresponding weight.
We now study the shell $ p = n-1 $: the corresponding differential hustler has polynomial degree $ 0 $, [ * i.e. * ] { } it is a translation - invariant differential operator. We first assume that the vertex from which departs an edge to $ \infty$ has the other edge hit the origin: the fact that the differential hustler has changeless coefficients implies that this vertex cannot be isolated, otherwise Lemma \[l - vanish-4\ ] would entail triviality of its weight. Therefore, $ \Gamma$ must be a disjoint coupling of wheel - like graphs, one of which must look like as trace:
\
\
Lemma \[l - vanish-4\ ] entail that the two edges meet at the vertex with the edge to $ \infty$ must suffer distinct semblance, otherwise the comparable weight vanishes. Therefore, the relations $ [ \mathfrak k_0,\mathfrak k_0]\subseteq \mathfrak k_0 $, $ [ \mathfrak k_0,\mathfrak p_0 | $ vegtices of the first type of $\Gamma$ and hnrting vhe orifin: by tfe previous arguments, we knox thqt $p\ltz (n-1)+1=n$, because the vdrtex witj an edgw to $\unfty$ has eh additljnal cdge, chmch may or may kot hit the origin. The ponyvolial degree of the differential opewator axslciated to $\Gamia$ eatals $n-(2n-1-p)=-n+1+p$, which must be greater or eqhal thai $0$: this forces ommediately $p\geq n-1$. This faft, clmbined with the pgevious coneitijb on $p$, yieldr that either $p=n$ or $p=n-1$.
Jn the case $p=n$, the correspondine difyerential okzeatlt has polynonial qegree $1$, and ngom eacv vertec of the first ty'e dwparts exactly one edje to the origin: thif implies tkat the graph $\Gamma$ mysr havg a varteb of tht fprj
\
\
Of cokrsx, the vertes from whicy departs the arrow tj $\infty$ may be also ysjlated, [*i.e.*]{} no edge has it as the endpoinu: in fhis case, as $n\geq 2$, staneard dimensional argulents impjy that the corresponding weight is trivial. If it is nkg iwojxref, the property $[\mathfrak k_0,\mathfrak p_0]\subset \matrrrsk p_0$ implies thab the two consecutovf rqges correspovd to 'dolagators of type $(+,+)$, wnd Lemia \[l-vabish-4\] yielqs ttiviality of the correspondung weight.
We bow consider the ccse $p=n-1$: the currexpondong differential operatur hzs polynomiwl degree $0$, [*i.e.*]{} it is a tranrlanion-hnvarianu differential opewator. We hirst assume thaj the vqrtex from whicm departs an edge tl $\infjy$ has the other edge hitting the origin: the farv that the digfarettial opzrator has constant coefficients nmplies chat tfis vertex cannot be isolateq, otherwise Lalma \[l-vanish-4\] wmuld impjy teiviqlity ow its weight. Thrrefore, $\Gcima$ nust be a disjoint unkkn of wheel-like grqphs, one of whivh ouse popk nike as follmws:
\
\
Ldmmx \[k-vanirh-4\] implies bhag thr two edges meeting dt tge vertex with the eqge to $\ibfty$ muse have distinvt colors, otherwist the rorres'ondinb wgight vanishes. Therefore, the remations $[\mwthnrak k_0,\mathfrah k_0]\smbsejeq \mathfral k_0$, $[\mathfrak k_0,\mathfrak p_0 | $ vertices of the first type of hitting origin: by previous arguments, we the with an edge $\infty$ has an edge, which may or may not the origin. The polynomial degree of the differential operator associated to $\Gamma$ equals which must be greater or equal than $0$: this forces immediately $p\geq n-1$. fact, with previous on $p$, yields that either $p=n$ or $p=n-1$. In the case $p=n$, the corresponding differential operator polynomial degree $1$, and from each vertex of first type departs exactly edge to the origin: this that graph $\Gamma$ have vertex the form \ Of course, the vertex from which departs the arrow to $\infty$ may be also isolated, [*i.e.*]{} no has it endpoint: in case, $n\geq standard dimensional arguments the corresponding weight is trivial. If isolated, the property $[\mathfrak k_0,\mathfrak p_0]\subset \mathfrak p_0$ that the consecutive edges correspond to propagators of $(+,+)$, and Lemma \[l-vanish-4\] yields triviality of the weight. We now consider the case $p=n-1$: the corresponding differential operator has polynomial degree $0$, is a translation-invariant differential We first assume the from departs edge to has the other edge hitting the origin: the fact that the operator has constant coefficients implies that this vertex cannot be Lemma would imply triviality its weight. Therefore, $\Gamma$ be disjoint union of wheel-like of must follows: \ \[l-vanish-4\] implies that the edges meeting at the vertex the edge to $\infty$ the corresponding weight vanishes. Therefore, the relations $[\mathfrak k_0]\subseteq \mathfrak k_0$, $[\mathfrak k_0,\mathfrak p_0 | $ vertices of the first type of $\GAmma$ and hitTing tHe oRigIn: By thE preVious arguments, WE knoW that $p\leq (n-1)+1=n$, because the vErtex WiTH an eDGe To $\infTy$ has an ADdITIonAl EdGe, wHiCH mAy or mAy nOt hit thE origin. The PolYnOmial degree oF ThE differentIal Operator assoCiaTed to $\GAmMa$ eQUals $n-(2N-1-p)=-n+1+P$, whicH must bE GreateR or equal tHaN $0$: This foRCes immeDIAtEly $p\Geq n-1$. This fact, combiNEd WIth the previous ConditIoN On $P$, YIelDs tHat either $p=N$ oR $p=n-1$.
In THe case $p=N$, ThE CORreSPonding differEntial operaTOr hAs polyNoMiaL Degree $1$, And frOm EAch Vertex of the FirsT type depaRts exaCTly one eDGe to the Origin: ThiS imPlieS ThAt The GrAPh $\GAMmA$ muST haVe a verteX oF tHe forM
\
\
Of cOURSE, the VerTex fRom whIch departs the ArrOw to $\INftY$ may bE also IsolAtEd, [*i.e.*]{} nO edge hAs it aS tHe endpoint: in thiS casE, as $n\geq 2$, stAndArD diMeNsionAL argumEntS imPly that The corrESpoNdING WeIght is trivial. If it iS nOT IsOlated, thE propeRTy $[\MaTHfrak k_0,\maThFraK p_0]\suBSEt \matHfraK P_0$ iMplies thAt the tWO cOnSecutivE eDges coRrEspOnd To proPAgatOrs of tYpe $(+,+)$, and LeMma \[l-vANish-4\] yields trivIAlity of the corREsPONdINg weIghT.
We now consiDer tHE casE $p=n-1$: tHE cOrrESpondIng diFfEReNTial operator has polyNoMial deGree $0$, [*i.E.*]{} it is a translaTion-invariANT DifferenTial OPeRAtor. We first assUme thAt the verteX From whicH depaRts an edgE to $\infty$ hAS The other EdgE hiTtiNg tHE OrIgin: the fact thAT The dIfFerentiAl oPerator Has ConStaNt cOeFficients Implies tHaT tHiS vErtEx canNOt be isolAtEd, oThErwIse LeMMa \[l-vanIsh-4\] woUld iMpLy TRivIality oF ItS WEighT. THeRefoRe, $\GAmMa$ musT be a DIsjOint uniOn of wheel-LikE GrapHs, OnE of whicH must look like As Follows:
\
\
LemMa \[L-vaNish-4\] imPLIes that tHe two edges meeting at the vERtex witH thE edge To $\inFty$ must haVe dIstincT coLOrs, othErwise The coRrEspONDing wEIGhT vaNiShes. TherefORE, thE relaTiOns $[\mAthfrak K_0,\mathfrak k_0]\subseteq \MAthFrak k_0$, $[\mathfrak K_0,\maThfrAK P_0 | $ vertices of the first ty pe of $\Ga mma$and hi tt ingtheorigin: by the prev ious arguments, we kno w tha t$ p\le q ( n-1)+ 1=n$, b e ca u s e t he v ert ex wi th an ed ge to $ \infty$ ha s a nadditional e d ge , which ma y o r may not hi t t he ori gi n.T he po lyn omial degre e of th e differe nt i al ope r ator as s o ci ated to $\Gamma$ equa l s$ n-(2n-1-p)=-n+ 1+p$,wh i ch m ust be greater o requal than $0 $ :t h i s f o rces immediat ely $p\geqn -1$ . This f act , combi ned w it h th e previouscond ition on$p$, y i elds th a t eithe r $p=n $ o r $ p=n- 1 $.
Inth e ca s e$p= n $,the corr es po nding dif f e r e ntia l o pera tor h as polynomial de gree $1$ , and from eac hverte x of t he fi rs t type departsexac tly one e dge t o t he orig i n: thi s i mpl ies tha t the g r aph $ \ G a mm a$ must have a ver te x of the for m
\
\
O fc ourse, t he ve rtex f rom w hich de parts th e arro w t o$\infty $may be a lso is olate d , [* i.e.*] {} no ed ge ha s it as the end p oint: in this ca s e ,a s $n \ge q 2$, stand ardd imen sion a larg u ments impl yt ha t the correspondingwe ight i s tri vial. If it i s not isol a t e d, the p rope r ty $[\mathfrak k_ 0,\ma thfrak p_0 ] \subset\math frak p_0 $ implies t hat thetwo co nse cut i v eedges corresp o n d to p ropagat ors of typ e $ (+, +)$ , a nd Lemma \[ l-vanish -4 \] y ie lds triv i ality of t heco rre spond i ng wei ght.
Weno wc ons ider th e c a s e $p =n -1 $: t heco rresp ondi n g d ifferen tial oper ato r has p ol ynomial degree $0$,[* i.e.*]{} i tisa tran s l ation-in variant differential op e rator.Wefirst ass ume thatthe verte x f r om whi ch dep artsan ed g e to $ \ i nf ty$ h as the oth e r ed ge hi tt ingthe ori gin: the fact that the differential op erat o r h asc on s tan tc oef f i cients impliesthat thisve r te x cannot b e is ol ated, o therwis e Lem m a \[l-v anish-4\] would im pl y tr i v ial ity of its weight. Therefor e , $\G a mm a$ mu stbe a d is joi nt un ion of whe el-li ke gra ph s, one of w hi ch mustlook like as follows:
\
\
L emma\[l -vanish-4 \]i mpl ies thatthetwo edgesmee tin g atthe verte x wi t hthe edgeto $ \ infty$ mu s thav e di stinct colo r s , ot herwi set he cor resp onding weight van i shes. Therefor e, t h e re lat i ons$[ \mathfrak k_0, \ma th f r ak k_0]\ su bseteq \mat hfrak k_ 0$ , $[\m athfra k k_0, \mathfr a k p _ 0 | $ vertices_of the_first type of $\Gamma$_and hitting_the_origin: by_the_previous arguments, we_know that $p\leq_(n-1)+1=n$, because the vertex_with an edge_to_$\infty$ has an additional edge, which may or may not hit the origin. The_polynomial_degree of_the_differential_operator associated to $\Gamma$ equals_$n-(2n-1-p)=-n+1+p$, which must be greater_or equal_than $0$: this forces immediately $p\geq n-1$. This_fact,_combined with the_previous condition on $p$, yields that either $p=n$ or_$p=n-1$.
In the case $p=n$, the corresponding_differential operator has_polynomial_degree_$1$, and from each_vertex of the first type departs_exactly one edge to the origin:_this implies that the graph $\Gamma$ must_have a vertex of the form
\
\
Of_course, the vertex from which_departs the_arrow to $\infty$ may be_also isolated, [*i.e.*]{}_no edge_has it as_the endpoint: in this case, as_$n\geq 2$, standard_dimensional arguments imply that the corresponding_weight_is trivial. If_it_is_not isolated,_the property $[\mathfrak_k_0,\mathfrak_p_0]\subset \mathfrak_p_0$_implies that the two consecutive edges_correspond_to propagators of type $(+,+)$, and Lemma \[l-vanish-4\]_yields triviality of the_corresponding_weight.
We now consider the_case $p=n-1$: the corresponding differential_operator has polynomial degree $0$, [*i.e.*]{}_it is_a translation-invariant_differential operator. We first assume that the vertex from which departs_an edge to $\infty$ has the_other edge hitting the_origin: the_fact_that the differential_operator_has constant_coefficients implies that this vertex cannot be_isolated, otherwise_Lemma \[l-vanish-4\] would imply triviality of its_weight. Therefore, $\Gamma$ must_be_a disjoint union of wheel-like graphs,_one of which must look like_as follows:
\
\
Lemma \[l-vanish-4\] implies that the_two_edges_meeting at the vertex with_the edge to $\infty$ must have_distinct colors, otherwise_the corresponding weight vanishes. Therefore, the relations_$[\mathfrak_k_0,\mathfrak k_0]\subseteq \mathfrak k_0$, $[\mathfrak k_0,\mathfrak_p_0 |
H)$ and assume for some $f\in \PF'(H)$, $\RF(f)$ has only one positive entry in each row. Then we have:
1. After suitable permutation of indices, we can assume $$\RF(f) = \left( \begin{array}{cccc} -1 & \alpha_2-1 & 0 & 0\\
0 & -1 & \al_3-1 & 0 \\
0 & 0 & -1 & \al_4-1\\
\al_1-1 & 0 & 0 & -1 \end{array}\right).$$
2. In this case, if we put $f' =\Fr(H)-f$, then $$\RF(f') = \left( \begin{array}{cccc} -1 & \alpha_2-2 & \al_3-1 & 0\\
0 & -1 & \al_3-2 & \al_4-1 \\
\al_1-1 & 0 & -1 & \al_4-2\\
\al_1-2 & \al_2-1 & 0 & -1 \end{array}\right)$$
3. We have $\type(H) = 3$ with $\PF(H)=
\{f,f', \Fr(H)\}$, $\mu(I_H)=6$ and the minimal generators of $I_H$ are obtained from taking differences of $2$ rows of $\RF(f)$, namely $I_H=(x_1^{\al_1}-x_2^{\al_2-1}x_4, x_2^{\al_2}-x_3^{\al_3-1}x_1, x_3^{\al_1}-x^{\al_4-1}x_2, x_4^{\al_4}-x_1^{\al_1-1}x_3,
x_1x_4^{\al_4-1}-x_2^{\al_2-1}x_3, x_1^{\al_1-1}x_2-x_3^{\al_3-1}x_4).$
4. We have $$n_1= (\al_2-1)(\al_3-1)\al_4 +\al_2, n_2 | H)$ and assume for some $ f\in \PF'(H)$, $ \RF(f)$ has only one positive entrance in each rowing. Then we have:
1. After suitable permutation of index, we can assume $ $ \RF(f) = \left (\begin{array}{cccc } -1 & \alpha_2 - 1 & 0 & 0\\
0 & -1 & \al_3 - 1 & 0 \\
0 & 0 & -1 & \al_4 - 1\\
\al_1 - 1 & 0 & 0 & -1 \end{array}\right).$$
2. In this case, if we arrange $ fluorine' = \Fr(H)-f$, then $ $ \RF(f') = \left (\begin{array}{cccc } -1 & \alpha_2 - 2 & \al_3 - 1 & 0\\
0 & -1 & \al_3 - 2 & \al_4 - 1 \\
\al_1 - 1 & 0 & -1 & \al_4 - 2\\
\al_1 - 2 & \al_2 - 1 & 0 & -1 \end{array}\right)$$
3. We have $ \type(H) = 3 $ with $ \PF(H)=
\{f, f', \Fr(H)\}$, $ \mu(I_H)=6 $ and the minimal generators of $ I_H$ are obtained from taking difference of $ 2 $ rows of $ \RF(f)$, namely $ I_H=(x_1^{\al_1}-x_2^{\al_2 - 1}x_4, x_2^{\al_2}-x_3^{\al_3 - 1}x_1, x_3^{\al_1}-x^{\al_4 - 1}x_2, x_4^{\al_4}-x_1^{\al_1 - 1}x_3,
x_1x_4^{\al_4 - 1}-x_2^{\al_2 - 1}x_3, x_1^{\al_1 - 1}x_2 - x_3^{\al_3 - 1}x_4).$
4. We have $ $ n_1= (\al_2 - 1)(\al_3 - 1)\al_4 + \al_2, n_2 | H)$ ajd assume for some $f\in \PN'(H)$, $\RF(f)$ has only one posmtive ehtry in dach row. Then we have:
1. After syitaboe permutation of indizes, we caj assume $$\RF(f) = \left( \begii{zrray}{ccge} -1 & \amiha_2-1 & 0 & 0\\
0 & -1 & \al_3-1 & 0 \\
0 & 0 & -1 & \dl_4-1\\
\al_1-1 & 0 & 0 & -1 \evd{crray}\right).$$
2. In this case, if we put $f' =\Fr(H)-f$, tnej $$\RF(f') = \left( \bedin{agrwy}{ccdb} -1 & \alpha_2-2 & \al_3-1 & 0\\
0 & -1 & \al_3-2 & \al_4-1 \\
\al_1-1 & 0 & -1 & \al_4-2\\
\al_1-2 & \al_2-1 & 0 & -1 \end{array}\right)$$
3. We jave $\type(H) = 3$ with $\PF(H)=
\{f,f', \Fr(H)\}$, $\mo(J_H)=6$ wbd the minimxl generators of $I_H$ arg obtained from taking differencds of $2$ rows of $\TY(d)$, nwkely $I_H=(x_1^{\al_1}-x_2^{\el_2-1}x_4, x_2^{\aj_2}-x_3^{\al_3-1}x_1, x_3^{\al_1}-x^{\al_4-1}w_2, x_4^{\al_4}-x_1^{\an_1-1}x_3,
x_1c_4^{\al_4-1}-x_2^{\al_2-1}x_3, x_1^{\al_1-1}x_2-x_3^{\ak_3-1}x_4).$
4. We have $$n_1= (\al_2-1)(\al_3-1)\al_4 +\al_2, n_2 | H)$ and assume for some $f\in \PF'(H)$, only positive entry each row. Then permutation indices, we can $$\RF(f) = \left( -1 & \alpha_2-1 & 0 & 0 & -1 & \al_3-1 & 0 \\ 0 & 0 & -1 \al_4-1\\ \al_1-1 & 0 & 0 & -1 \end{array}\right).$$ 2. In this case, we $f' then = \left( \begin{array}{cccc} -1 & \alpha_2-2 & \al_3-1 & 0\\ 0 & -1 & \al_3-2 & \\ \al_1-1 & 0 & -1 & \al_4-2\\ & \al_2-1 & 0 -1 \end{array}\right)$$ 3. We have = with $\PF(H)= \Fr(H)\}$, and minimal generators of are obtained from taking differences of $2$ rows of $\RF(f)$, namely $I_H=(x_1^{\al_1}-x_2^{\al_2-1}x_4, x_2^{\al_2}-x_3^{\al_3-1}x_1, x_3^{\al_1}-x^{\al_4-1}x_2, x_4^{\al_4}-x_1^{\al_1-1}x_3, x_1x_4^{\al_4-1}-x_2^{\al_2-1}x_3, x_1^{\al_1-1}x_2-x_3^{\al_3-1}x_4).$ We have +\al_2, n_2 | H)$ and assume for some $f\in \PF'(H)$, $\RF(F)$ has only onE posiTivE enTrY in eAch rOw. Then we have:
1. AfTEr suItable permutation of indIces, wE cAN assUMe $$\rF(f) = \leFt( \begin{ARrAY}{CccC} -1 & \aLpHa_2-1 & 0 & 0\\
0 & -1 & \aL_3-1 & 0 \\
0 & 0 & -1 & \aL_4-1\\
\Al_1-1 & 0 & 0 & -1 \End{arRay}\Right).$$
2. In This case, if We pUt $F' =\Fr(H)-f$, then $$\RF(f') = \LEfT( \begin{arraY}{ccCc} -1 & \alpha_2-2 & \al_3-1 & 0\\
0 & -1 & \al_3-2 & \aL_4-1 \\
\al_1-1 & 0 & -1 & \Al_4-2\\
\al_1-2 & \al_2-1 & 0 & -1 \EnD{arRAy}\rigHt)$$
3. WE have $\Type(H) = 3$ wITh $\PF(H)=
\{f,F', \Fr(H)\}$, $\mu(I_H)=6$ aNd THe miniMAl generATOrS of $I_h$ are obtained from tAKiNG differences of $2$ Rows of $\rF(F)$, NaMELy $I_h=(x_1^{\aL_1}-x_2^{\al_2-1}x_4, x_2^{\al_2}-x_3^{\aL_3-1}x_1, X_3^{\al_1}-x^{\aL_4-1}X_2, x_4^{\al_4}-x_1^{\al_1-1}X_3,
X_1x_4^{\AL_4-1}-X_2^{\Al_2-1}x_3, X_1^{\Al_1-1}x_2-x_3^{\al_3-1}x_4).$
4. We have $$N_1= (\al_2-1)(\al_3-1)\al_4 +\al_2, n_2 | H)$ and assume for some $f \in \PF'(H )$, $ \RF (f) $hasonly one positivee ntry in each row. Then wehave:
1 . A f te r sui table p e rm u t ati on o f i nd i ce s, we ca n assum e $$\RF(f) =\l eft( \begin{ a rr ay}{cccc}-1& \alpha_2-1 &0 & 0\ \ 0 & - 1 & \al_ 3-1 &0 \\
0 & 0 & - 1 & \al _ 4-1\\
\ al_1 -1 & 0 & 0 & -1 \ e nd { array}\right). $$
2. I nt h iscas e, if we p ut $f'= \Fr(H)- f $, t h en$ $\RF(f') = \l eft( \begin { arr ay}{cc cc } - 1 & \al pha_2 -2 & \ al_3-1 & 0\ \
0 & -1& \al_ 3 -2 & \a l _4-1 \\
\ al_ 1-1 & 0 &-1 &\a l _4- 2 \\
\a l_1-2 &\a l_ 2-1 & 0 & - 1 \end {ar ray} \righ t)$$
3. Wehav e $\ t ype (H) = 3$ w ith$\ PF(H) =
\{f,f ', \Fr(H)\}$, $\m u(I_ H)=6$ and th emin im al ge n erator s o f $ I_H$ ar e obtai n edfr o m ta king differences o f$ 2 $rows of$\RF(f ) $, n a mely $I_ H= (x_ 1^{\ a l _1}-x _2^{ \ al _2-1}x_4 , x_2^ { \a l_ 2}-x_3^ {\ al_3-1 }x _1, x_ 3^{\a l _1}- x^{\al _4-1}x_2 , x_4 ^ {\al_4}-x_1^{\ a l_1-1}x_3,
x _ 1 x_ 4 ^{\a l_4 -1}-x_2^{\a l_2- 1 }x_3 , x_ 1 ^{ \al _ 1-1}x _2-x_ 3^ { \a l _3-1}x_4).$
4. We h ave $$ n_1=(\al_2-1)(\al _3-1)\al_4 + \ al_2, n_ 2 | H)$ and_assume for_some $f\in \PF'(H)$, $\RF(f)$_has only_one_positive entry_in_each row. Then_we have:
1. _After suitable permutation of_indices, we can_assume_$$\RF(f) = \left( \begin{array}{cccc} -1 & \alpha_2-1 & 0 & 0\\
_0_& -1_&_\al_3-1_& 0 \\
_ 0 & 0 &_-1 &_\al_4-1\\
\al_1-1 & 0 &_0_& -1 \end{array}\right).$$
2._ In this case, if we put $f' =\Fr(H)-f$,_then $$\RF(f') = \left( \begin{array}{cccc} -1_& \alpha_2-2 &_\al_3-1_&_0\\
_0 & -1 & \al_3-2 &_\al_4-1 \\
\al_1-1_& 0 & -1 & \al_4-2\\
_ \al_1-2 & \al_2-1 &_0 & -1 \end{array}\right)$$
3. _We have_$\type(H) = 3$ with $\PF(H)=
_ _\{f,f', \Fr(H)\}$,_$\mu(I_H)=6$ and the_minimal generators of $I_H$ are obtained_from taking differences_of $2$ rows of $\RF(f)$, namely_$I_H=(x_1^{\al_1}-x_2^{\al_2-1}x_4,_x_2^{\al_2}-x_3^{\al_3-1}x_1, x_3^{\al_1}-x^{\al_4-1}x_2, x_4^{\al_4}-x_1^{\al_1-1}x_3,
___ x_1x_4^{\al_4-1}-x_2^{\al_2-1}x_3,_x_1^{\al_1-1}x_2-x_3^{\al_3-1}x_4).$
4. We_have_$$n_1= (\al_2-1)(\al_3-1)\al_4_+\al_2,_n_2 |
m)^*S^0$ for some $x\in S^0$;\
$(2)$ $xax=x$, $^\circ{x}=\!\!~^\circ(a^m)$, $^\circ(x^*)\subseteq\!\!~^\circ(a^m)$ and $^\circ(a^m)^* \subseteq\!\!~^\circ (x^m)$ for some $x\in S^0$.
Suppose $a$ is \*-DMP with index $m$. Then $a^{\scriptsize\textcircled{\tiny D}_m},~(a^m)^{\dag}$ exist with $(a^{\scriptsize\textcircled{\tiny D}_m})^m=(a^m)^{\dag}$ by Lemma \[1\]. Take $x=a^{\scriptsize\textcircled{\tiny D}_m}$, then $xax=x,~xS^0=x^*S^0=a^mS^0$ by Lemma \[7\]. Further, from $x^m=(a^m)^{\dag}$, it follows that $x^m=(x^ma^m)^*x^m=(a^m)^*(x^m)^*x^m\in (a^m)^*S^0$ and $(a^m)^*=(a^mx^ma^m)^*=x^ma^m(a^m)^*\in x^mS^0$. Hence (1) holds.\
$(1)\Rightarrow(2)$ is clear.\
(2). From $xax=x$, $^\circ{x}=\!\!~^\circ(a^m)$ and $^\circ(x^*)\subseteq\!\!~^\circ(a^m)$, it follows that $a^{\scriptsize\textcircled{\tiny D}_m}=x$ by Lemma \[7\]. Then $1-(x^ma^m)^*\in\!\!~ ^\circ (a^m)^*\subseteq\!\!~^\circ(x^m)$ implies $x^m=(x^ma^m)^*x^m$. So $x^ma^m=(x^ma^m)^*x^ma^m$. Therefore $(x^ma^m)^*=x^ma^m$, together with $xa^{m+1}=a^m,~ax^2=x$, implies $a$ is \*-DMP with index $m$ by Theorem | m)^*S^0 $ for some $ x\in S^0$;\
$ (2)$ $ xax = x$, $ ^\circ{x}=\!\!~^\circ(a^m)$, $ ^\circ(x^*)\subseteq\!\!~^\circ(a^m)$ and $ ^\circ(a^m)^ * \subseteq\!\!~^\circ (x^m)$ for some $ x\in S^0$.
Suppose $ a$ is \*-DMP with index $ m$. Then $ a^{\scriptsize\textcircled{\tiny D}_m},~(a^m)^{\dag}$ exist with $ (a^{\scriptsize\textcircled{\tiny D}_m})^m=(a^m)^{\dag}$ by Lemma \[1\ ]. Take $ x = a^{\scriptsize\textcircled{\tiny D}_m}$, then $ xax = x,~xS^0 = x^*S^0 = a^mS^0 $ by Lemma \[7\ ]. far, from $ x^m=(a^m)^{\dag}$, it take after that $ x^m=(x^ma^m)^*x^m=(a^m)^*(x^m)^*x^m\in (a^m)^*S^0 $ and $ (a^m)^*=(a^mx^ma^m)^*=x^ma^m(a^m)^*\in x^mS^0$. Hence (1) holds.\
$ (1)\Rightarrow(2)$ is clear.\
(2). From $ xax = x$, $ ^\circ{x}=\!\!~^\circ(a^m)$ and $ ^\circ(x^*)\subseteq\!\!~^\circ(a^m)$, it follows that $ a^{\scriptsize\textcircled{\tiny D}_m}=x$ by Lemma \[7\ ]. Then $ 1-(x^ma^m)^*\in\!\!~ ^\circ (a^m)^*\subseteq\!\!~^\circ(x^m)$ implies $ x^m=(x^ma^m)^*x^m$. then $ x^ma^m=(x^ma^m)^*x^ma^m$. Therefore $ (x^ma^m)^*=x^ma^m$, together with $ xa^{m+1}=a^m,~ax^2 = x$, incriminate $ a$ is \*-DMP with exponent $ m$ by Theorem | m)^*S^0$ vor some $x\in S^0$;\
$(2)$ $xax=x$, $^\circ{x}=\!\!~^\girc(a^m)$, $^\circ(x^*)\subsgtwq\!\!~^\circ(e^m)$ and $^\dirc(a^m)^* \sjbseteq\!\!~^\circ (x^m)$ for some $x\in D^0$.
Syppost $a$ is \*-DMP with inddx $m$. Then $a^{\scriptwize\uextcircled{\tiny D}_m},~(a^m)^{\dag}$ ewnst wjbh $(a^{\sermptsize\textcircked{\tiny D}_m})^k=(a^m)^{\dag}$ by Lemmd \[1\]. Gane $x=a^{\scriptsize\textcircled{\tiny D}_m}$, tren $xax=c,~xD^0=x^*S^0=a^mS^0$ by Lemmw \[7\]. Fugtrer, rgon $x^m=(a^m)^{\dag}$, it follows that $x^m=(x^ja^m)^*x^m=(a^m)^*(q^m)^*x^m\in (a^m)^*S^0$ and $(a^m)^*=(s^mx^ma^m)^*=x^ma^m(a^m)^*\in x^mS^0$. Hence (1) joldd.\
$(1)\Rightarrow(2)$ is clewr.\
(2). From $xax=z$, $^\ciwx{x}=\!\!~^\circ(a^m)$ and $^\circ(x^*)\substtzq\!\!~^\circ(a^m)$, it follows that $a^{\scriptsize\textcifcled{\ciny D}_m}=x$ by Lwmmw \[7\]. Then $1-(x^ma^m)^*\mn\!\!~ ^\cirb (a^m)^*\subseteq\!\!~^\clgc(x^m)$ im[lies $x^k=(x^ma^m)^*x^m$. So $x^ma^k=(x^me^m)^*x^mq^m$. Therefore $(x^ma^m)^*=x^ma^m$, together with $xa^{m+1}=w^m,~ax^2=x$, impnizs $a$ is \*-DMP with indez $n$ by Jheoram | m)^*S^0$ for some $x\in S^0$;\ $(2)$ $xax=x$, and \subseteq\!\!~^\circ (x^m)$ some $x\in S^0$. index Then $a^{\scriptsize\textcircled{\tiny D}_m},~(a^m)^{\dag}$ with $(a^{\scriptsize\textcircled{\tiny D}_m})^m=(a^m)^{\dag}$ Lemma \[1\]. Take $x=a^{\scriptsize\textcircled{\tiny D}_m}$, then by Lemma \[7\]. Further, from $x^m=(a^m)^{\dag}$, it follows that $x^m=(x^ma^m)^*x^m=(a^m)^*(x^m)^*x^m\in (a^m)^*S^0$ and $(a^m)^*=(a^mx^ma^m)^*=x^ma^m(a^m)^*\in Hence (1) holds.\ $(1)\Rightarrow(2)$ is clear.\ (2). From $xax=x$, $^\circ{x}=\!\!~^\circ(a^m)$ and $^\circ(x^*)\subseteq\!\!~^\circ(a^m)$, it that D}_m}=x$ Lemma Then $1-(x^ma^m)^*\in\!\!~ ^\circ (a^m)^*\subseteq\!\!~^\circ(x^m)$ implies $x^m=(x^ma^m)^*x^m$. So $x^ma^m=(x^ma^m)^*x^ma^m$. Therefore $(x^ma^m)^*=x^ma^m$, together with $xa^{m+1}=a^m,~ax^2=x$, implies $a$ is with index $m$ by Theorem | m)^*S^0$ for some $x\in S^0$;\
$(2)$ $xax=x$, $^\circ{x}=\!\!~^\cirC(a^m)$, $^\circ(x^*)\suBseteQ\!\!~^\ciRc(a^M)$ aNd $^\ciRc(a^m)^* \Subseteq\!\!~^\circ (x^m)$ FOr soMe $x\in S^0$.
Suppose $a$ is \*-DMP witH indeX $m$. tHen $a^{\SCrIptsiZe\textcIRcLED{\tiNy d}_m},~(A^m)^{\dAg}$ EXiSt witH $(a^{\sCriptsiZe\textcircLed{\TiNy D}_m})^m=(a^m)^{\dag}$ by lEmMa \[1\]. Take $x=a^{\scRipTsize\textcirCleD{\tiny D}_M}$, tHen $XAx=x,~xS^0=X^*S^0=a^MS^0$ by LEmma \[7\]. FuRTher, frOm $x^m=(a^m)^{\dag}$, It FOllows THat $x^m=(x^mA^M)^*X^m=(A^m)^*(x^m)^*X^m\in (a^m)^*S^0$ and $(a^m)^*=(a^mx^ma^M)^*=X^mA^M(a^m)^*\in x^mS^0$. Hence (1) hOlds.\
$(1)\RiGhTArROW(2)$ is CleAr.\
(2). From $xax=x$, $^\CiRc{x}=\!\!~^\ciRC(a^m)$ and $^\cIRc(X^*)\SUBseTEq\!\!~^\circ(a^m)$, it folLows that $a^{\scRIptSize\teXtCirCLed{\tinY D}_m}=x$ bY LEMma \[7\]. then $1-(x^ma^m)^*\in\!\!~ ^\cIrc (a^M)^*\subseteq\!\!~^\Circ(x^m)$ IMplies $x^M=(X^ma^m)^*x^m$. SO $x^ma^m=(x^Ma^m)^*X^ma^M$. TheREfOrE $(x^mA^m)^*=X^Ma^m$, TOgEthER wiTh $xa^{m+1}=a^m,~aX^2=x$, ImPlies $A$ is \*-Dmp WITh inDex $M$ by THeoreM | m)^*S^0$ for some $x\in S^ 0$;\
$(2)$ $xax =x$ , $ ^\ circ {x}= \!\!~^\circ(a^ m )$,$^\circ(x^*)\subseteq\ !\!~^ \c i rc(a ^ m) $ and $^\cir c (a ^ m )^* \ su bse te q \! \!~^\ cir c (x^m) $ for some $x \i n S^0$.
Su p po se $a$ is\*- DMP with ind ex$m$. T he n $ a ^{\sc rip tsize \textc i rcled{ \tiny D}_ m} , ~(a^m) ^ {\dag}$ e xi st w ith $(a^{\scripts i ze \ textcircled{\t iny D} _m } )^ m = (a^ m)^ {\dag}$ by L emma\ [1\]. T a ke $ x =a^ { \scriptsize\t extcircled{ \ tin y D}_m }$ , t h en $xa x=x,~ xS ^ 0=x ^*S^0=a^mS^ 0$ b y Lemma \ [7\].F urther, from $x ^m=(a^ m)^ {\d ag}$ , i tfol lo w s t h at $x ^ m=( x^ma^m)^ *x ^m =(a^m )^*( x ^ m ) ^*x^ m\i n (a ^m)^* S^0$ and $(a^ m)^ *=(a ^ mx^ ma^m) ^*=x^ ma^m (a ^m)^* \in x^ mS^0$ .Hence (1) holds .\
$ (1)\Right arr ow (2) $is cl e ar.\
( 2). Fr om $xax =x$, $^ \ cir c{ x } = \! \!~^\circ(a^m)$ an d$ ^ \c irc(x^*) \subse t eq \! \ !~^\circ (a ^m) $, i t follo ws t h at $a^{\sc riptsi z e\ te xtcircl ed {\tiny D }_m }=x $ byL emma \[7\] . Then $ 1-(x^ m a^m)^*\in\!\!~ ^\circ (a^m)^ * \s u b se t eq\! \!~ ^\circ(x^m) $ im p lies $x^ m =( x^m a ^m)^* x^m$. S o $ x ^ma^m=(x^ma^m)^*x^m a^ m$. Th erefo re $(x^ma^m)^ *=x^ma^m$, t o gether w ith$ xa ^ {m+1}=a^m,~ax^ 2=x$, implies $ a $ is \*- DMP w ith inde x $m$ byT h eorem | m)^*S^0$ for_some $x\in_S^0$;\
$(2)$ $xax=x$, $^\circ{x}=\!\!~^\circ(a^m)$, $^\circ(x^*)\subseteq\!\!~^\circ(a^m)$ and_$^\circ(a^m)^* \subseteq\!\!~^\circ_(x^m)$_for some_$x\in_S^0$.
Suppose $a$ is_\*-DMP with index_$m$. Then $a^{\scriptsize\textcircled{\tiny D}_m},~(a^m)^{\dag}$_exist with $(a^{\scriptsize\textcircled{\tiny_D}_m})^m=(a^m)^{\dag}$_by Lemma \[1\]. Take $x=a^{\scriptsize\textcircled{\tiny D}_m}$, then $xax=x,~xS^0=x^*S^0=a^mS^0$ by Lemma \[7\]. Further, from $x^m=(a^m)^{\dag}$, it follows_that_$x^m=(x^ma^m)^*x^m=(a^m)^*(x^m)^*x^m\in (a^m)^*S^0$_and_$(a^m)^*=(a^mx^ma^m)^*=x^ma^m(a^m)^*\in_x^mS^0$. Hence (1) holds.\
$(1)\Rightarrow(2)$ is_clear.\
(2). From $xax=x$, $^\circ{x}=\!\!~^\circ(a^m)$ and_$^\circ(x^*)\subseteq\!\!~^\circ(a^m)$, it_follows that $a^{\scriptsize\textcircled{\tiny D}_m}=x$ by Lemma \[7\]. Then_$1-(x^ma^m)^*\in\!\!~_^\circ (a^m)^*\subseteq\!\!~^\circ(x^m)$ implies_$x^m=(x^ma^m)^*x^m$. So $x^ma^m=(x^ma^m)^*x^ma^m$. Therefore $(x^ma^m)^*=x^ma^m$, together with $xa^{m+1}=a^m,~ax^2=x$, implies_$a$ is \*-DMP with index $m$_by Theorem |
:= \{ {\mathbf y}\in (L')^{n-1} \ | \langle {\mathbf y}, {\mathbf y}\rangle = T_{2} \text{ and } {\mathbb D}^+_{y} \cap \overline U = \emptyset \}.$$
Using the estimates and, and standard arguments for convergence of theta series, it follows that the sum $$\label{eqn:xi_0 sum}
\sum_{T = {\left( \begin{smallmatrix}* & * \\ * & T_2 \end{smallmatrix} \right)}} \sum_{\substack{({\mathbf x}_1, {\mathbf y}) \in \Omega(T) \\ {\mathbf y}\in S_2}}\varphi({\mathbf x}_1,{\mathbf y}) \varphi_{{\mathrm{KM}}} \left( \sqrt{v_1} x_1+ \frac{y \cdot v'_{12} }{ \sqrt{v_1}} \right)\wedge { \mathfrak{g}}(ya_2^*) \prod_{i=2}^d \varphi_{\infty_i}( \sigma_i ({\mathbf x}_1, {\mathbf y}) a_i) \ e( T \mathbf u )$$ converges absolutely to a smooth form on ${\mathbb H}^d_n \times U$. The (finitely many) remaining terms, corresponding to ${\mathbf y}\in S_1$, can be written as $$\sum_{{\mathbf y}\in S_1} f_{{\mathbf y}}({\bm{\tau}})(\varphi)\wedge { \mathfrak{g}}(y a_2^*)$$ where, for any ${\mathbf y}\in V^{n-1}$ and $\varphi \in S(L^n)$, we set $$f_{{\mathbf y}}({\bm{\tau}}) (\varphi) \ = \ \sum_{ {\mathbf x}_1 \in V} \varphi({\mathbf x}_1, {\mathbf y}) \varphi_{{\mathrm{KM}}} \left( \sqrt{v_1} x_1+ \frac{y \cdot v'_{12} }{ \sqrt{v_1}} \right) \prod_{i=2}^d \varphi_{\infty_i}( \sigma_i ({\mathbf x}_1, {\mathbf y})a_i) \ e( T({\mathbf x}_1, {\mathbf y})\mathbf u ),$$ where $T({\mathbf x}_1, | : = \ { { \mathbf y}\in (L')^{n-1 } \ | \langle { \mathbf y }, { \mathbf y}\rangle = T_{2 } \text { and } { \mathbb D}^+_{y } \cap \overline U = \emptyset \}.$$
Using the estimates and, and standard arguments for overlap of theta serial, it stick to that the sum $ $ \label{eqn: xi_0 kernel }
\sum_{T = { \left (\begin{smallmatrix } * & * \\ * & T_2 \end{smallmatrix } \right) } } \sum_{\substack{({\mathbf x}_1, { \mathbf y }) \in \Omega(T) \\ { \mathbf y}\in S_2}}\varphi({\mathbf x}_1,{\mathbf y }) \varphi_{{\mathrm{KM } } } \left (\sqrt{v_1 } x_1 + \frac{y \cdot v'_{12 } } { \sqrt{v_1 } } \right)\wedge { \mathfrak{g}}(ya_2^ *) \prod_{i=2}^d \varphi_{\infty_i } (\sigma_i ({ \mathbf x}_1, { \mathbf y }) a_i) \ e (T \mathbf u) $ $ converges absolutely to a fluent form on $ { \mathbb H}^d_n \times U$. The (finitely many) remaining condition, corresponding to $ { \mathbf y}\in S_1 $, can be written as $ $ \sum_{{\mathbf y}\in S_1 } f_{{\mathbf y}}({\bm{\tau}})(\varphi)\wedge { \mathfrak{g}}(y a_2^*)$$ where, for any $ { \mathbf y}\in V^{n-1}$ and $ \varphi \in S(L^n)$, we sic $ $ f_{{\mathbf y}}({\bm{\tau } }) (\varphi) \ = \ \sum _ { { \mathbf x}_1 \in V } \varphi({\mathbf x}_1, { \mathbf y }) \varphi_{{\mathrm{KM } } } \left (\sqrt{v_1 } x_1 + \frac{y \cdot v'_{12 } } { \sqrt{v_1 } } \right) \prod_{i=2}^d \varphi_{\infty_i } (\sigma_i ({ \mathbf x}_1, { \mathbf y})a_i) \ e (T({\mathbf x}_1, { \mathbf y})\mathbf u), $ $ where $ T({\mathbf x}_1, | := \{ {\mathbf y}\in (L')^{n-1} \ | \langle {\oathbf y}, {\mathbf y}\ranglx = T_{2} \test{ and } {\oathbb D}^+_{y} \cap \overline U = \em'tyswt \}.$$
Uwing the estimates and, and stanfard argymenus for convergencx of thebc serjcs, it hollows that thg sum $$\label{exn:xi_0 sum}
\rul_{T = {\left( \begin{smallmatrix}* & * \\ * & T_2 \enq{smallmstgix} \right)}} \sum_{\sobstabk{({\iathgf x}_1, {\mathbf y}) \in \Omega(T) \\ {\mathbf y}\ih S_2}}\varpii({\mathbf x}_1,{\mathbg y}) \varphi_{{\mathrm{KM}}} \left( \sqgt{v_1} d_1+ \frac{y \cdot v'_{12} }{ \sqgt{v_1}} \right)\weege { \nathfrak{g}}(ya_2^*) \prod_{i=2}^d \ncrphi_{\infty_i}( \sigma_i ({\mathbf x}_1, {\mathbf y}) a_i) \ d( T \mcthbf u )$$ cobvwrgfv absolutelb to a smooth form on ${\mathtb H}^d_n \yimes U$. The (fikitelb maby) remaining terms, cocresponding to ${\mathbs y}\in S_1$, cdn be written as $$\sun_{{\mqthbf y}\in S_1} d_{{\maghbr b}}({\bm{\fau}})(\varohi)\xedge { \mathrrak{g}}(y a_2^*)$$ whwre, for any ${\mathbf u}\ig V^{n-1}$ and $\varphj \in S(J^n)$, we set $$f_{{\mathbf y}}({\bm{\tau}}) (\varphi) \ = \ \sum_{ {\mdthgf x}_1 \in V} \varphi({\mathbf x}_1, {\mathbf y}) \varphi_{{\matjrm{KM}}} \lefe( \sqrt{v_1} x_1+ \frac{y \cdot v'_{12} }{ \sqrt{v_1}} \right) \prod_{i=2}^d \varpvi_{\infvy_k}( \snnix_u ({\lathbf x}_1, {\mathbf y})a_i) \ e( T({\mathbf x}_1, {\mathbf y})\matrgf u ),$$ where $T({\mathbn x}_1, | := \{ {\mathbf y}\in (L')^{n-1} \ | y}, y}\rangle = \text{ and } = \}.$$ Using the and, and standard for convergence of theta series, it that the sum $$\label{eqn:xi_0 sum} \sum_{T = {\left( \begin{smallmatrix}* & * \\ * T_2 \end{smallmatrix} \right)}} \sum_{\substack{({\mathbf x}_1, {\mathbf y}) \in \Omega(T) \\ {\mathbf y}\in S_2}}\varphi({\mathbf y}) \left( x_1+ \cdot v'_{12} }{ \sqrt{v_1}} \right)\wedge { \mathfrak{g}}(ya_2^*) \prod_{i=2}^d \varphi_{\infty_i}( \sigma_i ({\mathbf x}_1, {\mathbf y}) a_i) \ T \mathbf u )$$ converges absolutely to a form on ${\mathbb H}^d_n U$. The (finitely many) remaining corresponding ${\mathbf y}\in can written $$\sum_{{\mathbf y}\in S_1} y}}({\bm{\tau}})(\varphi)\wedge { \mathfrak{g}}(y a_2^*)$$ where, for any ${\mathbf y}\in V^{n-1}$ and $\varphi \in S(L^n)$, we set $$f_{{\mathbf (\varphi) \ \sum_{ {\mathbf \in \varphi({\mathbf {\mathbf y}) \varphi_{{\mathrm{KM}}} x_1+ \frac{y \cdot v'_{12} }{ \sqrt{v_1}} \sigma_i ({\mathbf x}_1, {\mathbf y})a_i) \ e( T({\mathbf {\mathbf y})\mathbf ),$$ where $T({\mathbf x}_1, | := \{ {\mathbf y}\in (L')^{n-1} \ | \langle {\mathbf y}, {\mAthbf y}\rangLe = T_{2} \teXt{ aNd } {\mAtHbb D}^+_{Y} \cap \Overline U = \emptySEt \}.$$
UsIng the estimates and, and sTandaRd ARgumENtS for cOnvergeNCe OF TheTa SeRieS, iT FoLlows ThaT the sum $$\Label{eqn:xi_0 Sum}
\SuM_{T = {\left( \begin{sMAlLmatrix}* & * \\ * & T_2 \enD{smAllmatrix} \rigHt)}} \sUm_{\subsTaCk{({\mAThbf x}_1, {\MatHbf y}) \iN \Omega(t) \\ {\Mathbf Y}\in S_2}}\varphI({\mAThbf x}_1,{\mAThbf y}) \vaRPHi_{{\MathRm{KM}}} \left( \sqrt{v_1} x_1+ \fraC{Y \cDOt v'_{12} }{ \sqrt{v_1}} \right)\wEdge { \maThFRaK{G}}(Ya_2^*) \pRod_{I=2}^d \varphi_{\inFtY_i}( \sigMA_i ({\mathbF X}_1, {\mATHBf y}) A_I) \ e( T \mathbf u )$$ conVerges absolUTelY to a smOoTh fORm on ${\maThbb H}^D_n \TImeS U$. The (finiteLy maNy) remainiNg termS, CorrespONding to ${\Mathbf Y}\in s_1$, caN be wRItTeN as $$\SuM_{{\MatHBf Y}\in s_1} F_{{\maThbf y}}({\bm{\tAu}})(\VaRphi)\wEdge { \MATHFrak{G}}(y a_2^*)$$ WherE, for aNy ${\mathbf y}\in V^{n-1}$ And $\VarpHI \in s(L^n)$, we Set $$f_{{\mAthbF y}}({\Bm{\tau}}) (\Varphi) \ = \ \Sum_{ {\maThBf x}_1 \in V} \varphi({\matHbf x}_1, {\Mathbf y}) \vaRphI_{{\mAthRm{kM}}} \lefT( \Sqrt{v_1} x_1+ \FraC{y \cDot v'_{12} }{ \sqrT{v_1}} \right) \PRod_{I=2}^d \VARPhI_{\infty_i}( \sigma_i ({\mathbF x}_1, {\MAThBf y})a_i) \ e( T({\mAthbf x}_1, {\MAtHbF Y})\mathbf u ),$$ WhEre $t({\matHBF x}_1, | := \{ {\mathbf y}\in (L' )^{n-1} \| \la ngl e { \m athb f y} , {\mathbf y}\ r angl e = T_{2} \text{ and } {\ma th b b D} ^ +_ {y} \ cap \ov e rl i n e U = \ emp ty s et \}. $$
Usingthe estima tes a nd, and stan d ar d argument s f or convergen ceof the ta se r ies,itfollo ws tha t the s um $$\lab el { eqn:xi _ 0 sum} \sum_{T = {\left( \b e gin{smallmatri x}* &*\ \* & T _2\end{small ma trix} \right) } }\ s u m_{ \ substack{({\m athbf x}_1, {\m athbfy} ) \ i n \Ome ga(T) \ \ {\ mathbf y}\i n S_ 2}}\varph i({\ma t hbf x}_ 1 ,{\math bf y}) \v arp hi_{ { \m at hrm {K M }}} \l eft ( \s qrt{v_1} x _1 + \fr ac{y \ c d ot v '_{ 12}}{ \s qrt{v_1}} \ri ght )\we d ge{ \ma thfra k{g} }( ya_2^ *) \prod _{ i=2}^d \varphi_ {\in fty_i}( \ sig ma _i({ \math b f x}_1 , { \ma thbf y} ) a_i)\ e (T \ ma thbf u )$$ converg es a bs olutelyto a s m oo th form on${ \ma thbb H }^d_n \ti m es U$. The (fini t el ymany) r em aining t erm s,corre s pond ing to ${\math bf y} \ in S_1$, can b e written as $ $ \s u m _{ { \mat hbf y}\in S_1} f _ {{\m athb f y }}( { \bm{\ tau}} )( \ va r phi)\wedge { \mathf ra k{g}}( y a_2 ^*)$$ where,for any ${ \ m a thbf y}\ in V ^ {n - 1}$ and $\varp hi \i n S(L^n)$, we set $ $f_{{ \mathbfy}}({\bm{ \ t au}}) (\ var phi ) \ =\ \ sum_{ {\mathb f x}_1 \ in V} \va rphi({\ mat hbf x} _1, { \mathbf y }) \varp hi _{ {\ ma thr m{KM} } } \left( \ sqr t{ v_1 } x_1 + \frac {y \c dotv' _{ 1 2}}{ \sqr t {v _ 1 }} \ ri gh t) \p ro d_{i= 2}^d \va rphi_{\ infty_i}( \s i gma_ i({ \mathbf x}_1, {\math bf y})a_i) \e(T({\ma t h bf x}_1, {\mathbf y})\mathbf u) ,$$ whe re$T({\ math bf x}_1, | :=_ \{_{\mathbf y}\in (L')^{n-1} \_| \langle_{\mathbf_y}, {\mathbf_y}\rangle_= T_{2} \text{_and } {\mathbb_D}^+_{y} \cap \overline U_= \emptyset _\}.$$
Using_the estimates and, and standard arguments for convergence of theta series, it follows that_the_sum $$\label{eqn:xi_0_sum}
__ _ \sum_{T = {\left( \begin{smallmatrix}*_& *_\\ * & T_2 \end{smallmatrix} \right)}} \sum_{\substack{({\mathbf x}_1,_{\mathbf_y}) \in \Omega(T)_\\ {\mathbf y}\in S_2}}\varphi({\mathbf x}_1,{\mathbf y}) \varphi_{{\mathrm{KM}}} \left( \sqrt{v_1}_x_1+ \frac{y \cdot v'_{12} }{ \sqrt{v_1}}_\right)\wedge { \mathfrak{g}}(ya_2^*)___ \prod_{i=2}^d \varphi_{\infty_i}( \sigma_i_({\mathbf x}_1, {\mathbf y}) a_i) _\ e( T \mathbf u )$$_converges absolutely to a smooth form on_${\mathbb H}^d_n \times U$. The (finitely_many) remaining terms, corresponding to_${\mathbf y}\in_S_1$, can be written as_$$\sum_{{\mathbf y}\in S_1}_ _f_{{\mathbf y}}({\bm{\tau}})(\varphi)\wedge {_\mathfrak{g}}(y a_2^*)$$ where, for any ${\mathbf_y}\in V^{n-1}$ and_$\varphi \in S(L^n)$, we set $$f_{{\mathbf_y}}({\bm{\tau}})_(\varphi) \ =_\__\sum_{ {\mathbf_x}_1 \in V}__\varphi({\mathbf x}_1,_{\mathbf_y}) \varphi_{{\mathrm{KM}}} \left( \sqrt{v_1} x_1+ \frac{y_\cdot_v'_{12} }{ \sqrt{v_1}} \right) \prod_{i=2}^d_\varphi_{\infty_i}( \sigma_i ({\mathbf x}_1,_{\mathbf_y})a_i) \ e(_T({\mathbf x}_1, {\mathbf y})\mathbf u_),$$ where $T({\mathbf x}_1, |
_i(s)^2 ds \right) \right| \geq\epsilon_n\right] \nonumber\\
\label{eq.estgitntnpl1}
&= 4\,\mathbb{P}\left[ B_{i,n}^*\left( \int_{t_n}^{t_{n+1}}g_i(s)^2 ds \right) \geq\epsilon_n\right]
= 4\,\left\{ 1-\Phi\left(\frac{\epsilon_n}{\sqrt{\int_{t_n}^{t_{n+1}}g_i(s)^2 ds}} \right) \right\},\end{aligned}$$ where we have used the fact that $\max_{0\leq s\leq t} W(s)$ has the same distribution as $|W(t)|$ when $W$ is a standard Brownian motion, the symmetry of the distribution of a standard Brownian motion, and $\Phi$ denotes the distribution function of a standard normal random variable. Now, $$\begin{aligned}
g_2(t)^2 &= \pi |b|^{1/2} (1+t)^{1/2} \sin^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq \pi |b|^{1/2} (1+t)^{1/2}, \\
g_4(t)^2 &= \pi |b|^{1/2} (1+t)^{1/2} \cos^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq \pi |b|^{1/2} (1+t)^{1/2}.\end{aligned}$$ Thus, by we have $$\label{eq:giasy}
\int_{t_n}^{t_{n+1}}g_i(s)^2 ds \leq \pi |b|^{1/2} (1+t_{n+1})^{1/2} (t_{n+1}-t_n) \sim \pi |b|^{1/2} t_n^{1/2} \Delta t_n, \quad \text{ as $n\to\infty$},$$ and therefore by and $$\limsup_{n\to\infty}
\frac{\sqrt{\int_{ | _ i(s)^2 ds \right) \right| \geq\epsilon_n\right ] \nonumber\\
\label{eq.estgitntnpl1 }
& = 4\,\mathbb{P}\left [ B_{i, n}^*\left (\int_{t_n}^{t_{n+1}}g_i(s)^2 ds \right) \geq\epsilon_n\right ]
= 4\,\left\ { 1-\Phi\left(\frac{\epsilon_n}{\sqrt{\int_{t_n}^{t_{n+1}}g_i(s)^2 ds } } \right) \right\},\end{aligned}$$ where we have used the fact that $ \max_{0\leq s\leq t } W(s)$ has the same distribution as $ |W(t)|$ when $ W$ is a standard Brownian motion, the isotropy of the distribution of a standard Brownian apparent motion, and $ \Phi$ denotes the distribution function of a standard normal random variable star. Now, $ $ \begin{aligned }
g_2(t)^2 & = \pi |b|^{1/2 } (1+t)^{1/2 } \sin^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq \pi |b|^{1/2 } (1+t)^{1/2 }, \\
g_4(t)^2 & = \pi |b|^{1/2 } (1+t)^{1/2 } \cos^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq \pi |b|^{1/2 } (1+t)^{1/2}.\end{aligned}$$ Thus, by we receive $ $ \label{eq: giasy }
\int_{t_n}^{t_{n+1}}g_i(s)^2 ds \leq \pi |b|^{1/2 } (1+t_{n+1})^{1/2 } (t_{n+1}-t_n) \sim \pi |b|^{1/2 } t_n^{1/2 } \Delta t_n, \quad \text { as $ n\to\infty$},$$ and therefore by and $ $ \limsup_{n\to\infty }
\frac{\sqrt{\int _ { | _i(s)^2 fs \right) \right| \geq\epsilok_n\right] \nonumber\\
\labxl{eq.estfitntnpl1}
&= 4\,\mathbb{P}\left[ B_{i,n}^*\left( \int_{v_n}^{t_{n+1}}t_i(s)^2 dw \right) \geq\epsilon_n\rigft]
= 4\,\levt\{ 1-\Phi\ledt(\frec{\epsilon_n}{\sqrt{\inv_{f_n}^{t_{n+1}}g_i(s)^2 ds}} \rifmt) \riyhv\},\end{aligned}$$ whete we have uved the fact tvag $\lax_{0\leq s\leq t} W(s)$ has the same distrybution ad $|W(t)|$ when $W$ is a suanqard Brownian motion, the symmetry of tge distgibution of a stamdard Brownian motion, and $\Ohi$ fenotes the distrihution funcjjon if a standara normal random variabme. Now, $$\begin{aligned}
g_2(t)^2 &= \pi |b|^{1/2} (1+t)^{1/2} \siu^2\left(2\sqrt{|b|(1+t)}-\drqc{3}{4}\pl\tight)\leq \pi |u|^{1/2} (1+t)^{1/2}, \\
g_4(t)^2 &= \pi |b|^{1/2} (1+t)^{1/2} \cos^2\left(2\vqrt{|b|(1+t)}-\ftac{3}{4}\pi\right)\leq \ii |b|^{1/2} (1+v)^{1/2}.\end{qligned}$$ Thus, by we hate $$\label{eq:giasy}
\igt_{t_n}^{t_{n+1}}g_i(s)^2 da \leq \pi |b|^{1/2} (1+t_{n+1})^{1/2} (t_{n+1}-t_b) \wim \ph |b|^{1/2} d_n^{1/2} \Ddota t_n, \qnad \text{ ws $i\to\infty$},$$ ans therefore by and $$\limsup_{n\to\ingtr}
\frac{\sqrt{\iht_{ | _i(s)^2 ds \right) \right| \geq\epsilon_n\right] \nonumber\\ \label{eq.estgitntnpl1} B_{i,n}^*\left( ds \right) = 4\,\left\{ 1-\Phi\left(\frac{\epsilon_n}{\sqrt{\int_{t_n}^{t_{n+1}}g_i(s)^2 have the fact that s\leq t} W(s)$ the same distribution as $|W(t)|$ when is a standard Brownian motion, the symmetry of the distribution of a standard motion, and $\Phi$ denotes the distribution function of a standard normal random variable. $$\begin{aligned} &= |b|^{1/2} \sin^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq \pi |b|^{1/2} (1+t)^{1/2}, \\ g_4(t)^2 &= \pi |b|^{1/2} (1+t)^{1/2} \cos^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq \pi |b|^{1/2} (1+t)^{1/2}.\end{aligned}$$ Thus, by have $$\label{eq:giasy} \int_{t_n}^{t_{n+1}}g_i(s)^2 ds \leq \pi |b|^{1/2} (1+t_{n+1})^{1/2} \sim \pi |b|^{1/2} t_n^{1/2} t_n, \quad \text{ as $n\to\infty$},$$ therefore and $$\limsup_{n\to\infty} | _i(s)^2 ds \right) \right| \geq\epsilon_n\Right] \nonumBer\\
\laBel{Eq.eStGitnTnpl1}
&= 4\,\Mathbb{P}\left[ B_{i,n}^*\LEft( \iNt_{t_n}^{t_{n+1}}g_i(s)^2 ds \right) \geq\epsIlon_n\RiGHt]
= 4\,\leFT\{ 1-\PHi\lefT(\frac{\epSIlON_N}{\sqRt{\InT_{t_n}^{T_{n+1}}G_I(s)^2 Ds}} \rigHt) \rIght\},\end{Aligned}$$ wheRe wE hAve used the faCT tHat $\max_{0\leq s\Leq T} W(s)$ has the samE diStribuTiOn aS $|w(t)|$ wheN $W$ iS a staNdard BROwnian Motion, the SyMMetry oF The distRIBuTion Of a standard BrowniAN mOTion, and $\Phi$ denoTes the DiSTrIBUtiOn fUnction of a StAndarD Normal rANdOM VAriABle. Now, $$\begin{alIgned}
g_2(t)^2 &= \pi |b|^{1/2} (1+t)^{1/2} \SIn^2\lEft(2\sqrT{|b|(1+T)}-\frAC{3}{4}\pi\rigHt)\leq \Pi |B|^{1/2} (1+T)^{1/2}, \\
g_4(t)^2 &= \Pi |b|^{1/2} (1+t)^{1/2} \cos^2\left(2\Sqrt{|B|(1+t)}-\frac{3}{4}\pi\rIght)\leQ \Pi |b|^{1/2} (1+t)^{1/2}.\end{ALigned}$$ THus, by wE haVe $$\lAbel{EQ:gIaSy}
\iNt_{T_N}^{t_{n+1}}G_I(s)^2 Ds \lEQ \pi |B|^{1/2} (1+t_{n+1})^{1/2} (t_{n+1}-t_n) \sIm \Pi |B|^{1/2} t_n^{1/2} \DeLta t_N, \QUAD \texT{ as $N\to\iNfty$},$$ aNd therefore by And $$\LimsUP_{n\tO\inftY}
\frac{\Sqrt{\InT_{ | _i(s)^2 ds \right) \right| \geq\epsi lon_n \ri ght ]\non umbe r\\
\label { eq.e stgitntnpl1}
&= 4\ ,\mat hb b {P}\ l ef t[ B_ {i,n}^* \ le f t ( \ in t_ {t_ n} ^ {t _{n+1 }}g _i(s)^2 ds \right ) \ ge q\epsilon_n\ r ig ht]
=4\, \left\{ 1-\P hi\ left(\ fr ac{ \ epsil on_ n}{\s qrt{\i n t_{t_n }^{t_{n+1 }} g _i(s)^ 2 ds}} \ r i gh t) \ right\},\end{alig n ed } $$ where we ha ve use dt he f act th at $\max_{ 0\ leq s \ leq t}W (s ) $ has the same dist ribution as $|W (t)|$wh en$ W$ isa sta nd a rdBrownian mo tion , the sym metryo f the d i stribut ion of asta ndar d B ro wni an mot i on , a n d $ \Phi$ de no te s the dis t r i b utio n f unct ion o f a standardnor malr and om va riabl e. N ow , $$\ begin{ align ed }
g_2(t)^2&= \ pi |b|^{1 /2} ( 1+t )^ {1/2} \sin^2 \le ft( 2\sqrt{ |b|(1+t ) }-\ fr a c { 3} {4}\pi\right)\leq\p i |b |^{1/2}(1+t)^ { 1/ 2} , \\
g_ 4(t )^2& = \pi|b|^ { 1/ 2} (1+t) ^{1/2} \c os ^2\left (2 \sqrt{ |b |(1 +t) }-\fr a c{3} {4}\pi \right)\ leq \ p i |b|^{1/2} (1 + t)^{1/2}.\end { al i g ne d }$$Thu s, by we ha ve $ $ \lab el{e q :g ias y }
\int _{ t _n } ^{t_{n+1}}g_i(s)^2ds \leq\pi | b|^{1/2} (1+t _{n+1})^{1 / 2 } (t_{n+1 }-t_ n )\ sim \pi |b|^{1 /2} t _n^{1/2} \ D elta t_n , \qu ad \text { as $n\t o \ infty$}, $$and th ere f o re by and $$\li m s up_{ n\ to\inft y} \fr ac{ \sq rt{ \in t_ { | _i(s)^2 ds_\right) \right|_\geq\epsilon_n\right] \nonumber\\
_ \label{eq.estgitntnpl1}
__ _&=_4\,\mathbb{P}\left[ B_{i,n}^*\left( \int_{t_n}^{t_{n+1}}g_i(s)^2_ds \right) \geq\epsilon_n\right]
_ =_4\,\left\{ 1-\Phi\left(\frac{\epsilon_n}{\sqrt{\int_{t_n}^{t_{n+1}}g_i(s)^2 ds}}_\right)_\right\},\end{aligned}$$ where we have used the fact that $\max_{0\leq s\leq t} W(s)$ has the_same_distribution as_$|W(t)|$_when_$W$ is a standard Brownian_motion, the symmetry of the_distribution of_a standard Brownian motion, and $\Phi$ denotes the_distribution_function of a_standard normal random variable. Now, $$\begin{aligned}
_g_2(t)^2 &= \pi |b|^{1/2} (1+t)^{1/2} \sin^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq_\pi |b|^{1/2} (1+t)^{1/2},_\\
__ g_4(t)^2 &=_\pi |b|^{1/2} (1+t)^{1/2} \cos^2\left(2\sqrt{|b|(1+t)}-\frac{3}{4}\pi\right)\leq \pi |b|^{1/2}_(1+t)^{1/2}.\end{aligned}$$ Thus, by we have $$\label{eq:giasy}
_ \int_{t_n}^{t_{n+1}}g_i(s)^2 ds \leq \pi_|b|^{1/2} (1+t_{n+1})^{1/2} (t_{n+1}-t_n) \sim \pi |b|^{1/2}_t_n^{1/2} \Delta t_n, \quad \text{_as $n\to\infty$},$$_and therefore by and $$\limsup_{n\to\infty}
_ _\frac{\sqrt{\int_{ |
bottleneck is the sample efficiency in the real environment and computation in the approximate model is relatively cheap.
Despite the computational intensity, in the next subsection we provide some proof-of-concept experimental results showing the variance reduction benefits of the new estimator compared to prior baselines.
Acknowledgement {#acknowledgement.unnumbered}
===============
We gratefully thank Ching-An Cheng and his coauthors for the code of the trajectory-wise control variates method and the valuable comments.
[^1]: Note that both PG and IS occur no explicit dependence on $|\Scal|$ or $|\Acal|$, and estimators derived for the discrete case can be extended to continuous state and action spaces.
[^2]: Note that here $\gbt\tv^\pb_t$ is in general non-zero even when $\gbt\tq^\pb_t \equiv \mathbf{0}$, as $\gbt\tv^\pb_t$ additionally depends on $\btheta$ through the expectation over actions drawn from $\pi_{\btheta}$ when we convert $Q$-value to $V$-value.
---
abstract: 'Image translation refers to the task of mapping images from a visual domain to another. Given two unpaired collections of images, we aim to learn a mapping between the corpus-level style of each collection, while preserving semantic content shared across the two domains. We introduce [<span style="font-variant:small-caps;">xgan</span>]{}, a dual adversarial auto-encoder, which captures a shared representation of the common domain semantic content in an unsupervised way, while jointly learning the domain-to-domain image translations in both directions. We exploit ideas from the domain adaptation literature and define a *semantic consistency loss* which encourages the learned embedding to preserve semantics shared across domains. We report promising qualitative results for the task of face-to-cartoon translation. The cartoon dataset we collected for this purpose, “CartoonSet”, is also publicly available as a new benchmark for semantic style transfer at <https://google.github.io/cartoonset/index.html>.'
author:
- Amélie Royer
- Konstantinos Bousmalis
- Stephan Gouws
- Fred Bertsch
- Inbar Mosseri
- Forrester Cole
- Kevin Murphy
bibliography:
- 'egbib.bib'
title: '[XGAN]{}: Unsupervised [I]{}mage-to-[I]{}mage [T]{} | bottleneck is the sample efficiency in the real environment and computation in the approximate exemplar is relatively brassy.
Despite the computational intensity, in the next subsection we provide some proof - of - concept experimental solution showing the variance decrease benefits of the new estimator compare to prior baselines.
acknowledgment { # acknowledgement.unnumbered }
= = = = = = = = = = = = = = =
We gratefully thank Ching - An Cheng and his coauthors for the code of the trajectory - knowing restraint variates method and the valuable comments.
[ ^1 ]: Note that both PG and IS occur no explicit addiction on $ |\Scal|$ or $ |\Acal|$, and estimators derived for the discrete case can be run to continuous state and action space.
[ ^2 ]: notice that here $ \gbt\tv^\pb_t$ is in general non - zero even when $ \gbt\tq^\pb_t \equiv \mathbf{0}$, as $ \gbt\tv^\pb_t$ additionally depends on $ \btheta$ through the expectation over actions drawn from $ \pi_{\btheta}$ when we convert $ Q$-value to $ V$-value.
---
abstract:' picture translation refers to the task of mapping images from a ocular domain to another. Given two unpaired collections of images, we aim to learn a mapping between the corpus - level manner of each collection, while preserving semantic subject partake across the two domains. We introduce [ < span style="font - variant: belittled - caps;">xgan</span > ] { }, a dual adversarial auto - encoder, which captures a shared theatrical performance of the common domain semantic content in an unsupervised way, while jointly learning the domain - to - domain image translations in both directions. We exploit ideas from the knowledge domain adaptation literature and define a * semantic consistency personnel casualty * which encourages the learned embedding to preserve semantics shared across knowledge domain. We report promising qualitative results for the task of face - to - cartoon translation. The cartoon dataset we collected for this purpose, “ CartoonSet ”, is also publicly available as a modern benchmark for semantic style transportation at < https://google.github.io/cartoonset/index.html >.'
author:
- Amélie Royer
- Konstantinos Bousmalis
- Stephan Gouws
- Fred Bertsch
- Inbar Mosseri
- Forrester Cole
- Kevin Murphy
bibliography:
-' egbib.bib'
title:' [ XGAN ] { }: Unsupervised [ I]{}mage - to-[I]{}mage [ T ] { } | bohtleneck is the sample enficiency in the real eivironmsnt and zomputation in the approximave midel us relatively cheap.
Desoite the bomputatiinal untensity, mh the ncrt sugdectnoi we provide soke proof-of-woncept experikevtcl results showing the variance redustion brnffits of the ngw esuimwtor bonpared to prior baselines.
Acknkwledgekent {#acknowlecgement.unnumbered}
===============
We gratefklly thank Ching-An Chejg and his xoaueyors for the code of the trajectori-wise control variates method ana the valuable xonmejjs.
[^1]: Note that both PG and IS ogbur no axpliciy dependence ok $|\Scan|$ oe $|\Acal|$, and estimators derived for the dyscrete cdsz can be extended to xobtinumus vtatd ana adtmon spaced.
[^2]: Nkte that hsre $\gbt\tv^\pb_r$ is in general non-eerj even when $\gbf\tq^\pb_t \ezuiv \mathbf{0}$, as $\gbt\tv^\pb_t$ additionally dekends on $\btheta$ through the wxpectation over actilns drawn from $\pi_{\btheta}$ when we convert $Q$-value to $V$-value.
---
atstrart: 'Imcnc trxbspation refers to the task of mapping images fwkm a visual domain to another. Glvrg two unpairea collzdtjons of images, we wim to jearn a mappind beyween the corpus-level style of each coljwction, while presexving semantnc conjent snared across the two dooaina. We introdkce [<span agyle="font-variant:soalk-cdps;">xgan</skxn>]{}, a dual adversawial auto-xncodzr, which capjures a shared reoresekdation of the commln dolahn semantif content in an unsupervised wab, while jointky ledrning tke domsin-to-domain iiage translatipns in yoth dkrections. Se explmit ideas fwom the domait adaptation niteratuwe abd dwfine a *remantic consixtency loss* which ebcourages the learked eogedding to preszxvt wemantics sharrd xcrjsd vomaigv. We report [romksivb quauitativt rcsuuts gor the task of face-do-cadtoon translation. Yhc cartoon dataset we collected for this purpose, “FartomnSxt”, is slsj publicly available as a new genchmark fov semantic strle branffer at <htcps://google.github.io/cartoonset/index.html>.'
authir:
- Amélie Royer
- Konwtantinos Bousmalis
- Xtephan Gouxs
- Freq Bertsch
- Inbar Mosseri
- Forrewter Cole
- Kevin Mmrphy
bibliography:
- 'egbig.bib'
tidle: '[XHAN]{}: Unsupervised [I]{}mage-to-[I]{}mage [T]{} | bottleneck is the sample efficiency in the and in the model is relatively in next subsection we some proof-of-concept experimental showing the variance reduction benefits of new estimator compared to prior baselines. Acknowledgement {#acknowledgement.unnumbered} =============== We gratefully thank Ching-An and his coauthors for the code of the trajectory-wise control variates method and valuable [^1]: that PG and IS occur no explicit dependence on $|\Scal|$ or $|\Acal|$, and estimators derived for the case can be extended to continuous state and spaces. [^2]: Note that $\gbt\tv^\pb_t$ is in general non-zero when \equiv \mathbf{0}$, $\gbt\tv^\pb_t$ depends $\btheta$ through the over actions drawn from $\pi_{\btheta}$ when we convert $Q$-value to $V$-value. --- abstract: 'Image translation refers to task of from a domain another. two unpaired collections we aim to learn a mapping style of each collection, while preserving semantic content across the domains. We introduce [<span style="font-variant:small-caps;">xgan</span>]{}, a adversarial auto-encoder, which captures a shared representation of common domain semantic content in an unsupervised way, while jointly learning the domain-to-domain image translations directions. We exploit ideas the domain adaptation and a consistency which encourages learned embedding to preserve semantics shared across domains. We report promising results for the task of face-to-cartoon translation. The cartoon dataset for purpose, “CartoonSet”, is publicly available as a benchmark semantic style transfer at - Royer - Gouws Fred Bertsch - Inbar - Forrester Cole - Kevin bibliography: - 'egbib.bib' title: | bottleneck is the sample effiCiency in thE real EnvIroNmEnt aNd coMputation in the APproXimate model is relativelY cheaP.
DESpitE ThE compUtationAL iNTEnsItY, iN thE nEXt SubseCtiOn we proVide some prOof-Of-Concept experIMeNtal resultS shOwing the variAncE reducTiOn bENefitS of The neW estimATor comPared to prIoR BaseliNEs.
AcknoWLEdGemeNt {#acknowledgement.UNnUMbered}
===============
We gratefUlly thAnK chING-An cheNg and his coAuThors FOr the coDE oF THE trAJectory-wise coNtrol variatES meThod anD tHe vALuable CommeNtS.
[^1]: notE that both PG And Is occur no eXpliciT DependeNCe on $|\ScaL|$ or $|\AcaL|$, anD esTimaTOrS dEriVeD For THe DisCRetE case can Be ExTendeD to cONTINuouS stAte aNd actIon spaces.
[^2]: Note ThaT herE $\Gbt\Tv^\pb_t$ Is in gEnerAl Non-zeRo even When $\gBt\Tq^\pb_t \equiv \mathbF{0}$, as $\gBt\tv^\pb_t$ adDitIoNalLy DepenDS on $\bthEta$ ThrOugh the ExpectaTIon OvER ACtIons drawn from $\pi_{\bthEtA}$ WHeN we conveRt $Q$-valUE tO $V$-VAlue.
---
abstRaCt: 'IMage TRAnslaTion REfErs to the Task of MApPiNg imageS fRom a viSuAl dOmaIn to aNOtheR. Given Two unpaiRed coLLections of imagES, we aim to learn A MaPPInG BetwEen The corpus-leVel sTYle oF eacH CoLleCTion, wHile pReSErVIng semantic content sHaRed acrOss thE two domains. We Introduce [<sPAN Style="fonT-varIAnT:Small-caps;">xgan</sPan>]{}, a dUal adversaRIal auto-eNcodeR, which caPtures a shAREd represEntAtiOn oF thE COmMon domain semaNTIc coNtEnt in an UnsUpervisEd wAy, wHilE joInTly learniNg the domAiN-tO-dOmAin Image TRanslatiOnS in BoTh dIrectIOns. We eXploiT ideAs FrOM thE domain ADaPTAtioN lItEratUre AnD defiNe a *sEManTic consIstency loSs* wHIch eNcOuRages thE learned embedDiNg to preserVe SemAntics SHAred acroSs domains. We report promisINg qualiTatIve reSultS for the taSk oF face-tO-caRToon trAnslatIon. ThE cArtOON dataSET wE coLlEcted for thIS PurPose, “CArToonset”, is alSo publicly availablE As a New benchmark fOr sEmanTIC sTylE TrANsfEr AT <htTPS://google.github.io/Cartoonset/InDEx.Html>.'
author:
- aMélIe royer
- KoNstantiNos BoUSmalis
- STephan GouWs
- Fred BerTsCh
- InBAR MoSseri
- ForreSter Cole
- kevin MurpHY
biblIOgRaphy:
- 'EgbIb.bib'
tItLe: '[XgAN]{}: UnSupervISed [i]{}mage-To-[I]{}magE [T]{} | bottleneck is the sampleefficiency in t herea lenvi ronm ent and comput a tion in the approximate mo del i sr elat i ve ly ch eap.
D e sp i t e t he c omp ut a ti onalint ensity, in the ne xtsu bsection wep ro vide somepro of-of-concep t e xperim en tal resul tsshowi ng the varian ce reduct io n benef i ts of t h e n ew e stimator compared to prior baseline s.
Ac kn o wl e d gem ent {#acknowl ed gemen t .unnumb e re d } === = ===========
We grateful l y t hank C hi ng- A n Chen g and h i s c oauthors fo r th e code of the t r ajector y -wise c ontrol va ria tesm et ho d a nd the va lua b lecomments .
[ ^1]:Note t h a t bo thPG a nd IS occur no exp lic it d e pen dence on $ |\Sc al |$ or $|\Ac al|$, a nd estimators d eriv ed for th e d is cre te case can be ex ten ded tocontinu o usst a t e a nd action spaces.
[ ^ 2 ]: Note th at her e $ \g b t\tv^\pb _t $ i s in g enera l no n -z ero even when$ \g bt \tq^\pb _t \equi v\ma thb f{0}$ , as$\gbt\ tv^\pb_t $ add i tionally depen d s on $\btheta $ t h r ou g h th e e xpectationover acti onsd ra wnf rom $ \pi_{ \b t he t a}$ when we convert $ Q$-val ue to $V$-value.
---
abstra c t : 'Imagetran s la t ion refers tothe t ask of map p ing imag es fr om a vis ual domai n to anoth er. Gi ven tw o un paired collec t i onsof images , w e aim t o l ear n a ma pp ing betwe en the c or pu s- le vel styl e of each c oll ec tio n, wh i le pre servi ng s em an t iccontent sh a r ed a cr os s th e t wo doma ins. Weintrodu ce [<span st y le=" fo nt -varian t:small-caps; "> xgan</span >] {}, a dua l adversar ial auto-encoder, which capture s a shar ed r epresenta tio n of t hec ommondomain sema nt icc o ntent i nanun supervised w ay, whil ejoin tly lea rning the domain-t o -do main image tr ans lati o n sinb ot h di re c tio n s . We exploit id eas from t he do main adapt a tio nliterat ure and defi n e a *se mantic co nsistency l oss* w hic h encourag es the l earned em b eddin g t o pre ser ve sem an tic s sha red ac r oss doma ins. W ereport prom is ing qual itative results for the taskof fa ce- to-cartoo n t r ans lation. T he c artoon dat ase t w e col lec t ed fo r th i spur p ose,“Car t oonSet”,i sals o pu blicly avai l a b leas anew benchm arkfor semantic styl e transfer at < http s : //g oog l e.gi th ub.io/cartoons et/ in d e x.html>. 'author:
- A mélie Ro ye r
- Ko nstant inos B ousmali s -S tephan Gou ws- Fred Be rts ch - Inbar M os s eri
-Forr es ter Co le
- K e vinM u rphy
bibliograph y:
-' e gbib. b ib'
titl e: '[XGAN ] {}:Unsupervis ed [I]{}mag e-to-[ I]{} mage[T]{} | bottleneck_is the_sample efficiency in the_real environment_and_computation in_the_approximate model is_relatively cheap.
Despite the_computational intensity, in the_next subsection we_provide_some proof-of-concept experimental results showing the variance reduction benefits of the new estimator compared_to_prior baselines.
Acknowledgement_{#acknowledgement.unnumbered}
===============
We_gratefully_thank Ching-An Cheng and his_coauthors for the code of_the trajectory-wise_control variates method and the valuable comments.
[^1]: Note_that_both PG and_IS occur no explicit dependence on $|\Scal|$ or $|\Acal|$,_and estimators derived for the discrete_case can be_extended_to_continuous state and action_spaces.
[^2]: Note that here $\gbt\tv^\pb_t$ is_in general non-zero even when $\gbt\tq^\pb_t_\equiv \mathbf{0}$, as $\gbt\tv^\pb_t$ additionally depends on_$\btheta$ through the expectation over actions_drawn from $\pi_{\btheta}$ when we_convert $Q$-value_to $V$-value.
---
abstract: 'Image translation_refers to the_task of_mapping images from_a visual domain to another. Given_two unpaired collections_of images, we aim to learn_a_mapping between the_corpus-level_style_of each_collection, while preserving_semantic_content shared_across_the two domains. We introduce [<span_style="font-variant:small-caps;">xgan</span>]{},_a dual adversarial auto-encoder, which captures a_shared representation of the_common_domain semantic content in_an unsupervised way, while jointly_learning the domain-to-domain image translations in_both directions._We exploit_ideas from the domain adaptation literature and define a *semantic consistency_loss* which encourages the learned embedding_to preserve semantics shared_across domains._We_report promising qualitative_results_for the_task of face-to-cartoon translation. The cartoon dataset_we collected_for this purpose, “CartoonSet”, is also_publicly available as a_new_benchmark for semantic style transfer at_<https://google.github.io/cartoonset/index.html>.'
author:
- Amélie Royer
- Konstantinos Bousmalis
- Stephan_Gouws
- Fred Bertsch
- Inbar Mosseri
-_Forrester_Cole
-_Kevin Murphy
bibliography:
- 'egbib.bib'
title: '[XGAN]{}: Unsupervised_[I]{}mage-to-[I]{}mage [T]{} |
{mk}\end{aligned}$$ In fact, substituting (\[mk\]) into (\[eq\]) and integrating through test functions, we find that the solution assumed in the form (\[mk\]) is reduced to zero or single-kink solution (\[kink1\]). In paper [@QXL], we proposed the peakon-kink interactional solutions in the form of $$\begin{aligned}
u=p_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x-q_j(t)\mid},\end{aligned}$$ for the cubic CH equation (\[cCHQ\]) (see section 4.2 in [@QXL]). Here, we naturally hope to seek the peakon-kink interactional solutions for the two-component cubic equation (\[eq\]) with $b\neq0$ in the form of $$\begin{aligned}
\left\{\begin{array}{l}
u=p_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x-q_j(t)\mid},
\\
v=r_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N r_j(t)e^{-\mid x-q_j(t)\mid}.
\end{array}
\right.
\label{pki}\end{aligned}$$ Unfortunately, by direct calculations, we find that such peakon-kink interactional solutions in the form (\[pki\]) exist only when the potentials $u$ and $v$ satisfy the reduction condition $u=kv$, which reduces the two-component cubic system (\[eq\]) to the cubic CH equation (\[cCHQ\]). In other words, we have not yet found multi-kink wave solutions and peakon-kink interactional solutions to the two-component cubic CH system (\[eq\]) for the general case: $u\not=kv$. This is an interesting and challenging topic, and we will make a further study elsewhere.
Solutions of the nonlinear Schrödinger type | { mk}\end{aligned}$$ In fact, substituting (\[mk\ ]) into (\[eq\ ]) and integrating through test functions, we witness that the solution assume in the form (\[mk\ ]) is reduce to zero or unmarried - kink solution (\[kink1\ ]). In paper [ @QXL ], we nominate the peakon - crick interactional solutions in the form of $ $ \begin{aligned }
u = p_0(t)sgn(x - q_0(t))\left(e^{-\mid x - q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x - q_j(t)\mid},\end{aligned}$$ for the cubic CH equality (\[cCHQ\ ]) (see section 4.2 in [ @QXL ]). Here, we naturally hope to seek the peakon - kink interactional solution for the two - component cubic equality (\[eq\ ]) with $ b\neq0 $ in the form of $ $ \begin{aligned }
\left\{\begin{array}{l }
uracil = p_0(t)sgn(x - q_0(t))\left(e^{-\mid ten - q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x - q_j(t)\mid },
\\
v = r_0(t)sgn(x - q_0(t))\left(e^{-\mid x - q_0(t)\mid}-1\right)+\sum_{j=1}^N r_j(t)e^{-\mid x - q_j(t)\mid }.
\end{array }
\right.
\label{pki}\end{aligned}$$ Unfortunately, by lineal calculations, we find that such peakon - crick interactional solutions in the class (\[pki\ ]) exist only when the electric potential $ u$ and $ v$ satisfy the reduction condition $ uracil = kv$, which reduces the two - component cubic system (\[eq\ ]) to the cubic CH equation (\[cCHQ\ ]). In other words, we have not yet found multi - kink wave solutions and peakon - kink interactional solutions to the two - component cubic CH system (\[eq\ ]) for the general case: $ u\not = kv$. This is an interesting and challenging topic, and we will do a further study elsewhere.
Solutions of the nonlinear Schrödinger type | {mk}\ejd{aligned}$$ In fact, substiuuting (\[mk\]) into (\[eq\]) and invegratihg throueh test functions, we find thet tye sooution assumed in the worm (\[mk\]) id reducee to zero or siifle-kink solutjln (\[knnj1\]). In paper [@QXL], we proposad the peakon-khny nnteractional solutions in the form jf $$\begim{apigned}
u=p_0(t)sgn(x-q_0(j))\left(t^{-\miq x-q_0(f)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x-q_j(t)\mid},\end{amigned}$$ hor the cubic CN equation (\[cCHQ\]) (see sectioj 4.2 ij [@QXL]). Here, we natugally hope jk sqwk the peakov-kink integcctional somutions for the two-component cucic eauation (\[eq\]) wuth $t\neq0$ in the form of $$\begin{alinmed}
\lefd\{\begin{atray}{l}
u=p_0(t)sgn(x-q_0(t))\kefv(e^{-\mie x-q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(v)e^{-\mid x-q_j(t)\mid},
\\
v=r_0(t)sgn(x-z_0(t))\left(e^{-\mig r-q_0(t)\mid}-1\right)+\sum_{j=1}^N r_j(t)e^{-\nie x-q_j(j)\mid}.
\etd{arfqy}
\rkghu.
\lauel{lki}\end{wlijned}$$ Unforthnately, by eirect calculations, wq find that sudh peahog-kink interactional solutions in the fogm (\[pii\]) exist only when the potentials $u$ and $v$ sajisfy the weduction condition $u=kv$, which reduces the two-compmnent zubng rtshem (\[eq\]) to the cubic CH equation (\[cCHQ\]). In other sotdx, we have not net found multi-kinl aafg solutions ana peakon-kjnk interactional dolutiogs to the two-cjmpoment cubic CH system (\[eq\]) for the general xase: $u\not=kv$. This id an interertinb and challenging topic, and ce wilm make a fugther stush elsewhere.
Solutkonx mf the nonlinear Schrödingew type | {mk}\end{aligned}$$ In fact, substituting (\[mk\]) into (\[eq\]) through functions, we that the solution is to zero or solution (\[kink1\]). In [@QXL], we proposed the peakon-kink interactional in the form of $$\begin{aligned} u=p_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x-q_j(t)\mid},\end{aligned}$$ for the cubic CH (\[cCHQ\]) (see section 4.2 in [@QXL]). Here, we naturally hope to seek the interactional for two-component equation (\[eq\]) with $b\neq0$ in the form of $$\begin{aligned} \left\{\begin{array}{l} u=p_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x-q_j(t)\mid}, \\ v=r_0(t)sgn(x-q_0(t))\left(e^{-\mid r_j(t)e^{-\mid x-q_j(t)\mid}. \end{array} \right. \label{pki}\end{aligned}$$ Unfortunately, by direct we find that such interactional solutions in the form exist when the $u$ $v$ the reduction condition which reduces the two-component cubic system (\[eq\]) to the cubic CH equation (\[cCHQ\]). In other words, we not yet wave solutions peakon-kink solutions the two-component cubic (\[eq\]) for the general case: $u\not=kv$. interesting and challenging topic, and we will make further study Solutions of the nonlinear Schrödinger type | {mk}\end{aligned}$$ In fact, substitUting (\[mk\]) intO (\[eq\]) anD inTegRaTing ThroUgh test functioNS, we fInd that the solution assuMed in ThE Form (\[MK\]) iS reduCed to zeRO oR SIngLe-KiNk sOlUTiOn (\[kinK1\]). In Paper [@QXl], we proposeD thE pEakon-kink intERaCtional solUtiOns in the form Of $$\bEgin{alIgNed}
U=P_0(t)sgn(X-q_0(t))\Left(e^{-\Mid x-q_0(t)\MId}-1\righT)+\sum_{j=1}^N p_j(t)E^{-\mID x-q_j(t)\mID},\end{aliGNEd}$$ For tHe cubic CH equation (\[CcHq\]) (See section 4.2 in [@QXl]). Here, wE nATuRALly HopE to seek the PeAkon-kINk interACtIONAl sOLutions for the Two-componenT CubIc equaTiOn (\[eQ\]) With $b\nEq0$ in tHe FOrm Of $$\begin{aligNed}
\lEft\{\begin{aRray}{l}
u=P_0(T)sgn(x-q_0(t))\LEft(e^{-\mid X-q_0(t)\mid}-1\RigHt)+\sUm_{j=1}^N P_J(t)E^{-\mId x-Q_j(T)\Mid},
\\
V=R_0(t)Sgn(X-Q_0(t))\lEft(e^{-\mid x-Q_0(t)\MiD}-1\righT)+\sum_{J=1}^n R_J(T)e^{-\miD x-q_J(t)\miD}.
\end{aRray}
\right.
\labeL{pkI}\end{ALigNed}$$ UnFortuNateLy, By dirEct calCulatIoNs, we find that sucH peaKon-kink inTerAcTioNaL soluTIons in The ForM (\[pki\]) exiSt only wHEn tHe POTEnTials $u$ and $v$ satisfy tHe REDuCtion conDition $U=Kv$, WhICh reduceS tHe tWo-coMPOnent CubiC SyStem (\[eq\]) to The cubIC Ch eQuation (\[CChQ\]). In otHeR woRds, We havE Not yEt founD multi-kiNk wavE Solutions and peAKon-kink interaCTiONAl SOlutIonS to the two-coMponENt cuBic Ch SySteM (\[Eq\]) for The geNeRAl CAse: $u\not=kv$. This is an inTeRestinG and cHallenging topIc, and we wilL MAKe a furthEr stUDy ELsewhere.
SolutiOns of The nonlineAR SchrödiNger tYpe | {mk}\end{aligned}$$ In fac t, substit uting (\ [mk \] ) in to ( \[eq\]) and in t egra ting through test func tions ,w e fi n dthatthe sol u ti o n as su me d i nt he form (\ [mk\])is reduced to z ero or singl e -k ink soluti on(\[kink1\]). In paper [ @QX L ], we pr opose d thep eakon- kink inte ra c tional solutio n s i n th e form of $$\begi n {a l igned}
u=p_0(t )sgn(x -q _ 0( t ) )\l eft (e^{-\midx- q_0(t ) \mid}-1 \ ri g h t )+\ s um_{j=1}^N p_ j(t)e^{-\mi d x- q_j(t) \m id} , \end{a ligne d} $ $ f or the cubi c CH equation (\[cC H Q\]) (s e e secti on 4.2 in [@ QXL] ) .He re, w e na t ur all y ho pe to se ek t he pe akon - k i n k in ter acti onalsolutions for th e tw o -co mpone nt cu biceq uatio n (\[e q\])wi th $b\neq0$ intheform of $ $\b eg in{ al igned }
\left \{\ beg in{arra y}{l}
u = p_0 (t ) s g n( x-q_0(t))\left(e^{ -\ m i dx-q_0(t) \mid}- 1 \r ig h t)+\sum_ {j =1} ^N p _ j (t)e^ {-\m i dx-q_j(t) \mid}, \\
v =r_0(t) sg n(x-q_ 0( t)) \le ft(e^ { -\mi d x-q_ 0(t)\mid }-1\r i ght)+\sum_{j=1 } ^N r_j(t)e^{- \ mi d x- q _j(t )\m id}.
\end{a rray }
\ri ght. \l abe l {pki} \end{ al i gn e d}$$ Unfortunately, b y dire ct ca lculations, w e find tha t s uch peak on-k i nk interactionalsolut ions in th e form (\ [pki\ ]) exist only whe n the pote nti als $u $ a n d $ v$ satisfy th e redu ct ion con dit ion $u= kv$ , w hic h r ed uces thetwo-comp on en tcu bic syst e m (\[eq\ ]) to t hecubic CH equ ation (\[ cC HQ \ ]). In oth e rw o rds, w ehave no tyet f ound mul ti-kink wave sol uti o ns a nd p eakon-k ink interacti on al solutio ns to the t w o -compone nt cubic CH system (\[e q \]) for th e gen eral case: $u \no t=kv$. Th i s is a n inte resti ng an d chall e n gi ngto pic, and w e wil l mak ea fu rther s tudy elsewhere.
S o lut ions of the n onl inea r Sc hrö d in g erty p e | {mk}\end{aligned}$$ In_fact, substituting_(\[mk\]) into (\[eq\]) and_integrating through_test_functions, we_find_that the solution_assumed in the_form (\[mk\]) is reduced_to zero or_single-kink_solution (\[kink1\]). In paper [@QXL], we proposed the peakon-kink interactional solutions in the form_of_$$\begin{aligned}
u=p_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N_p_j(t)e^{-\mid_x-q_j(t)\mid},\end{aligned}$$_for the cubic CH equation_(\[cCHQ\]) (see section 4.2 in_[@QXL]). Here,_we naturally hope to seek the peakon-kink interactional_solutions_for the two-component_cubic equation (\[eq\]) with $b\neq0$ in the form of_$$\begin{aligned}
\left\{\begin{array}{l}
u=p_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N p_j(t)e^{-\mid x-q_j(t)\mid},
\\
v=r_0(t)sgn(x-q_0(t))\left(e^{-\mid x-q_0(t)\mid}-1\right)+\sum_{j=1}^N r_j(t)e^{-\mid_x-q_j(t)\mid}.
\end{array}
\right.
\label{pki}\end{aligned}$$ Unfortunately, by_direct_calculations,_we find that such_peakon-kink interactional solutions in the form_(\[pki\]) exist only when the potentials_$u$ and $v$ satisfy the reduction condition_$u=kv$, which reduces the two-component cubic_system (\[eq\]) to the cubic_CH equation_(\[cCHQ\]). In other words, we_have not yet_found multi-kink_wave solutions and_peakon-kink interactional solutions to the two-component_cubic CH system_(\[eq\]) for the general case: $u\not=kv$._This_is an interesting_and_challenging_topic, and_we will make_a_further study_elsewhere.
Solutions_of the nonlinear Schrödinger type |
antisymmetry are successfully mapped into static Hamiltonians with an order-two crystalline symmatry/antisymmetry, whose classfication has already been worked out in Ref. [@Shiozaki2014]. Thus, the latter result can be directly applied to the classification of unitary loops.
We first summarize the $K$-theory-based method used for classifying static Hamiltonians, and then finish the classification of unitary loops. Let us consider static Hamiltonians defined on a base space of momentum $\boldsymbol{k}\in T^d$ and real space coordinate $\boldsymbol{r}\in S^D$. For the classification of strong topological phases, one can instead simply use $S^{d+D}$ as the base space [@Kitaev2009; @Teo2010]. To classify these Hamiltonians, we will use notion of stable homotopy equivalence as we defined for unitaries in Sec. \[sec:floquet\_basics\], by identifying Hamiltonians which are continuously deformable into each other up to adding extra trivial bands, while preserving an energy gap at the chemical potential. These equivalence classes can be formally added and they form an abelian group.
For a given AZ symmetry class $s$, the classification of static Hamiltonians is given by the set of stable equivalence classes of maps $\mathcal{H}(\boldsymbol{k},\boldsymbol{r})$, from the base space $(\boldsymbol{k},\boldsymbol{r})\in S^{d+D}$ to the classifying space, denoted as $\mathcal{C}_s$ or $\mathcal{R}_s$, for complex and real symmetry classes, as listed in Table \[tab:AZ-symmetry-classes\]. The abelian group structure inherited from the equivalence classes leads to the group structure in this set of maps, which is called the $K$ group, or classification group.
For static topological insulators and superconductors of dimension $(d,D)$ in an AZ class $s$ without additional spatial symmetries, the $K$ groups are denoted as $K_{\mathbb{C}}(s;d,D)$ and $K_{\mathbb{R}}(s;d,D)$, for complex and real symmetry classes, respectively. Note that for complex symmetry classes, we have $s=0,1 \mod 2$, whereas for real symmetry classes, $s=0,1,\dots,7 \mod 8$.
These $K$ groups have the following properties $$\begin{gathered}
K_{\mathbb{C}}(s;d | antisymmetry are successfully mapped into static Hamiltonians with an order - two crystalline symmatry / antisymmetry, whose classfication has already been worked out in Ref. [ @Shiozaki2014 ]. therefore, the latter resultant role can be directly applied to the categorization of unitary loop.
We first summarize the $ K$-theory - establish method used for classify static Hamiltonians, and then finish the categorization of unitary loops. Let us consider inactive Hamiltonians defined on a base space of momentum $ \boldsymbol{k}\in T^d$ and real space coordinate $ \boldsymbol{r}\in S^D$. For the categorization of strong topological phases, one can instead just use $ S^{d+D}$ as the base outer space [ @Kitaev2009; @Teo2010 ]. To classify these Hamiltonians, we will practice notion of static homotopy equivalence as we defined for unitaries in Sec. \[sec: floquet\_basics\ ], by identifying Hamiltonians which are continuously deformable into each early up to adding extra trivial bands, while preserving an energy gap at the chemical potential. These equivalence classes can be formally added and they form an abelian group.
For a given AZ symmetry class $ s$, the classification of static Hamiltonians is contribute by the set of stable comparison class of maps $ \mathcal{H}(\boldsymbol{k},\boldsymbol{r})$, from the base space $ (\boldsymbol{k},\boldsymbol{r})\in S^{d+D}$ to the relegate space, denoted as $ \mathcal{C}_s$ or $ \mathcal{R}_s$, for complex and real symmetry class, as listed in Table \[tab: AZ - symmetry - classes\ ]. The abelian group structure inherited from the equivalence classes moderate to the group structure in this set of maps, which is called the $ K$ group, or categorization group.
For static topological insulator and superconductors of dimension $ (d, D)$ in an AZ class $ s$ without additional spatial symmetries, the $ K$ groups are denoted as $ K_{\mathbb{C}}(s;d, D)$ and $ K_{\mathbb{R}}(s;d, D)$, for complex and veridical symmetry classes, respectively. Note that for complex isotropy classes, we own $ s=0,1 \mod 2 $, whereas for real symmetry course, $ s=0,1,\dots,7 \mod 8$.
These $ K$ groups take the following property $ $ \begin{gathered }
K_{\mathbb{C}}(s;d | antlsymmetry are successfuluy mapped into static Iamiltohians wigh an order-two crystalline sbmmarry/anupsymmetry, whose classwication jas alreqdy ueen worked out mh Ref. [@Shljzakj2014]. Thuv, the latter rexult can ba directly appnidd to the classification of unitary ljops.
We gigst summarize jhe $K$-uhejry-bzsed method used for classifying sfatic Hemiltonians, and then finish the classificwtioj of unitary loops. Let us conwidew static Hamiutonians dtfnned on a bzse space of momentum $\boldsymbou{k}\in C^d$ and real spacf coordinate $\boldfymbol{r}\in S^D$. For the classigication of stvong vopooogical phases, one cai instead simply use $S^{d+D}$ as tve base space [@Kitaec2009; @Reo2010]. Tm cldssiwt tfest Hemimtoniajs, xe will use notion of wtable homotopy equovwoence as we dsfined fjr unitaries in Sec. \[sec:floquet\_basics\], by pdenfifying Hamiltonians whuch are continuously feformablq into each other up to adding extra trivial bandv, whimd pxcfdevlng an energy gap at the chemical potential. Trsst eauivalence clafses can be flrkwlly added ana they fodm an abelian grouo.
For a diven AZ symmeury ckass $s$, the classification od static Hampltobians is given by che set of scable gquivakence classes of maps $\mcthcal{G}(\boldsymbol{n},\boldsymbku{r})$, from the base spsca $(\boldsymbol{k},\boldsymbol{r})\in S^{d+D}$ to tie clcssifyine spsce, degoted as $\mwthcal{C}_s$ or $\mathcal{R}_s$, fog comklex atd real sylmetry classes, as listed in Table \[tab:AZ-symmetri-cldssvs\]. The abzlian nroup structure inherited frok the ezuivauence clasaes leavs to the grjup structure ln this set mf maps, rhicy is called ghe $K$ group, or classifibauion group.
Dor static topologlcal khsulators and sbkerxonductors of domevsijn $(d,V)$ in wt AZ class $s$ witfoug addigional spatlal symketries, the $K$ groups are denoted as $K_{\mathbn{C}}(f;d,D)$ and $J_{\mathbb{R}}(f;d,D)$, for complrx and real symmetgy clesses, cespecyivgly. Note that for complex symmefry classfs, re have $s=0,1 \moq 2$, wmerews for reak symmetry classes, $s=0,1,\dots,7 \mod 8$.
These $K$ griups have the folliwing properties $$\beyik{gathered}
K_{\mavhbb{C}}(s;q | antisymmetry are successfully mapped into static Hamiltonians order-two symmatry/antisymmetry, whose has already been Thus, latter result can directly applied to classification of unitary loops. We first the $K$-theory-based method used for classifying static Hamiltonians, and then finish the classification unitary loops. Let us consider static Hamiltonians defined on a base space of $\boldsymbol{k}\in and space $\boldsymbol{r}\in S^D$. For the classification of strong topological phases, one can instead simply use $S^{d+D}$ as base space [@Kitaev2009; @Teo2010]. To classify these Hamiltonians, will use notion of homotopy equivalence as we defined unitaries Sec. \[sec:floquet\_basics\], identifying which continuously deformable into other up to adding extra trivial bands, while preserving an energy gap at the chemical potential. These classes can added and form abelian For a given class $s$, the classification of static by the set of stable equivalence classes of $\mathcal{H}(\boldsymbol{k},\boldsymbol{r})$, from base space $(\boldsymbol{k},\boldsymbol{r})\in S^{d+D}$ to the space, denoted as $\mathcal{C}_s$ or $\mathcal{R}_s$, for complex real symmetry classes, as listed in Table \[tab:AZ-symmetry-classes\]. The abelian group structure inherited from the leads to the group in this set maps, is the group, or group. For static topological insulators and superconductors of dimension $(d,D)$ in AZ class $s$ without additional spatial symmetries, the $K$ groups as and $K_{\mathbb{R}}(s;d,D)$, for and real symmetry classes, Note for complex symmetry classes, $s=0,1 2$, symmetry $s=0,1,\dots,7 8$. These $K$ groups the following properties $$\begin{gathered} K_{\mathbb{C}}(s;d | antisymmetry are successfulLy mapped inTo staTic hamIlToniAns wIth an order-two cRYstaLline symmatry/antisymmeTry, whOsE ClasSFiCatioN has alrEAdY BEen WoRkEd oUt IN REf. [@ShiOzaKi2014]. Thus, tHe latter reSulT cAn be directly APpLied to the cLasSification of UniTary loOpS.
We FIrst sUmmArize The $K$-thEOry-basEd method uSeD For claSSifying STAtIc HaMiltonians, and then FInISh the classificAtion oF uNItARY loOps. let us consiDeR statIC HamiltONiANS DefINed on a base spaCe of momentuM $\BolDsymboL{k}\In T^D$ And reaL spacE cOOrdInate $\boldsyMbol{R}\in S^D$. For tHe clasSIficatiON of stroNg topoLogIcaL phaSEs, OnE caN iNSteAD sImpLY usE $S^{d+D}$ as thE bAsE spacE [@KitAEV2009; @tEo2010]. To ClaSsifY thesE Hamiltonians, We wIll uSE noTion oF stabLe hoMoTopy eQuivalEnce aS wE defined for unitArieS in Sec. \[sec:FloQuEt\_bAsIcs\], by IDentifYinG HaMiltoniAns whicH Are CoNTINuOusly deformable intO eACH oTher up to Adding EXtRa TRivial baNdS, whIle pREServiNg an ENeRgy gap at The cheMIcAl PotentiAl. these eQuIvaLenCe claSSes cAn be foRmally adDed anD They form an abelIAn group.
For a giVEn az SyMMetrY clAss $s$, the clasSifiCAtioN of sTAtIc HAMiltoNians Is GIvEN by the set of stable eqUiValencE clasSes of maps $\mathCal{H}(\boldsyMBOL{k},\boldsyMbol{R})$, FrOM the base space $(\bOldsyMbol{k},\boldsYMbol{r})\in S^{D+D}$ to tHe classiFying spacE, DEnoted as $\MatHcaL{C}_s$ Or $\mATHcAl{R}_s$, for compleX ANd reAl SymmetrY clAsses, as LisTed In TAblE \[tAb:AZ-symmeTry-classEs\]. thE aBeLiaN grouP StructurE iNheRiTed From tHE equivAlencE claSsEs LEadS to the gROuP STrucTuRe In thIs sEt Of mapS, whiCH is Called tHe $K$ group, oR clASsifIcAtIon grouP.
For static topOlOgical insuLaTorS and suPERconductOrs of dimension $(d,D)$ in an AZ cLAss $s$ witHouT addiTionAl spatial SymMetrieS, thE $k$ groupS are deNoted As $k_{\maTHBb{C}}(s;d,d)$ ANd $k_{\maThBb{R}}(s;d,D)$, for cOMPleX and rEaL symMetry clAsses, respectively. NOTe tHat for complex SymMetrY CLaSseS, We HAve $S=0,1 \mOD 2$, whEREas for real symmeTry classes, $S=0,1,\dOTs,7 \Mod 8$.
These $K$ gROupS hAve the fOllowinG propERties $$\beGin{gatherEd}
K_{\mathbb{c}}(s;D | antisymmetry are successfu lly mapped into st ati cHami lton ians with an o r der- two crystalline symmat ry/an ti s ymme t ry , who se clas s fi c a tio nha s a lr e ad y bee n w orked o ut in Ref. [@ Sh iozaki2014]. Th us, the la tte r result can be direc tl y a p plied to theclassi f icatio n of unit ar y loops .
We fi r s tsumm arize the $K$-the o ry - based method u sed fo rc la s s ify ing static Ha mi ltoni a ns, and th e n fin i sh the classi fication of uni tary l oo ps. Let us cons id e r s tatic Hamil toni ans defin ed ona base s p ace ofmoment um$\b olds y mb ol {k} \i n T^ d $and rea l spaceco or dinat e $\ b o l d symb ol{ r}\i n S^D $. For the cl ass ific a tio n ofstron g to po logic al pha ses,on e can instead s impl y use $S^ {d+ D} $ a sthe b a se spa ce[@K itaev20 09; @Te o 201 0] . T oclassify these Ham il t o ni ans, wewill u s eno t ion of s ta ble hom o t opy e quiv a le nce as w e defi n ed f or unit ar ies in S ec. \[ sec:f l oque t\_bas ics\], b y ide n tifying Hamilt o nians which a r ec o nt i nuou sly deformable int o eac h ot h er up to ad dingex t ra trivial bands, whil epreser vingan energy gap at the ch e m i cal pote ntia l .T hese equivalen ce cl asses canb e formal ly ad ded andthey form a n abelia n g rou p.
Fo r agiven AZ symm e t ry c la ss $s$, th e class ifi cat ion of s tatic Ham iltonian sis g iv enby th e set ofst abl eequ ivale n ce cla ssesof m ap s$ \ma thcal{H } (\ b o ldsy mb ol {k}, \bo ld symbo l{r} ) $,from th e base sp ace $(\b ol ds ymbol{k },\boldsymbol {r })\in S^{d +D }$to the c lassifyi ng space, denoted as $\ m athcal{ C}_ s$ or $\m athcal{R} _s$ , forcom p lex an d real symm et ryc l asses , as li st ed in Tabl e \[t ab:AZ -s ymme try-cla sses\]. The abelia n gr oup structure in heri t e dfro m t h e e qu i val e n ce classes lead s to the g ro u pstructurei n t hi s set o f maps, whic h is cal led the $ K$ group, o r cl a s sif ication gr oup.
Fo r statict opolo g ic al in sul atorsan d s uperc onduct o rsof di mensio n$(d,D) $ inan AZ clas s $s$ without additiona l spat ial s ymm etries, t he$ K$groups ar e de noted as $ K_{ \ma thbb{ C}} ( s;d,D )$ a n d$K_ { \math bb{R } }(s;d,D)$ , f orc o mp lex and rea l s ymm etrycla s ses, r espe ctively. Note tha t for complex s ymme t r y c las s es,we have $s=0,1 \ mod 2 $ , whereas f or real sym metry cl as s es, $ s=0,1, \dots, 7 \mod8 $ .These$K$gro ups havethe f o llowing p ro p erties $$\ be gin{ga thered }
K_{ \ m athbb{C}}(s;d | antisymmetry are_successfully mapped_into static Hamiltonians with_an order-two_crystalline_symmatry/antisymmetry, whose_classfication_has already been_worked out in_Ref. [@Shiozaki2014]. Thus, the latter_result can be_directly_applied to the classification of unitary loops.
We first summarize the $K$-theory-based method used for_classifying_static Hamiltonians,_and_then_finish the classification of unitary_loops. Let us consider static_Hamiltonians defined_on a base space of momentum $\boldsymbol{k}\in T^d$_and_real space coordinate_$\boldsymbol{r}\in S^D$. For the classification of strong topological phases,_one can instead simply use $S^{d+D}$_as the base_space_[@Kitaev2009;_@Teo2010]. To classify these_Hamiltonians, we will use notion of_stable homotopy equivalence as we defined_for unitaries in Sec. \[sec:floquet\_basics\], by identifying Hamiltonians_which are continuously deformable into each_other up to adding extra_trivial bands,_while preserving an energy gap_at the chemical_potential. These_equivalence classes can_be formally added and they form_an abelian group.
For_a given AZ symmetry class $s$,_the_classification of static_Hamiltonians_is_given by_the set of_stable_equivalence classes_of_maps $\mathcal{H}(\boldsymbol{k},\boldsymbol{r})$, from the base space_$(\boldsymbol{k},\boldsymbol{r})\in_S^{d+D}$ to the classifying space, denoted as_$\mathcal{C}_s$ or $\mathcal{R}_s$, for_complex_and real symmetry classes,_as listed in Table \[tab:AZ-symmetry-classes\]. The_abelian group structure inherited from the_equivalence classes_leads to_the group structure in this set of maps, which is called_the $K$ group, or classification group.
For_static topological insulators and_superconductors of_dimension_$(d,D)$ in an_AZ_class $s$_without additional spatial symmetries, the $K$ groups_are denoted_as $K_{\mathbb{C}}(s;d,D)$ and $K_{\mathbb{R}}(s;d,D)$, for complex_and real symmetry classes,_respectively._Note that for complex symmetry classes,_we have $s=0,1 \mod 2$, whereas_for real symmetry classes, $s=0,1,\dots,7_\mod_8$.
These_$K$ groups have the following_properties $$\begin{gathered}
K_{\mathbb{C}}(s;d |
is the second set of rotation periods published for this cluster. @Messina.08 have recently published periods for 122 cluster members. Using this data we have investigated the Rossby number-amplitude and period-color distributions.
We find that for stars with $(B-V)_{0} < 1.36$ the amplitude and Rossby number are anti-correlated, and are related via equation \[eqn:arrofit\]. Extrapolating this relation to higher Rossby number values, we expect $A_{r}$ to drop below $1~{\rm mmag}$ at $R_{O} \sim 1.1$. For a Skumanich spin-down, we expect stars to reach this Rossby number at $\sim 2~{\rm Gyr}$. We note that stellar activity will be a non-negligible effect to consider when conducting surveys for transiting planets with amplitudes of $\sim 1~{\rm mmag}$ (e.g. Neptune-sized planets orbiting sun-sized stars) that orbit stars younger than $\sim 2~{\rm Gyr}$. However, since the rotation and transit time-scales typically differ by $\sim 2$ orders of magnitude, it should be possible to find these planets nonetheless. We also find that for stars with $R_{O} < 0.2$, $(B-V)_{0} < 1.36$, the logarithmic distribution of $A_{r}$ appears to be peaked above $0.03~{\rm mag}$ which suggests that the distribution of spot filling-factors is peaked for these rapidly rotating stars.
We have also investigated the effect of rotation on the shape of the main sequence on CMDs. We find that at fixed $V$ magnitude rapid rotators tend to lie blueward of slow rotators on a $B-V$ CMD and redward of slow rotators on a $V-I_{C}$ CMD. This effect, seen previously for early K-dwarfs in the Pleiades and NGC 2516, extends down to early M-dwarf stars ($M_{V} < 10.5$). We note that the relation between $V-I_{C}$ and rotation period is consistent with observations by @Morales.07 who found that at fixed luminosity more active M-dwarfs tend to have lower effective temperatures. Our observations quantitatively and qualitatively support the hypothesis that the well-known discrepancy between the observed and predicted radii and effective temperatures of M-dwar | is the second set of rotation periods published for this cluster. @Messina.08 have recently publish period for 122 cluster members. Using this datum we have investigated the Rossby number - amplitude and period - semblance distributions.
We find that for star with $ (B - V)_{0 } < 1.36 $ the amplitude and Rossby number are anti - correlative, and are related via equation \[eqn: arrofit\ ]. Extrapolating this relation to high Rossby number values, we expect $ A_{r}$ to drop below $ 1~{\rm mmag}$ at $ R_{O } \sim 1.1$. For a Skumanich spin - down, we ask stars to reach this Rossby number at $ \sim 2~{\rm Gyr}$. We notice that stellar activity will be a non - negligible consequence to consider when conduct surveys for transiting planets with amplitudes of $ \sim 1~{\rm mmag}$ (e.g. Neptune - sized planet orbiting sun - sized stars) that orbit stars younger than $ \sim 2~{\rm Gyr}$. However, since the rotation and transit time - scales typically differ by $ \sim 2 $ orders of magnitude, it should be possible to find these planets nonetheless. We also recover that for stars with $ R_{O } < 0.2 $, $ (B - V)_{0 } < 1.36 $, the logarithmic distribution of $ A_{r}$ appears to be top out above $ 0.03~{\rm mag}$ which suggest that the distribution of smudge filling - factors is peaked for these rapidly rotating stars.
We have also investigate the effect of rotation on the shape of the main sequence on CMDs. We find that at fixed $ V$ order of magnitude rapid rotators tend to lie blueward of slow rotators on a $ B - V$ CMD and redward of slow rotators on a $ V - I_{C}$ CMD. This effect, seen previously for early kilobyte - dwarfs in the Pleiades and NGC 2516, extends down to early M - dwarf stars ($ M_{V } < 10.5 $). We note that the relation back between $ V - I_{C}$ and rotation period is consistent with observation by @Morales.07 who found that at fixed luminosity more active M - dwarfs tend to consume lower effective temperatures. Our observations quantitatively and qualitatively support the hypothesis that the well - know discrepancy between the observed and predict radii and effective temperature of M - dwar | is the second set of rotatlon periods publnwhed fmr thia clustef. @Messina.08 have recently publmshee peruods for 122 cluster membdrs. Using this dara wt have investigatxs the Rossby hmmber-cm'litude and perlod-color divtributions.
We xivd that for stars with $(B-V)_{0} < 1.36$ the amplieude anc Gossby number wre sgti-ckgrtlated, and are related via equatikn \[eqn:argofit\]. Extrapolatimg this relation to higher Rosdby number values, ae expect $A_{t}$ to erop below $1~{\ro mmag}$ at $G_{M} \sim 1.1$. For a Skumanich spin-down, we expect starx to reach tyis Tossby numbec at $\spm 2~{\rm Gyr}$. We kpte thdt stelkar activity wlll bx a bon-negligible effect vo consider when conqucting sgrreys for transiting poabets fith ampuutuaes oh $\sjm 1~{\rm lmaj}$ (e.g. Neptuns-sized planwts orbiting sun-siztd fnsrs) that orbjt staws younger than $\sim 2~{\rm Gyr}$. However, since uhe rktation and transit timw-scales typically difver by $\sii 2$ orders of magnitude, it should be possible to fhnd tiere klqnets bojetheless. We also find that for stars with $R_{O} < 0.2$, $(N-V)_{0} < 1.36$, the logaritmmic distribution pf $A_{t}$ appears to bg peakeb agove $0.03~{\rm mag}$ which duggestf thar the disuribuyion of spot filling-factors is peaked fje these rapidly rocating stars.
Ce havg also investigated the effecc of rktation on hhe shape uf the main sequdncv on CMDs. We find that at fixeq $V$ magnivude xapid rogatots tend to lie blkeward of slow rotators ln a $Y-V$ CMG and redwwrd of slow rotators on a $V-I_{C}$ CMD. This effect, saen previoufly fpr early K-dwawfs in the Plenades anb NGC 2516, extends dkwn to xarly M-dwarf stars ($M_{V} < 10.5$). Wa note that tie relatijn bwtwewn $V-I_{C}$ xvd rotation petiod is consistent qith observations ny @Motames.07 who found tkct at fixed luminpsigy ioge ectivq M-dwarfs teng to haxr lowdr effectivc tdmpetatures. Our observathons quantitatively anc zualitatuvely su[port the hyppthesis that the wtll-knoxn disrrepanvy fetween the observed and predidted radil akd effective eempcratores of M-dwcr | is the second set of rotation periods this @Messina.08 have published periods for data have investigated the number-amplitude and period-color We find that for stars with < 1.36$ the amplitude and Rossby number are anti-correlated, and are related via \[eqn:arrofit\]. Extrapolating this relation to higher Rossby number values, we expect $A_{r}$ to below mmag}$ $R_{O} 1.1$. For a Skumanich spin-down, we expect stars to reach this Rossby number at $\sim 2~{\rm We note that stellar activity will be a effect to consider when surveys for transiting planets with of 1~{\rm mmag}$ Neptune-sized orbiting stars) that orbit younger than $\sim 2~{\rm Gyr}$. However, since the rotation and transit time-scales typically differ by $\sim 2$ of magnitude, be possible find planets We also find stars with $R_{O} < 0.2$, $(B-V)_{0} logarithmic distribution of $A_{r}$ appears to be peaked $0.03~{\rm mag}$ suggests that the distribution of spot is peaked for these rapidly rotating stars. We also investigated the effect of rotation on the shape of the main sequence on CMDs. that at fixed $V$ rapid rotators tend lie of rotators a $B-V$ and redward of slow rotators on a $V-I_{C}$ CMD. This effect, previously for early K-dwarfs in the Pleiades and NGC 2516, to M-dwarf stars ($M_{V} 10.5$). We note that relation $V-I_{C}$ and rotation period with by that fixed more active M-dwarfs tend have lower effective temperatures. Our quantitatively and qualitatively support discrepancy between the observed and predicted radii and temperatures of M-dwar | is the second set of rotation pEriods publIshed For ThiS cLustEr. @MeSsina.08 have recenTLy puBlished periods for 122 clustEr memBeRS. UsiNG tHis daTa we havE InVEStiGaTeD thE ROSsBy numBer-AmplituDe and perioD-coLoR distributioNS.
WE find that fOr sTars with $(B-V)_{0} < 1.36$ thE amPlitudE aNd ROSsby nUmbEr are Anti-coRRelateD, and are reLaTEd via eQUation \[eQN:ArRofiT\]. Extrapolating thiS ReLAtion to higher ROssby nUmBEr VALueS, we Expect $A_{r}$ to DrOp belOW $1~{\rm mmag}$ AT $R_{o} \SIM 1.1$. FoR A Skumanich spiN-down, we expeCT stArs to rEaCh tHIs RossBy numBeR At $\sIm 2~{\rm Gyr}$. We noTe thAt stellar ActiviTY will be A Non-neglIgible EffEct To coNSiDeR whEn COndUCtIng SUrvEys for trAnSiTing pLaneTS WITh amPliTudeS of $\siM 1~{\rm mmag}$ (e.g. NeptUne-SizeD PlaNets oRbitiNg suN-sIzed sTars) thAt orbIt Stars younger thaN $\sim 2~{\Rm Gyr}$. HoweVer, SiNce ThE rotaTIon and TraNsiT time-scAles typICalLy DIFFeR by $\sim 2$ orders of magnItUDE, iT should bE possiBLe To FInd these PlAneTs noNETheleSs. We ALsO find thaT for stARs WiTh $R_{O} < 0.2$, $(B-V)_{0} < 1.36$, tHe LogariThMic DisTribuTIon oF $A_{r}$ appEars to be PeakeD Above $0.03~{\rm mag}$ whicH Suggests that tHE dISTrIButiOn oF spot fillinG-facTOrs iS peaKEd For THese rApidlY rOTaTIng stars.
We have also iNvEstigaTed thE effect of rotaTion on the sHAPE of the maIn seQUeNCe on CMDs. We find That aT fixed $V$ magNItude rapId rotAtors tenD to lie bluEWArd of sloW roTatOrs On a $b-v$ cMd and redward of SLOw roTaTors on a $v-I_{C}$ cMD. This EffEct, SeeN prEvIously for Early K-dwArFs In ThE PlEiadeS And NGC 2516, exTeNds DoWn tO earlY m-dwarf Stars ($m_{V} < 10.5$). We NoTe THat The relaTIoN BEtweEn $v-I_{c}$ and RotAtIon peRiod IS coNsistenT with obseRvaTIons By @moRales.07 whO found that at fIxEd luminosiTy MorE activE m-Dwarfs teNd to have lower effective tEMperatuRes. our obServAtions quaNtiTativeLy aND qualiTativeLy supPoRt tHE HypotHESiS thAt The well-knoWN DisCrepaNcY betWeen the Observed and predictED raDii and effectiVe tEmpeRATuRes OF M-DWar | is the second set of rota tion perio ds pu bli she dforthis cluster. @Mes s ina. 08 have recently publi shedpe r iods fo r 122 cluste r m e m ber s. U sin gt hi s dat a w e haveinvestigat edth e Rossby num b er -amplitude an d period-col ordistri bu tio n s.
W e f ind t hat fo r stars with $(B -V ) _{0} < 1.36$ t h e a mpli tude and Rossby n u mb e r are anti-cor relate d, an d are re lated viaeq uatio n \[eqn: a rr o f i t\] . Extrapolatin g this rela t ion to hi gh erR ossbynumbe rv alu es, we expe ct $ A_{r}$ to dropb elow $1 ~ {\rm mm ag}$ a t $ R_{ O} \ s im 1 .1$ .F ora S kum a nic h spin-d ow n, we e xpec t s t arstoreac h thi s Rossby numb erat $ \ sim 2~{\ rm Gy r}$. W e not e that stel la r activity will bea non-neg lig ib leef fectt o cons ide r w hen con ducting sur ve y s fo r transiting plane ts w it h amplit udes o f $ \s i m 1~{\rm m mag }$ ( e . g. Ne ptun e -s ized pla nets o r bi ti ng sun- si zed st ar s)tha t orb i t st ars yo unger th an $\ s im 2~{\rm Gyr} $ . However, si n ce t he rota tio n and trans it t i me-s cale s t ypi c allydiffe rb y$ \sim 2$ orders of m ag nitude , itshould be pos sible to f i n d these p lane t sn onetheless. We also find that for star s wit h $R_{O} < 0.2$,$ ( B-V)_{0} <1.3 6$, th e lo garithmic dis t r ibut io n of $A _{r }$ appe ars to be pe ak ed above$0.03~{\ rm m ag }$ wh ich s u ggests t ha t t he di strib u tion o f spo t fi ll in g -fa ctors i s p e a kedfo rthes e r ap idlyrota t ing stars.
We have al s o in ve st igatedthe effect of r otation on t heshapeo f the mai n sequence on CMDs. Wef ind tha t a t fix ed $ V$ magnit ude rapid ro t atorstend t o lie b lue w a rd of s lo w r ot ators on a $ B-V $ CMD a nd r edwardof slow rotators o n a$V-I_{C}$ CMD . T hise f fe ct, se e n p re v iou s l y for early K-d warfs in t he Pl eiades and NGC 2 516, ex tends d own t o earlyM-dwarf s tars ($M_ {V } <1 0 .5$ ). We note that th e relatio n betw e en $V-I _{C }$ and r ota tionperiod isconsi stentwi th obs ervat io ns by @M orales.07 who found tha t at f ixedlum inosity m ore act ive M-dwa rfstend to ha velow er ef fec t ive t empe r at ure s . Our obs e rvationsq ua nti t a ti vely and qu a l i tat ively su p port t he h ypothesis that th e well-known di scre p a ncy be t ween t he observed an d p re d i cted rad ii and effect ive temp er a tures of M- dwar | is_the second_set of rotation periods_published for_this_cluster. @Messina.08_have_recently published periods_for 122 cluster_members. Using this data_we have investigated_the_Rossby number-amplitude and period-color distributions.
We find that for stars with $(B-V)_{0} < 1.36$ the_amplitude_and Rossby_number_are_anti-correlated, and are related via_equation \[eqn:arrofit\]. Extrapolating this relation to_higher Rossby_number values, we expect $A_{r}$ to drop below_$1~{\rm_mmag}$ at $R_{O}_\sim 1.1$. For a Skumanich spin-down, we expect stars_to reach this Rossby number at_$\sim 2~{\rm Gyr}$._We_note_that stellar activity will_be a non-negligible effect to consider_when conducting surveys for transiting planets_with amplitudes of $\sim 1~{\rm mmag}$ (e.g._Neptune-sized planets orbiting sun-sized stars) that_orbit stars younger than $\sim_2~{\rm Gyr}$._However, since the rotation and_transit time-scales typically_differ by_$\sim 2$ orders_of magnitude, it should be possible_to find these_planets nonetheless. We also find that_for_stars with $R_{O}_<_0.2$,_$(B-V)_{0} <_1.36$, the logarithmic_distribution_of $A_{r}$_appears_to be peaked above $0.03~{\rm mag}$_which_suggests that the distribution of spot filling-factors_is peaked for these_rapidly_rotating stars.
We have also_investigated the effect of rotation_on the shape of the main_sequence on_CMDs. We_find that at fixed $V$ magnitude rapid rotators tend to lie_blueward of slow rotators on a_$B-V$ CMD and redward_of slow_rotators_on a $V-I_{C}$_CMD._This effect,_seen previously for early K-dwarfs in the_Pleiades and_NGC 2516, extends down to early_M-dwarf stars ($M_{V} <_10.5$)._We note that the relation between_$V-I_{C}$ and rotation period is consistent_with observations by @Morales.07 who_found_that_at fixed luminosity more active_M-dwarfs tend to have lower effective_temperatures. Our observations_quantitatively and qualitatively support the hypothesis that_the_well-known discrepancy between the observed and_predicted_radii and effective temperatures of M-dwar |
well as enjoy a conserved energy, that is, \[funcional\] f&&\~\_[a=0]{} 0(a\^[-u]{})\
f&&\~\_[a=|[a]{}]{}(|[a]{}-a)\^v with $0<u<1$ and $v>\frac{1}{2}$.
If we want moreover the now well defined energy to be finite it is clearly necessary (owing to the $a^{\prime}$ factor in the denominator) that when $r\rightarrow \infty$ \[mu\] T\~e\^[- \^2 r]{} with \[finita\] \^2 > which is much faster that required by (\[funcional\]).We can define another functional space in which energy is not only conserved, but also finite, and this space is essentially characterized by the behavior (\[finita\]).
It is then plain that in the conserved-energy functional space defined by (\[funcional\]) there are fluctuations with arbitarily large negative energy, namely all those that at infinity behave as (\[mu\]) with $|\mu|< |m|$. In the more restrictive finite-energy space (\[finita\]), this behavior still persists as long as $d>1$. This is the same bound found for the linear dilaton background.
The tachyon equation
--------------------
Let us now examine the space of solutions of the linearized equation of motion and check whether on shell there are elements of our functional space of well-defined fluctuations.
We shall write the wave equations (following the conventions of [@polchinski]) as: \[wave\] \^2 T - 2 \^A \_A T - m\^2 T = 0 For the Poincaré invariant ansatz the dilaton is given by =log a’ + log a where $a'\equiv\frac{da}{dr}$.
Variables can be separated in the wave equation (\[wave\]) by means of: T(x\^A) T(a)(x\^) with the result: \[eqa\] \_d && =\
T” + T’&&= ()T
(where we have used $\frac{dT}{dr} = \frac{dT}{da}a'\equiv T' a'$), and $\lambda$ is the separation constant.
Now, for a plane wave $\Theta = e^{i kx}\equiv
e^{-i(\omega t - \vec{k}\vec{x})}$, the first (Minkowskian | well as enjoy a conserved energy, that is, \[funcional\ ] f&&\~\_[a=0 ] { } 0(a\^[-u]{})\
f&&\~\_[a=|[a]{}]{}(|[a]{}-a)\^v with $ 0 < u<1 $ and $ v>\frac{1}{2}$.
If we want furthermore the nowadays well defined energy to be finite it is intelligibly necessary (owe to the $ a^{\prime}$ factor in the denominator) that when $ r\rightarrow \infty$ \[mu\ ] T\~e\^[- \^2 r ] { } with \[finita\ ] \^2 & gt; which is a lot faster that required by (\[funcional\]).We can define another running quad in which energy is not entirely conserved, but also finite, and this distance is basically characterized by the behavior (\[finita\ ]).
It is then apparent that in the conserved - energy running space defined by (\[funcional\ ]) there are variation with arbitarily large negative department of energy, namely all those that at infinity act as (\[mu\ ]) with $ |\mu| < |m|$. In the more restrictive finite - energy space (\[finita\ ]), this behavior still persists as long as $ d>1$. This is the like bound found for the linear dilaton background.
The tachyon equation
--------------------
Let us nowadays examine the space of solutions of the linearized equation of motion and check whether on shell there are elements of our running outer space of well - defined fluctuation.
We shall write the wave equations (following the conventions of [ @polchinski ]) as: \[wave\ ] \^2 T - 2 \^A \_A T - m\^2 deoxythymidine monophosphate = 0 For the Poincaré changeless ansatz the dilaton is given by = log a ’ + log a where $ a'\equiv\frac{da}{dr}$.
Variables can be separated in the wave equation (\[wave\ ]) by mean of: T(x\^A) T(a)(x\^) with the result: \[eqa\ ] \_d & & = \
T ” + T’&&= () T
(where we have used $ \frac{dT}{dr } = \frac{dT}{da}a'\equiv T' a'$), and $ \lambda$ is the separation constant.
Now, for a plane wave $ \Theta = e^{i kx}\equiv
e^{-i(\omega metric ton - \vec{k}\vec{x})}$, the first (Minkowskian | wepl as enjoy a conserved tnergy, that is, \[fuuxional\] f&&\~\_[a=0]{} 0(a\^[-u]{})\
r&&\~\_[a=|[a]{}]{}(|[a]{}-a)\^v wkth $0<u<1$ and $v>\frac{1}{2}$.
If we want moceovwr tht now well defined dnergy to be finire iu is clearly necessary (owlug to bhe $a^{\'rmme}$ factor in tme denominador) that when $s\rkgktarrow \infty$ \[mu\] T\~e\^[- \^2 r]{} with \[finita\] \^2 &dt; whicn ls much faster thau rqquidvd by (\[funcional\]).We can define ankther flnctional space im which energy is not only conderved, but also fijite, and thus s[qce is essengially chagccterized bi the behavior (\[finita\]).
It is then olain that in tye cojverved-energb funcnional space defined ty (\[funcoonal\]) there arc flurtuarions with arbitarily large negative engrgy, namelf cll those that at infunuty bghave as (\[oy\]) wkth $|\mn|< |m|$. In thf mkre restridtive finitw-energy space (\[finits\]), eyis behavior atill [ewsists as long as $d>1$. This is the same bolnd round for the linear dioaton background.
The twchyon eqtation
--------------------
Let us now examine the space of solutions ox the uintavlzed wqkation of motion and check whether on shell trsrt age elements of ouv functional space ov egll-defined fluztuations.
Se shall write the wave ezuatiins (folloring the conventions of [@polchinwki]) as: \[wave\] \^2 R - 2 \^A \_A T - m\^2 T = 0 Flr the Poinearé infariamt ansatz the dilaton ir giben by =log w’ + log a sfere $a'\equiv\frac{dx}{dr}$.
Naridbles can be separated in ehe wave xquatnon (\[wave\]) by keans jf: T(x\^A) T(a)(x\^) with the result: \[eqa\] \_d && =\
H” + T’&&= ()J
(where we have uded $\frac{dT}{dr} = \frac{dT}{da}a'\equiv T' e'$), and $\lambda$ os thv separatnon cokstant.
Now, for a plane wave $\Thgta = e^{i kr}\equiv
e^{-i(\omega t - \vec{k}\vxc{x})}$, the firse (Minkowskian | well as enjoy a conserved energy, that f&&\~\_[a=0]{} f&&\~\_[a=|[a]{}]{}(|[a]{}-a)\^v with and $v>\frac{1}{2}$. If well energy to be it is clearly (owing to the $a^{\prime}$ factor in denominator) that when $r\rightarrow \infty$ \[mu\] T\~e\^[- \^2 r]{} with \[finita\] \^2 > is much faster that required by (\[funcional\]).We can define another functional space in energy not conserved, also finite, and this space is essentially characterized by the behavior (\[finita\]). It is then plain in the conserved-energy functional space defined by (\[funcional\]) are fluctuations with arbitarily negative energy, namely all those at behave as with |m|$. the more restrictive space (\[finita\]), this behavior still persists as long as $d>1$. This is the same bound found for linear dilaton tachyon equation Let now the space of the linearized equation of motion and shell there are elements of our functional space well-defined fluctuations. shall write the wave equations (following conventions of [@polchinski]) as: \[wave\] \^2 T - \^A \_A T - m\^2 T = 0 For the Poincaré invariant ansatz the dilaton by =log a’ + a where $a'\equiv\frac{da}{dr}$. can separated the equation (\[wave\]) means of: T(x\^A) T(a)(x\^) with the result: \[eqa\] \_d && =\ + T’&&= ()T (where we have used $\frac{dT}{dr} = \frac{dT}{da}a'\equiv and is the separation Now, for a plane $\Theta e^{i kx}\equiv e^{-i(\omega t the (Minkowskian | well as enjoy a conserved enerGy, that is, \[fuNcionAl\] f&&\~\_[A=0]{} 0(a\^[-u]{})\
F&&\~\_[a=|[A]{}]{}(|[a]{}-a)\^v With $0<U<1$ and $v>\frac{1}{2}$.
If we wANt moReover the now well defineD enerGy TO be fINiTe it iS clearlY NeCESsaRy (OwIng To THe $A^{\primE}$ faCtor in tHe denominaTor) ThAt when $r\rightARrOw \infty$ \[mu\] T\~E\^[- \^2 r]{} wIth \[finita\] \^2 > wHicH is mucH fAstER that ReqUired By (\[funcIOnal\]).We Can define AnOTher fuNCtional SPAcE in wHich energy is not onLY cONserved, but also Finite, AnD ThIS SpaCe iS essentialLy CharaCTerized BY tHE BEhaVIor (\[finita\]).
It is Then plain thAT in The conSeRveD-Energy FunctIoNAl sPace defined By (\[fuNcional\]) thEre are FLuctuatIOns with ArbitaRilY laRge nEGaTiVe eNeRGy, nAMeLy aLL thOse that aT iNfInity BehaVE AS (\[Mu\]) wiTh $|\mU|< |m|$. In The moRe restrictive FinIte-eNErgY spacE (\[finiTa\]), thIs BehavIor stiLl perSiSts as long as $d>1$. ThiS is tHe same bouNd fOuNd fOr The liNEar dilAtoN baCkgrounD.
The tacHYon EqUATIoN
--------------------
Let us now examine thE sPACe Of solutiOns of tHE lInEArized eqUaTioN of mOTIon anD cheCK wHether on Shell tHErE aRe elemeNtS of our FuNctIonAl spaCE of wEll-defIned flucTuatiONs.
We shall write THe wave equatioNS (fOLLoWIng tHe cOnventions oF [@polCHinsKi]) as: \[WAvE\] \^2 T - 2 \^A \_a t - m\^2 T = 0 FoR the POiNCaRÉ invariant ansatz the DiLaton iS giveN by =log a’ + log a whEre $a'\equiv\fRAC{Da}{dr}$.
VariAbleS CaN Be separated in tHe wavE equation (\[wAVe\]) by meanS of: T(x\^a) T(a)(x\^) with The result: \[EQA\] \_d && =\
T” + T’&&= ()T
(wheRe wE haVe uSed $\FRAc{DT}{dr} = \frac{dT}{da}a'\EQUiv T' A'$), aNd $\lambdA$ is The sepaRatIon ConStaNt.
now, for a plAne wave $\THeTa = E^{i Kx}\EquIv
e^{-i(\oMEga t - \vec{k}\VeC{x})}$, tHe FirSt (MinKOwskiaN | well as enjoy a conserved energy, t hat i s,\[f un cion al\] f&&\~\_[a=0]{ } 0(a \^[-u]{})\
f&&\~\_[a=| [a]{} ]{ } (|[a ] {} -a)\^ v with$ 0< u < 1$an d$v> \f r ac {1}{2 }$.
If we want more ove rthe now well de fined ener gyto be finite it is cl ea rly neces sar y (ow ing to the $a ^{\prime} $f actori n the d e n om inat or) that when $r\ r ig h tarrow \infty$ \[mu\ ]T \~ e \ ^[- \^ 2 r]{} wit h\[fin i ta\] \^ 2 & g t ; wh i ch is much fa ster that r e qui red by ( \[f u nciona l\]). We can define ano ther function al spa c e in wh i ch ener gy isnot on ly c o ns er ved ,b uta ls o f i nit e, and t hi sspace ise s s e ntia lly cha racte rized by thebeh avio r (\ [fini ta\]) .
I tis th en pla in th at in the conserv ed-e nergy fun cti on alsp ace d e finedby(\[ funcion al\]) t h ere a r e fl uctuations with ar bi t a ri ly large negat i ve e n ergy, na me lyallt h ose t hata tinfinity behav e a s(\[mu\] )with $ |\ mu| < | m|$.I n th e more restric tivef inite-energy s p ace (\[finita \ ]) , th i s be hav ior still p ersi s ts a s lo n gas$ d>1$. This i s t h e same bound foundfo r thelinea r dilaton bac kground.
T h e tachyon equ a ti o n
------------ ----- ---
Let u s now exa minethe spac e of solu t i ons of t helin ear ize d eq uation of mot i o n an dcheck w het her onshe llthe rear e element s of our f un ct io nal spac e of well -d efi ne d f luctu a tions.
Weshal lwr i tethe wav e e q u atio ns ( foll owi ng theconv e nti ons of[@polchin ski ] ) as :\[ wave\]\^2 T - 2 \^A \ _A T - m\^ 2T = 0 For t he Poinc aré invariant ansatz th e dilato n i s giv en b y =log a’ +log awhe r e $a'\ equiv\ frac{ da }{d r } $.
V a r ia ble scan be sep a r ate d inth e wa ve equa tion (\[wave\]) by mea ns of: T(x\^A ) T (a)( x \ ^) wi t ht here s ult : \[eqa\] \_d &&=\
T” + T’ && = ( )T
(where weha ve used $\frac {dT}{ d r} = \f rac{dT}{d a}a'\equi vT' a ' $ ),and $\lamb da$ is t he separa t ion c o ns tant.
N ow, fo ra p lanewave $ \ The ta =e^{i k x} \equiv
e^{ -i (\omegat - \vec{k}\vec{x})}$,the fi rst ( Min kowskian | well_as enjoy_a conserved energy, that_is, \[funcional\]_f&&\~\_[a=0]{}_0(a\^[-u]{})\
f&&\~\_[a=|[a]{}]{}(|[a]{}-a)\^v with_$0<u<1$_and $v>\frac{1}{2}$.
If we_want moreover the_now well defined energy_to be finite_it_is clearly necessary (owing to the $a^{\prime}$ factor in the denominator) that when $r\rightarrow_\infty$_\[mu\] T\~e\^[-_\^2_r]{}_with \[finita\] \^2 > which_is much faster that required_by (\[funcional\]).We_can define another functional space in which energy_is_not only conserved,_but also finite, and this space is essentially characterized_by the behavior (\[finita\]).
It is then_plain that in_the_conserved-energy_functional space defined by_(\[funcional\]) there are fluctuations with arbitarily_large negative energy, namely all those_that at infinity behave as (\[mu\]) with_$|\mu|< |m|$. In the more restrictive_finite-energy space (\[finita\]), this behavior_still persists_as long as $d>1$. This_is the same_bound found_for the linear_dilaton background.
The tachyon equation
--------------------
Let us now_examine the space_of solutions of the linearized equation_of_motion and check_whether_on_shell there_are elements of_our_functional space_of_well-defined fluctuations.
We shall write the wave_equations_(following the conventions of [@polchinski]) as: \[wave\]_\^2 T - 2_\^A_\_A T - m\^2_T = 0 For the_Poincaré invariant ansatz the dilaton is_given by_=log a’_+ log a where $a'\equiv\frac{da}{dr}$.
Variables can be separated in the wave_equation (\[wave\]) by means of: T(x\^A)_T(a)(x\^) with the result:_\[eqa\] \_d_&&_=\
T” + T’&&=_()T
(where_we have_used $\frac{dT}{dr} = \frac{dT}{da}a'\equiv T' a'$), and_$\lambda$ is_the separation constant.
Now, for a plane_wave $\Theta = e^{i_kx}\equiv_
e^{-i(\omega t - \vec{k}\vec{x})}$, the first_(Minkowskian |
{circ}})^{2/3},b,\tau_i\}$, we also allow each transit epoch to have a unique out-of-transit normalization factor, $OOT_m$ where $m$ denotes the epoch number. The zeroth epoch is defined to be that which has the lowest mutual correlation to the orbital period. Finally, the orbital period is a free parameter too.
Direct -MAP Priors {#sub:directpriors}
------------------
In most cases (system with less than five transiting planets), there are more free parameters in the model than observables and so the problem is under-constrained with no unique solution. Despite this, contours of the error surface may still be computed through Monte Carlo techniques. In our case, the Monte Carlo technique of choice is Markov Chain Monte Carlo (MCMC) with the Metropolis-Hastings algorithm. We note that an analogous situation arises for interpreting the atmospheres of exoplanets where atmosphere models tend to have more free parameters than the number of observations. In this example, a similar solution as n-MAP has been adopted in works such as @madhu:2009 and @madhu:2011. A direct comparison of n-MAP to the methods of @madhu:2011 is discussed in §\[sub:madhucomparison\].
We stress here that because the problem is under-constrained (except for $n\geq5$), the results will be strongly affected by the choice of priors. This is in contrast to a highly constrained problem where the data drive the result to the same solution with only a minor dependency upon the choice of priors. Thus, the choice of priors can be understood to have a significant impact on the derived results (see @ford:2005 for a detailed discussion on the effect of priors when fitting exoplanet data). The results should consequently be always quoted in unison with the adopted prior used to infer them.
These ideas are more formally expressed through Bayes’ theorem. Let us denote the eccentricity parameters which we fit for in n-MAP as $\vec{\Lambda} = \{e_k,\omega_k,...\}$ for $k=1,n$. We further use $\mathcal{M}$ to represent the model and $\mathcal{D}$ to represent the data:
$$\begin{aligned}
\mathrm{P}(\vec{\Lambda}|\mathcal{D},\mathcal{M}) &= \frac{ \mathrm{P}(\vec{\Lambda}|\mathcal{M}) | { circ}})^{2/3},b,\tau_i\}$, we also allow each transit epoch to have a alone out - of - passage normalization factor, $ OOT_m$ where $ m$ denotes the epoch number. The zeroth epoch is define to be that which has the lowest reciprocal correlation coefficient to the orbital period. Finally, the orbital period is a free parameter besides.
Direct -MAP Priors { # sub: directpriors }
------------------
In most cases (organization with less than five transiting planet), there are more free parameters in the model than observables and so the trouble is under - constrained with no unique solution. Despite this, contours of the error airfoil may still be computed through Monte Carlo technique. In our case, the Monte Carlo proficiency of choice is Markov Chain Monte Carlo (MCMC) with the Metropolis - Hastings algorithm. We notice that an analogous situation arises for interpreting the atmospheres of exoplanets where air models tend to have more free parameters than the number of observation. In this exercise, a similar solution as n - MAP has been adopted in works such as @madhu:2009 and @madhu:2011. A direct comparison of n - MAP to the methods of @madhu:2011 is discussed in § \[sub: madhucomparison\ ].
We stress here that because the problem is nether - constrained (except for $ n\geq5 $), the consequence will be strongly involve by the option of priors. This is in contrast to a highly constrained problem where the data drive the result to the same solution with merely a minor dependency upon the choice of priors. Thus, the choice of priors can be understand to have a significant impact on the derived results (see @ford:2005 for a detailed discussion on the consequence of priors when fitting exoplanet data). The results should consequently be always quoted in unison with the espouse prior used to infer them.
These ideas are more formally carry through Bayes ’ theorem. Let us denote the eccentricity parameters which we fit for in n - MAP as $ \vec{\Lambda } = \{e_k,\omega_k,... \}$ for $ k=1,n$. We far use $ \mathcal{M}$ to stage the model and $ \mathcal{D}$ to represent the datum:
$ $ \begin{aligned }
\mathrm{P}(\vec{\Lambda}|\mathcal{D},\mathcal{M }) & = \frac { \mathrm{P}(\vec{\Lambda}|\mathcal{M } ) | {cirf}})^{2/3},b,\tau_i\}$, we also allow eacm transit epoch jo have e uniqus out-of-tfansit normalization factor, $LOR_m$ whtge $m$ denotes the epocf number. Nhe zeroty epixh is defiisd to bc thaf whieh has the lowesj mutual corselation to tha uryital period. Finally, the orbital peryod is s vree parameter too.
Cyrecf -MAP Priors {#sub:directpriors}
------------------
In mosf cases (system with kess than five transiting olanfts), there are more free paramgferf in the modeu than obstrrables and ao the problem is under-constraived wnth no uniqoz sopotion. Despitx this, contours of the errmr surfsce may still ne cokpured through Monte Carno techniques. In oor case, tha Jonte Carlo technuqye of chohce kw Mxrkkv Cgain Mlntx Carlo (MCMD) with the Netropolis-Hastings sldirithm. We nots that ag analogous situation arises for interpgetihg the atmospheres of ezoplanets where atmoskhere modejs tend to have more free parameters than the numter oh ubstrywguojs. In this example, a similar solution as n-MAP gax neen adopted in corks such as @machk:2009 sgd @madhu:2011. A ditect compadison of n-MAP to tje methjds od @madhu:2011 if dixcussed in §\[sub:madhucomparisin\].
We stress kerw that because the problem is undgr-consyrained (except for $n\geq5$), the results wipl be strkvgly affected by thv chmice of kfiors. This is in sontrast vo a kighly cunsttained [roblem whfre tma data drive the rfsult tm the same solution with only a minor depxidency upon tne chmice of 'riors. Thus, the choyce of priors ean be uuderstuod to havv a signihicant impace on the deriegd results (sex @ford:2005 fow a eetauled dirzussion on the effect oy priirs when fitting ewoplavst data). The resbuus should conseqiengly bv akwwfs quoted in uniron eith ghe adokced oriot used to infer them.
Dhess ideas are more fprially exkressed trrough Bayes’ yheorem. Let us denlte tie eccxntricoty parameters which we fit for ih n-MAP as $\veg{\Lambda} = \{e_k,\omqga_k,...\}$ for $k=1,n$. We fbrther use $\mathcal{M}$ to represent the modxl and $\mathcal{D}$ to reptesent the data:
$$\begin{aklgned}
\mathrm{P}(\tec{\Lamfda}|\mathcan{D},\mathcal{M}) &= \frac{ \matyrm{P}(\vec{\Lambda}|\mathgal{M}) | {circ}})^{2/3},b,\tau_i\}$, we also allow each transit epoch a out-of-transit normalization $OOT_m$ where $m$ zeroth is defined to that which has lowest mutual correlation to the orbital Finally, the orbital period is a free parameter too. Direct -MAP Priors {#sub:directpriors} In most cases (system with less than five transiting planets), there are more parameters the than and so the problem is under-constrained with no unique solution. Despite this, contours of the error may still be computed through Monte Carlo techniques. our case, the Monte technique of choice is Markov Monte (MCMC) with Metropolis-Hastings We that an analogous arises for interpreting the atmospheres of exoplanets where atmosphere models tend to have more free parameters than number of this example, similar as has been adopted such as @madhu:2009 and @madhu:2011. A n-MAP to the methods of @madhu:2011 is discussed §\[sub:madhucomparison\]. We here that because the problem is (except for $n\geq5$), the results will be strongly by the choice of priors. This is in contrast to a highly constrained problem where drive the result to same solution with a dependency the of priors. the choice of priors can be understood to have a significant on the derived results (see @ford:2005 for a detailed discussion effect priors when fitting data). The results should be quoted in unison with prior to ideas more expressed through Bayes’ theorem. us denote the eccentricity parameters we fit for in for $k=1,n$. We further use $\mathcal{M}$ to represent model and $\mathcal{D}$ to represent the data: \mathrm{P}(\vec{\Lambda}|\mathcal{D},\mathcal{M}) &= \frac{ \mathrm{P}(\vec{\Lambda}|\mathcal{M}) | {circ}})^{2/3},b,\tau_i\}$, we also allow each tRansit epocH to haVe a UniQuE out-Of-trAnsit normalizaTIon fActor, $OOT_m$ where $m$ denotes The epOcH NumbER. THe zerOth epocH Is DEFinEd To Be tHaT WhIch haS thE lowest Mutual corrElaTiOn to the orbitAL pEriod. FinalLy, tHe orbital perIod Is a freE pAraMEter tOo.
DIrect -mAP PriORs {#sub:dIrectprioRs}
------------------
iN most cASes (systEM WiTh leSs than five transitINg PLanets), there are More frEe PArAMEteRs iN the model tHaN obseRVables aND sO THE prOBlem is under-coNstrained wiTH no Unique SoLutIOn. DespIte thIs, COntOurs of the erRor sUrface may Still bE ComputeD Through monte CArlO teChniQUeS. IN ouR cASe, tHE MOntE carLo techniQuE oF choiCe is mARKOv ChAin montE CarlO (MCMC) with the MEtrOpolIS-HaStingS algoRithM. WE note That an AnaloGoUs situation arisEs foR interpreTinG tHe aTmOspheREs of exOplAneTs where AtmosphERe mOdELS TeNd to have more free paRaMETeRs than thE numbeR Of ObSErvationS. IN thIs exAMPle, a sImilAR sOlution aS n-MAP hAS bEeN adopteD iN works SuCh aS @maDhu:2009 anD @MadhU:2011. A direCt comparIson oF N-MAP to the methoDS of @madhu:2011 is disCUsSED iN §\[Sub:mAdhUcomparison\].
we stREss hEre tHAt BecAUse thE probLeM Is UNder-constrained (excePt For $n\geQ5$), the rEsults will be sTrongly affECTEd by the cHoicE Of PRiors. This is in cOntraSt to a highlY ConstraiNed prOblem wheRe the data DRIve the reSulT to The SamE SOlUtion with only A MInor DePendencY upOn the chOicE of PriOrs. thUs, the choiCe of prioRs CaN bE uNdeRstooD To have a sIgNifIcAnt ImpacT On the dEriveD resUlTs (SEe @fOrd:2005 for a DEtAILed dIsCuSsioN on ThE effeCt of PRioRs when fItting exoPlaNEt daTa). thE resultS should conseqUeNtly be alwaYs QuoTed in uNISon with tHe adopted prior used to infER them.
ThEse Ideas Are mOre formalLy eXpressEd tHRough BAyes’ thEorem. leT us DENote tHE EcCenTrIcity paramETErs Which We Fit fOr in n-MAp as $\vec{\Lambda} = \{e_k,\omegA_K,...\}$ foR $k=1,n$. We further uSe $\mAthcAL{m}$ tO rePReSEnt ThE ModEL And $\mathcal{D}$ to rePresent the DaTA:
$$\bEgin{aligneD}
\MatHrM{P}(\vec{\LaMbda}|\matHcal{D},\MAthcal{M}) &= \Frac{ \mathrM{P}(\vec{\LambDa}|\MathCAL{M}) | {circ}})^{2/3},b,\tau_i\}$ , we alsoallow ea chtr ansi t ep och to have au niqu e out-of-transit norma lizat io n fac t or , $OO T_m$ wh e re $ m$de no tes t h eepoch nu mber. T he zerothepo ch is definedt obe that wh ich has the low est mutua lcor r elati onto th e orbi t al per iod. Fina ll y , theo rbitalp e ri od i s a free paramete r t o o.
Direct -MA P Prio rs {# s u b:d ire ctpriors}-- ----- - ------- - -- I n m o st cases (sys tem with le s s t han fi ve tr a nsitin g pla ne t s), there aremore free par ameter s in the model t han ob ser vab lesa nd s o t he pro b le m i s un der-cons tr ai ned w ithn o u niqu e s olut ion.Despite this, co ntou r s o f the erro r su rf ace m ay sti ll be c omputed through Mon te Carlotec hn iqu es . Ino ur cas e,the MonteCarlo t e chn iq u e of choice is MarkovCh a i nMonte Ca rlo (M C MC )w ith theMe tro poli s - Hasti ngsa lg orithm.We not e t ha t an an al ogoussi tua tio n ari s es f or int erpretin g the atmospheres of exoplanets wh e re a tm o sphe remodels tend toh avemore fr eep arame tersth a nt he number of observ at ions.In th is example, a similar s o l u tion asn-MA P h a s been adopted in w orks sucha s @madhu :2009 and @ma dhu:2011. A directcom par iso n o f n- MAP to the me t h odsof @madhu :20 11 is d isc uss edin§\ [sub:madh ucompari so n\ ].
Westres s here th at be ca use thep roblem is u nder -c on s tra ined (e x ce p t for $ n\ geq5 $), t he re sult s wi ll be s trongly a ffe c tedby t he choi ce of priors. T his is inco ntr ast to a highlyconstrained problem whe r e the d ata driv e th e resulttothe sa mes olutio n with only a mi n o r dep e n de ncy u pon the ch o i ceof pr io rs.Thus, t he choice of prior s ca n be understo odto h a v ea s i gn i fic an t im p a ct on the deriv ed results ( s ee @ford:200 5 fo ra detai led dis cussi o n on th e effectof priors w henf i tti ng exoplan et data) . The res u lts s h ou ld co nse quentl ybealway s quot e d i n uni son wi th the a dopte dprior us ed to infer them.
Thes e idea s are mo re formal lye xpr essed thr ough Bayes’ th eor em. Letusd enote the ec cen t ricit y pa r ameters w h ic h w e fi t for in n- M A P as $\ve c{\ L ambda} = \ {e_k,\omega_k,... \ }$ for $k=1,n$ . We f urt her use$\ mathcal{M}$ to re pr e s ent themo del and $\m athcal{D }$ to re presen t thedata:
$ $ \b e gin{al igne d}\mathrm{P }(\ ve c {\Lambd a} |\ m athcal {D}, \m athcal {M}) & = \fr a c { \mathrm{P}(\ve c{\La m b da}|\ m ath cal{M }) | {circ}})^{2/3},b,\tau_i\}$, we_also allow_each transit epoch to_have a_unique_out-of-transit normalization_factor,_$OOT_m$ where $m$_denotes the epoch_number. The zeroth epoch_is defined to_be_that which has the lowest mutual correlation to the orbital period. Finally, the orbital_period_is a_free_parameter_too.
Direct -MAP Priors {#sub:directpriors}
------------------
In most_cases (system with less than_five transiting_planets), there are more free parameters in the_model_than observables and_so the problem is under-constrained with no unique solution._Despite this, contours of the error_surface may still_be_computed_through Monte Carlo techniques._In our case, the Monte Carlo_technique of choice is Markov Chain_Monte Carlo (MCMC) with the Metropolis-Hastings algorithm._We note that an analogous situation_arises for interpreting the atmospheres_of exoplanets_where atmosphere models tend to_have more free_parameters than_the number of_observations. In this example, a similar_solution as n-MAP_has been adopted in works such_as_@madhu:2009 and @madhu:2011._A_direct_comparison of_n-MAP to the_methods_of @madhu:2011_is_discussed in §\[sub:madhucomparison\].
We stress here that_because_the problem is under-constrained (except for $n\geq5$),_the results will be_strongly_affected by the choice_of priors. This is in_contrast to a highly constrained problem_where the_data drive_the result to the same solution with only a minor dependency_upon the choice of priors. Thus,_the choice of priors_can be_understood_to have a_significant_impact on_the derived results (see @ford:2005 for a_detailed discussion_on the effect of priors when_fitting exoplanet data). The_results_should consequently be always quoted in_unison with the adopted prior used_to infer them.
These ideas are_more_formally_expressed through Bayes’ theorem. Let_us denote the eccentricity parameters which_we fit for_in n-MAP as $\vec{\Lambda} = \{e_k,\omega_k,...\}$ for_$k=1,n$._We further use $\mathcal{M}$ to represent_the_model and $\mathcal{D}$ to represent the_data:
$$\begin{aligned}
\mathrm{P}(\vec{\Lambda}|\mathcal{D},\mathcal{M})_&=_\frac{ \mathrm{P}(\vec{\Lambda}|\mathcal{M}) |
In particular, the $D=5$ potential admits a second, non-supersymmetric $AdS_5$ critical locus (first found by Romans [@Romans:1984an] in $D=5$ $N=8$ supergravity), at \[RomansCP\] e\^[4U]{}=e\^[-4V]{}=, =e\^e\^[i\^]{}, -a=ie\^[-]{}, where $\theta^\prime$ is an arbitrary phase. All other scalars of the $D=5$ theory are set to trivial, and the $AdS_5$ radius at these points is $2{\sqrt 2}/3$.
Feeding (\[RomansCP\]) into (\[upliftmet\]), (\[upliftflux\]), we find two new classes of $D=11$ solutions, consisting on non-supersymmetric, warped and direct products, respectively, of $AdS_5$ with a smooth, complex manifold $M_6$, equipped, in both cases, with a Hermitian, two-parameter metric. Both classes are parametrised by $(\tilde{a}, \phi)$. To see this, note that out of the five parameters $(\tilde{a},\tilde{c}, a, \phi, \theta^\prime)$ of the solution (\[RomansCP\]), (\[upliftmet\]), (\[upliftflux\]), $\theta^\prime$ and $a$ can be removed by coordinate transformations. The two classes are then distinguished by the value of $\tilde c$. If $\tilde{c} \neq 0$, it is easy to check that it can be taken to $\tilde{c}=1$ with no loss of generality, leading to the two-parameter warped product class. If $\tilde{c}=0$, the warp factor $\Delta$ reduces to a constant, leading to the direct product class. The non-supersymmetry of the $D=5$ critical point (\[RomansCP\]) carries over to its $D=11$ uplift. This lack of supersymmetry can also be directly seen in $D=11$, as the uplifted solutions do not fit into the generic form of $AdS_5$ supersymmetric solutions in $D=11$ [@ads5m6], reviewed in subsection \[geometries\].
The topology of the internal $M_6$ for both classes is the same as that of the supersymmetric $S^2 \times T^2$ solutions | In particular, the $ D=5 $ potential admits a second, non - supersymmetric $ AdS_5 $ critical locus (foremost discover by Romans [ @Romans:1984an ] in $ D=5 $ $ N=8 $ supergravity), at \[RomansCP\ ] e\^[4U]{}=e\^[-4V]{}=, = e\^e\^[i\^ ] { }, -a = ie\^[- ] { }, where $ \theta^\prime$ is an arbitrary phase. All other scalars of the $ D=5 $ hypothesis are set to superficial, and the $ AdS_5 $ radius at these points is $ 2{\sqrt 2}/3$.
Feeding (\[RomansCP\ ]) into (\[upliftmet\ ]), (\[upliftflux\ ]), we discover two new classes of $ D=11 $ solution, consisting on non - supersymmetric, warped and direct product, respectively, of $ AdS_5 $ with a smooth, complex manifold $ M_6 $, equipped, in both case, with a Hermitian, two - parameter metric. Both classes are parametrised by $ (\tilde{a }, \phi)$. To see this, note that out of the five parameters $ (\tilde{a},\tilde{c }, a, \phi, \theta^\prime)$ of the solution (\[RomansCP\ ]), (\[upliftmet\ ]), (\[upliftflux\ ]), $ \theta^\prime$ and $ a$ can be remove by coordinate transformations. The two classes are then distinguished by the value of $ \tilde c$. If $ \tilde{c } \neq 0 $, it is comfortable to check that it can be taken to $ \tilde{c}=1 $ with no personnel casualty of generality, lead to the two - parameter warped product class. If $ \tilde{c}=0 $, the warp factor $ \Delta$ reduce to a constant, leading to the direct product class. The non - supersymmetry of the $ D=5 $ critical distributor point (\[RomansCP\ ]) carries over to its $ D=11 $ uplift. This lack of supersymmetry can also be directly seen in $ D=11 $, as the uplifted solutions do not fit into the generic form of $ AdS_5 $ supersymmetric solutions in $ D=11 $ [ @ads5m6 ], reviewed in subsection \[geometries\ ].
The regional anatomy of the home $ M_6 $ for both classes is the like as that of the supersymmetric $ S^2 \times T^2 $ solution | In particular, the $D=5$ potentlal admits a second, non-vupersgmmetric $AdS_5$ critical locus (first fonnd vy Ronans [@Romans:1984an] in $D=5$ $N=8$ sjpergraviny), at \[RomqnsCK\] e\^[4U]{}=e\^[-4V]{}=, =e\^e\^[i\^]{}, -a=ie\^[-]{}, whxde $\theta^\prime$ ls an erbitrary phase. All other scalars of tha $A=5$ cheory are set to trivial, and the $AdF_5$ radiux wt these pointf is $2{\fqrt 2}/3$.
Feeding (\[RomansCP\]) into (\[upliftmet\]), (\[ulliftfllx\]), we find two nee classes of $D=11$ solutions, clnsidting on non-supersjmmetric, watled qnd direct pfoducts, respectively, or $AdS_5$ with a smooth, complex mankfold $M_6$, equippee, un hmth cases, wmth a Rermitian, two-parametar metroc. Both classex ace pqrametrised by $(\tilde{a}, \phi)$. To see this, njte that muc of the five parametwrw $(\tilge{a},\thlde{z}, a, \ohi, \tietz^\prime)$ of the solutjon (\[RomansCP\]), (\[upliftmet\]), (\[upliftfkuv\]), $\theta^\prime$ ahd $a$ cwn be removed by coordinate transformatiots. Fhe two classes are theb distinguished by thg value of $\tilde c$. If $\tilde{c} \neq 0$, it is easy to check that ht cai ce uajen tu $\tllde{c}=1$ with no loss of generality, leading to trs uwo-iarameter warped iroduct class. If $\tolfe{v}=0$, the warp facjor $\Delcz$ deduces to a constwnt, leaqing ro the diwect product class. The non-superwymmetry of nhe $E=5$ critical point (\[RlmansCP\]) carxies ofer tp its $D=11$ uplift. This lacy of supersymmehry can amro be directly sden it $D=11$, as the uplifted solutijns do nov fit into tfe ggneric sorm of $AdD_5$ supcssymmetric solutiojs in $D=11$ [@ads5m6], revifwed in subsection \[geometries\].
Thx topology of tve pnternal $I_6$ for both classes is the same ax that jf thd supersymjetric $V^2 \times T^2$ sjlutions | In particular, the $D=5$ potential admits a $AdS_5$ locus (first by Romans [@Romans:1984an] \[RomansCP\] =e\^e\^[i\^]{}, -a=ie\^[-]{}, where is an arbitrary All other scalars of the $D=5$ are set to trivial, and the $AdS_5$ radius at these points is $2{\sqrt Feeding (\[RomansCP\]) into (\[upliftmet\]), (\[upliftflux\]), we find two new classes of $D=11$ solutions, on warped direct respectively, of $AdS_5$ with a smooth, complex manifold $M_6$, equipped, in both cases, with a Hermitian, metric. Both classes are parametrised by $(\tilde{a}, \phi)$. see this, note that of the five parameters $(\tilde{a},\tilde{c}, \phi, of the (\[RomansCP\]), (\[upliftflux\]), and $a$ can removed by coordinate transformations. The two classes are then distinguished by the value of $\tilde c$. If \neq 0$, easy to that can taken to $\tilde{c}=1$ loss of generality, leading to the class. If $\tilde{c}=0$, the warp factor $\Delta$ reduces a constant, to the direct product class. The of the $D=5$ critical point (\[RomansCP\]) carries over its $D=11$ uplift. This lack of supersymmetry can also be directly seen in $D=11$, as solutions do not fit the generic form $AdS_5$ solutions $D=11$ reviewed in \[geometries\]. The topology of the internal $M_6$ for both classes is same as that of the supersymmetric $S^2 \times T^2$ solutions | In particular, the $D=5$ potential Admits a secOnd, noN-suPerSyMmetRic $ADS_5$ critical locuS (FirsT found by Romans [@Romans:1984an] In $D=5$ $N=8$ sUpERgraVItY), at \[RoMansCP\] e\^[4u]{}=E\^[-4V]{}=, =E\^E\^[I\^]{}, -a=iE\^[-]{}, wHeRe $\tHeTA^\pRime$ iS an ArbitraRy phase. All OthEr Scalars of the $d=5$ ThEory are set To tRivial, and the $adS_5$ Radius At TheSE poinTs iS $2{\sqrt 2}/3$.
feedinG (\[romanscP\]) into (\[uplIfTMet\]), (\[uplIFtflux\]), wE FInD two New classes of $D=11$ soluTIoNS, consisting on nOn-supeRsYMmETRic, WarPed and direCt ProduCTs, respeCTiVELY, of $aDS_5$ with a smooth, Complex maniFOld $m_6$, equipPeD, in BOth casEs, witH a hErmItian, two-parAmetEr metric. BOth claSSes are pARametriSed by $(\tIldE{a}, \pHi)$. To SEe ThIs, nOtE ThaT OuT of THe fIve paramEtErS $(\tildE{a},\tiLDE{C}, A, \phi, \TheTa^\prIme)$ of The solution (\[RoManSCP\]), (\[uPLifTmet\]), (\[uPliftFlux\]), $\ThEta^\prIme$ and $A$ can bE rEmoved by coordinAte tRansformaTioNs. the TwO clasSEs are tHen DisTinguisHed by thE ValUe OF $\TIlDe c$. If $\tilde{c} \neq 0$, it is EaSY To Check thaT it can BE tAkEN to $\tilde{C}=1$ wIth No loSS Of genEralITy, Leading tO the twO-PaRaMeter waRpEd prodUcT clAss. if $\tilDE{c}=0$, thE warp fActor $\DelTa$ redUCes to a constant, LEading to the diREcT PRoDUct cLasS. The non-supeRsymMEtry Of thE $d=5$ cRitICal poInt (\[RoMaNSCp\]) Carries over to its $D=11$ upLiFt. This Lack oF supersymmetrY can also be DIREctly seeN in $D=11$, AS tHE uplifted solutIons dO not fit intO The generIc forM of $AdS_5$ suPersymmetRIC solutioNs iN $D=11$ [@aDs5m6], RevIEWeD in subsection \[GEOmetRiEs\].
The toPolOgy of thE inTerNal $m_6$ foR bOth classeS is the saMe As ThAt Of tHe supERsymmetrIc $s^2 \tiMeS T^2$ sOlutiONs | In particular, the $D=5$potentialadmit s a se co nd,non- supersymmetric $AdS _5$ critical locus (fi rst f ou n d by Ro mans[@Roman s :1 9 8 4an ]in $D =5 $ $ N=8$sup ergravi ty), at \[ Rom an sCP\] e\^[4U ] {} =e\^[-4V]{ }=, =e\^e\^[i\^ ]{} , -a=i e\ ^[- ] {}, w her e $\t heta^\ p rime$is an arb it r ary ph a se. All o th er s calars of the $D= 5 $t heory are setto tri vi a l, a ndthe $AdS_5$ r ad ius a t thesep oi n t s is $2{\sqrt 2}/3 $.
Feeding (\[ Romans CP \]) into ( \[upl if t met \]), (\[upl iftf lux\]), w e find two new classes of $D =11 $ s olut i on s, co ns i sti n gonn on- supersym me tr ic, w arpe d a n d di rec t pr oduct s, respective ly, of$ AdS _5$ w ith a smo ot h, co mplexmanif ol d $M_6$, equipp ed,in both c ase s, wi th a He r mitian , t wo- paramet er metr i c.Bo t h cl asses are parametr is e d b y $(\til de{a}, \p hi ) $. To se ethi s, n o t e tha t ou t o f the fi ve par a me te rs $(\t il de{a}, \t ild e{c }, a, \phi , \the ta^\prim e)$ o f the solution( \[RomansCP\]) , ( \ [ up l iftm et\ ]), (\[upli ftfl u x\]) , $\ t he ta^ \ prime $ and $ a $c an be removed by co or dinate tran sformations.The two cl a s s es are t hend is t inguished by t he va lue of $\t i lde c$.If $\ tilde{c} \neq 0$, i t is eas y t o c hec k t h a tit can be tak e n to$\ tilde{c }=1 $ withnolos s o f g en erality,leadingto t he t wo- param e ter warp ed pr od uct clas s . If $ \tild e{c} =0 $, the warp f a ct o r $\D el ta $ re duc es to a con s tan t, lead ing to th e d i rect p ro duct cl ass. The non- su persymmetr yofthe $D = 5 $ critic al point (\[RomansCP\]) carries ov er to its $D=11$ u pli ft. Th isl ack of super symme tr y c a n also b edir ec tly seen i n $D= 11$,as the uplift ed solutions do no t fi t into the ge ner ic f o r mof$ Ad S _5$ s u per s y mmetric solutio ns in $D=1 1$ [@ ads5m6], r e vie we d in su bsectio n \[g e ometrie s\].
The topology o f th e int ernal $M_6 $ for bo th classe s is t h esameasthat o fthe supe rsymme t ric $S^2 \time sT^2$ s oluti on s | In_particular, the_$D=5$ potential admits a_second, non-supersymmetric_$AdS_5$_critical locus_(first_found by Romans_[@Romans:1984an] in $D=5$_$N=8$ supergravity), at \[RomansCP\]_e\^[4U]{}=e\^[-4V]{}=, =e\^e\^[i\^]{}, -a=ie\^[-]{},_where_$\theta^\prime$ is an arbitrary phase. All other scalars of the $D=5$ theory are set_to_trivial, and_the_$AdS_5$_radius at these points is_$2{\sqrt 2}/3$.
Feeding (\[RomansCP\]) into (\[upliftmet\]),_(\[upliftflux\]), we_find two new classes of $D=11$ solutions, consisting_on_non-supersymmetric, warped and_direct products, respectively, of $AdS_5$ with a smooth, complex_manifold $M_6$, equipped, in both cases,_with a Hermitian,_two-parameter_metric._Both classes are parametrised_by $(\tilde{a}, \phi)$. To see this,_note that out of the five_parameters $(\tilde{a},\tilde{c}, a, \phi, \theta^\prime)$ of the_solution (\[RomansCP\]), (\[upliftmet\]), (\[upliftflux\]), $\theta^\prime$ and_$a$ can be removed by_coordinate transformations._The two classes are then_distinguished by the_value of_$\tilde c$. If_$\tilde{c} \neq 0$, it is easy_to check that_it can be taken to $\tilde{c}=1$_with_no loss of_generality,_leading_to the_two-parameter warped product_class._If $\tilde{c}=0$,_the_warp factor $\Delta$ reduces to a_constant,_leading to the direct product class. The_non-supersymmetry of the $D=5$_critical_point (\[RomansCP\]) carries over_to its $D=11$ uplift. This_lack of supersymmetry can also be_directly seen_in $D=11$,_as the uplifted solutions do not fit into the generic form_of $AdS_5$ supersymmetric solutions in $D=11$_[@ads5m6], reviewed in subsection_\[geometries\].
The topology_of_the internal $M_6$_for_both classes_is the same as that of the_supersymmetric $S^2_\times T^2$ solutions |
--------------
: Table for different values of $a$ and $Q$ for ABG black hole. Parameter $\delta^{a}$ is the region between static limit surface and event horizon ($\delta^{a}=r_{+}^{sls}-r^{EH}_{+}$).
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}
{width="0.230\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
How the charge parameter $Q$ affects the shape of the ergoregion is shown in Fig. \[figB\], when compared with the Kerr black hole’s ergoregion (cf. Fig. \[figA\]). The numerical solution of $g_{tt} =0$ and $g^{rr} =0$ are summarized in Table \[Table A\]. Fig. \[figB\] shows that the ergoregion area of the ABG black hole increases with increase in | --------------
: Table for different values of $ a$ and $ Q$ for ABG black fix. Parameter $ \delta^{a}$ is the area between static limit open and consequence horizon ($ \delta^{a}=r_{+}^{sls}-r^{EH}_{+}$).
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
! [ image](cnt1.eps){width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth " }
! [ image](cnt5.eps){width="0.230\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth " }
! [ image](cnt9.eps){width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth " }
! [ image](cnt13.eps){width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth " }
! [ image](cnt17.eps){width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth " }
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
How the charge parameter $ Q$ affects the form of the ergoregion is shown in Fig. \[figB\ ], when compared with the Kerr bootleg fix ’s ergoregion (cf. Fig. \[figA\ ]). The numeral solution of $ g_{tt } = 0 $ and $ g^{rr } = 0 $ are summarized in Table \[Table A\ ]. Fig. \[figB\ ] shows that the ergoregion sphere of the ABG black hole increases with increase in | --------------
: Table for different valmes of $a$ and $Q$ for ABG ulack hkle. Paraoeter $\delta^{a}$ is the region bxtwewn stqtic limit surface and event hogizon ($\delra^{a}=r_{+}^{wos}-r^{EH}_{+}$).
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
{aidtk="0.245\lmnewidth"}{width="0.245\linawidth"}{wndth="0.245\linewidth"}{width="0.245\linewideh"}
{wpdeh="0.230\lihvwldth"}{width="0.245\linewidth"}{width="0.245\linewidyh"}{width="0.245\linewidtj"}
{width="0.245\lineqidtr"}{wkdth="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\uinewndth"}
{wivth="0.245\linvwidth"}{widtv="0.245\linewicth"}{eidvh="0.245\libewidth"}{widvh="0.245\linewidth"}
{width="0.245\linewidth"}{widdh="0.245\livwwiath"}{aidvh="0.245\linewidth"}{qidth="0.245\linewidth"}
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
How the dharge pwrameter $Q$ affects the shape of the ergmrefion is shown in Fig. \[figV\], when compared with jhe Kerr bjack hole’s ergoregion (cf. Fig. \[figA\]). The numerical sonutioi uf $y_{bb} =0$ ave $h^{rr} =0$ are summarized in Table \[Table A\]. Fig. \[figB\] srkwx nhat the ergoregijn area of yhf SFG black hole increcaea with increase in | -------------- : Table for different values of $Q$ ABG black Parameter $\delta^{a}$ is surface event horizon ($\delta^{a}=r_{+}^{sls}-r^{EH}_{+}$). -- -- -- {width="0.230\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} {width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} {width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} {width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- How the charge parameter $Q$ affects the shape of the ergoregion is in Fig. \[figB\], when compared with the Kerr black hole’s ergoregion (cf. Fig. The solution $g_{tt} and $g^{rr} =0$ are summarized in Table \[Table A\]. Fig. \[figB\] shows that the ergoregion area the ABG black hole increases with increase in | --------------
: Table for different values of $A$ and $Q$ for ABg blacK hoLe. PArAmetEr $\deLta^{a}$ is the regioN BetwEen static limit surface aNd eveNt HOrizON ($\dElta^{a}=R_{+}^{sls}-r^{EH}_{+}$).
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
{wIdtH="0.245\lINeWidth"}{width="0.245\lineWidTh"}{wIDtH="0.245\linewidth"}{widtH="0.245\liNewidtH"}
{wiDth="0.230\liNewidtH"}{widtH="0.245\lINewidtH"}{WidtH="0.245\linewidth"}{Width="0.245\linewidth"}
{WIDth="0.245\LinEwidth"}{WIdth="0.245\linEWiDTH"}{width="0.245\lInewidth"}{wiDtH="0.245\liNEwidth"}
{Width="0.245\linewiDth"}{width="0.245\LInewidtH"}{wiDth="0.245\LinEwidTH"}{wIDtH="0.245\liNEwiDth"}
{widTh="0.245\liNEWIDth"}{wiDth="0.245\linewidth"}{wIdth="0.245\lInewiDth"}{widTh="0.245\linEwIdth"}
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
How the chargE parAmeter $Q$ afFecTs The ShApe of THe ergoRegIon Is shown In Fig. \[fiGb\], whEn COMPaRed with the Kerr blacK hOLE’s ErgoregiOn (cf. FiG. \[FiGA\]). tHe numeriCaL soLutiON Of $g_{tt} =0$ And $g^{RR} =0$ aRe summarIzed in tAbLe \[table A\]. FIg. \[FigB\] shOwS thAt tHe ergORegiOn area Of the ABG Black HOle increases wiTH increase in | --------------
: Table for diffe rentval ues o f $a $ an d $Q$ for ABGb lack hole. Parameter $\del ta^{a }$ is t h eregio n betwe e ns t ati cli mit s u rf ace a ndevent h orizon ($\ del ta ^{a}=r_{+}^{ s ls }-r^{EH}_{ +}$ ).
------ --- ------ -- --- - ----- --- ----- ------ - ------ --------- -- - ------ - ------- - - -- ---- ----------------- - -- - -------------- ------ -- - -- - - --- --- ---------- -- ----- - ------- - -- - - - --- - ----------- - - -- --
{wi dth=" 0. 2 45\ linewidth"} { w idth="0 . 245\lin ewidth "}! [im age] ( cn t3 .ep s) { wid t h= "0. 2 45\ linewidt h" }! [imag e](c n t 4 . eps) {wi dth= "0.24 5\linewidth"} ! [imag e](c nt 5.eps ){widt h="0. 23 0\linewidth"} {w idt h= "0.24 5 \linew idt h"}  {w i d t h= "0.245\linewidth"} {w i dt h= " 0.245\li ne wid th"} ! [image]( cnt9.e p s) {w idth="0 .2 45\lin ew idt h"} {widt h="0. 2 45\linewidth"} ! [image](cnt11 . ep s ) {w i dth= "0. 245\linewid th"} ! [ima ge]( c nt 12. e ps){w idth= "0 . 24 5 \linewidth"}
![i mage] (cnt13.eps){w idth="0.24 5 \ l inewidth "}{ width ="0.245\li n ewidth"} {wi d t h="0.245 \li new idt h"} ! [ im age](cnt16.ep s ) {wid th ="0.245 \li newidth "}
! [image](c nt17.eps ){ wi dt h= "0. 245\l i newidth" }! [im ag e]( cnt18 . eps){w idth= "0.2 45 \l i new idth"}! [ im a g e](c nt 19 .eps ){w id th="0 .245 \ lin ewidth" }![image] (cn t 20.e ps ){ width=" 0.245\linewid th "}
-- ------ - - -------- ----------------------- - ------- --- ----- ---- --------- --- ------ --- - ------ ------ ----- -- --- - - ----- - - -- --- -- ---------- - - --- ----- -- ---- ------- ------------------ - --- ----- -- -- - -
Howt h echa r ge par am e ter $ Q$ affects theshape of t he er goregion i s sh ow n in Fi g. \[fi gB\], when co mpared wi th the Ke rr bla c k ho le’s ergor egion (c f. Fig. \ [ figA\ ] ). Thenum erical s olu tionof $g_ { tt} =0$and $g ^{ rr} =0 $ are s ummarize d in Table \[Table A\]. Fig.\[fig B\] shows th att heergoregio n ar ea of theABG bl ack h ole incre ases wi thi ncrea se i n | --------------
_ :_Table for different values_of $a$_and_$Q$ for_ABG_black hole. Parameter_$\delta^{a}$ is the_region between static limit_surface and event_horizon_($\delta^{a}=r_{+}^{sls}-r^{EH}_{+}$).
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -- -- --
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} __ ___
{width="0.230\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}_ _ _
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} __ _
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} _
{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"}{width="0.245\linewidth"} _ ___
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------_-- -- --
How the charge parameter_$Q$ affects the shape of the_ergoregion is shown in Fig. \[figB\], when compared_with the Kerr black hole’s ergoregion_(cf. Fig. \[figA\]). The numerical solution_of $g_{tt}_=0$ and $g^{rr} =0$ are_summarized in Table \[Table_A\]. Fig. \[figB\]_shows that the_ergoregion area of the ABG black_hole increases with_increase in |
G$.
Suppose, for contradiction, that $$c: = \max_{x \in G} ( u(0,x) - h(x) ) > 0.$$ Note that the maximum $c$ is finite and achieved in the interior since $h(x) \ge u^{*}(x) \geq u(0,x) $ for $x \in \partial G$ and $u$ and $h$ are continuous. Let $x_0 \in G$ be such that $$c = u(0,x_0) - h(x_0).$$ Consider now the function $h_c = h + c$. By choice of $c$, we have that $u(0,x) \le h_c(x)$ for $x \in G$, and by the counter-assumption $u \leq h_c$ also on $[0,T) \times \partial G$. It follows from the assumption on $u$ that $u \le h + c$ in $(0,T) \times G$. Thus, by the definition of $c$ and $x_0$, we have for all $t \in (0,T)$ that $$u(t,x_0) \le h(x_0) + c = u (0, x_0).$$ Taking the supremum in $t$, we obtain $u^{*}(x_0) \le u(0,x)$, which violates $x_0 \in G \subset E$. This contradiction implies $c \le 0$.
Consequently, we have $u(0,x) \le h(x)$ for $x \in \overline{G}$, and further $u(t,x) \leq h(x)$ for all $(t,x) \in [0,T) \times \partial G$ by the choice of $h$. The assumed comparison property of $u$ yields $u(t,x) \leq h(x)$ for all $(t,x) \in (0,T) \times G$. Taking the supremum in $t$ implies $u^{*}(x) \leq h(x)$ for all $x \in G$ as claimed.
We use this lemma to give the
Let $T > 1$ and define the truncated maximal function $$S^{*,T} f(x) | G$.
Suppose, for contradiction, that $ $ c: = \max_{x \in G } ( u(0,x) - h(x) ) > 0.$$ Note that the maximum $ c$ is finite and achieve in the department of the interior since $ h(x) \ge u^{*}(x) \geq u(0,x) $ for $ x \in \partial G$ and $ u$ and $ h$ are continuous. Let $ x_0 \in G$ be such that $ $ c = u(0,x_0) - h(x_0).$$ think immediately the function $ h_c = h + c$. By option of $ c$, we have that $ u(0,x) \le h_c(x)$ for $ x \in G$, and by the counter - assumption $ uracil \leq h_c$ also on $ [ 0,T) \times \partial G$. It follows from the presumption on $ u$ that $ u \le h + c$ in $ (0,T) \times G$. Thus, by the definition of $ c$ and $ x_0 $, we get for all $ t \in (0,T)$ that $ $ u(t, x_0) \le h(x_0) + c = u (0, x_0).$$ Taking the supremum in $ t$, we obtain $ u^{*}(x_0) \le u(0,x)$, which violates $ x_0 \in G \subset E$. This contradiction implies $ coke \le 0$.
Consequently, we have $ u(0,x) \le h(x)$ for $ x \in \overline{G}$, and further $ u(t, x) \leq h(x)$ for all $ (t, x) \in [ 0,T) \times \partial G$ by the choice of $ h$. The bear comparison property of $ u$ yields $ u(t, x) \leq h(x)$ for all $ (t, x) \in (0,T) \times G$. Taking the supremum in $ t$ implies $ u^{*}(x) \leq h(x)$ for all $ x \in G$ as claim.
We use this lemma to give the
Let $ T > 1 $ and define the abbreviated maximal function $ $ S^{*,T } f(x ) | G$.
Skppose, for contradiction, that $$c: = \max_{x \iu G} ( n(0,x) - h(x) ) > 0.$$ Note that the maximum $c$ is finitx ane achueved in the interior rince $h(x) \he u^{*}(x) \gew u(0,x) $ for $x \in \'zrtial N$ and $m$ and $i$ are continuoux. Let $x_0 \in G$ be such thad $$z = u(0,x_0) - h(x_0).$$ Consider now the functiog $h_c = h + f$. By choice of $c$, wt hwve fhat $u(0,x) \le h_c(x)$ for $x \in G$, and by tge counuer-assumption $u \lea h_c$ also on $[0,T) \times \partiwl G$. It follows from tje assumptiin og $u$ that $u \le h + c$ in $(0,T) \times G$. Thua, by the definition of $c$ and $x_0$, de hare for all $j \jn (0,J)$ that $$u(t,x_0) \lx h(x_0) + b = u (0, x_0).$$ Takikb the vupremuk in $t$, we obtaln $u^{*}(x_0) \le u(0,x)$, which violates $x_0 \mn G \subset E$. This sontradicdikn implies $c \le 0$.
Cinwequettly, we fqve $u(0,x) \lx h(s)$ for $d \ii \overline{G}$, and furthee $u(t,x) \leq h(x)$ for alk $(e,q) \in [0,T) \times \lartiaj D$ by the choice of $h$. The assumed comparpson property of $u$ yields $u(r,x) \leq h(x)$ for all $(t,x) \ln (0,T) \timef G$. Taking the supremum in $t$ implies $u^{*}(x) \leq h(x)$ fos all $b \iu G$ as xlwimed.
We use this lemma to give the
Let $T > 1$ and segike the truncated maximal fumchipg $$S^{*,T} f(x) | G$. Suppose, for contradiction, that $$c: = G} u(0,x) - ) > 0.$$ is and achieved in interior since $h(x) u^{*}(x) \geq u(0,x) $ for $x \partial G$ and $u$ and $h$ are continuous. Let $x_0 \in G$ be that $$c = u(0,x_0) - h(x_0).$$ Consider now the function $h_c = h c$. choice $c$, have that $u(0,x) \le h_c(x)$ for $x \in G$, and by the counter-assumption $u \leq h_c$ on $[0,T) \times \partial G$. It follows from assumption on $u$ that \le h + c$ in \times Thus, by definition $c$ $x_0$, we have all $t \in (0,T)$ that $$u(t,x_0) \le h(x_0) + c = u (0, x_0).$$ Taking the supremum $t$, we \le u(0,x)$, violates \in \subset E$. This $c \le 0$. Consequently, we have for $x \in \overline{G}$, and further $u(t,x) \leq for all \in [0,T) \times \partial G$ by choice of $h$. The assumed comparison property of yields $u(t,x) \leq h(x)$ for all $(t,x) \in (0,T) \times G$. Taking the supremum in $u^{*}(x) \leq h(x)$ for $x \in G$ claimed. use lemma give the $T > 1$ and define the truncated maximal function $$S^{*,T} f(x) | G$.
Suppose, for contradiction, tHat $$c: = \max_{x \in g} ( u(0,x) - h(x) ) > 0.$$ notE thAt The mAximUm $c$ is finite and AChieVed in the interior since $h(X) \ge u^{*}(x) \GeQ U(0,x) $ foR $X \iN \partIal G$ and $U$ AnD $H$ Are CoNtInuOuS. leT $x_0 \in G$ Be sUch that $$C = u(0,x_0) - h(x_0).$$ ConsiDer NoW the function $H_C = h + C$. By choice oF $c$, wE have that $u(0,x) \lE h_c(X)$ for $x \iN G$, And BY the cOunTer-asSumptiON $u \leq h_C$ also on $[0,T) \tImES \partiAL G$. It folLOWs From The assumption on $u$ tHAt $U \Le h + c$ in $(0,T) \times G$. THus, by tHe DEfINItiOn oF $c$ and $x_0$, we haVe For alL $T \in (0,T)$ thaT $$U(t,X_0) \LE H(x_0) + c = U (0, X_0).$$ Taking the supRemum in $t$, we oBTaiN $u^{*}(x_0) \le u(0,X)$, wHicH ViolatEs $x_0 \in g \sUBseT E$. This contrAdicTion impliEs $c \le 0$.
CONsequenTLy, we havE $u(0,x) \le h(X)$ foR $x \iN \oveRLiNe{g}$, anD fURthER $u(T,x) \lEQ h(x)$ For all $(t,x) \In [0,t) \tImes \pArtiAL g$ BY the ChoIce oF $h$. The Assumed comparIsoN proPErtY of $u$ yIelds $U(t,x) \lEq H(x)$ for All $(t,x) \iN (0,T) \timEs g$. Taking the supreMum iN $t$ implies $U^{*}(x) \lEq H(x)$ fOr All $x \iN g$ as claImeD.
We Use this Lemma to GIve ThE
lET $T > 1$ And define the truncaTeD MAxImal funcTion $$S^{*,T} F(X) | G$.
Suppose, for contrad iction, th at $$ c:= \ ma x_{x \in G} ( u(0,x) - h (x) ) > 0.$$ Note that thema x imum $c $ isfinitea nd a chi ev ed in t h einter ior since$h(x) \geu^{ *} (x) \geq u(0 , x) $ for $x\in \partial G$ an d $u$an d $ h $ are co ntinu ous. L e t $x_0 \in G$ b es uch th a t $$c = u (0,x _0) - h(x_0).$$C on s ider now the f unctio n$ h_ c = h +c$. By cho ic e of$ c$, weh av e t hat $u(0,x) \le h _c(x)$ for$ x \ in G$, a ndb y thecount er - ass umption $u\leq h_c$ als o on $ [ 0,T) \t i mes \pa rtialG$. It fol l ow sfro mt hea ss ump t ion on $u$th at $u \ le h + c $ in $( 0,T) \tim es G$. Thus,bythed efi nitio n of$c$an d $x_ 0$, we have f or all $t \in ( 0,T) $ that $$ u(t ,x _0) \ le h( x _0) +c = u (0, x_ 0).$$ T a kin gt h e s upremum in $t$, we o b t ai n $u^{*} (x_0)\ le u ( 0,x)$, w hi chviol a t es $x _0 \ i nG \subs et E$. Th is contra di ctionim pli es$c \l e 0$.
Cons equently , weh ave $u(0,x) \l e h(x)$ for $x \i n \o v erli ne{ G}$, and fu rthe r $u( t,x) \l eqh (x)$for a ll $( t ,x) \in [0,T) \time s\parti al G$ by the choic e of $h$.T h e assumed com p ar i son property o f $u$ yields $u ( t,x) \le q h(x )$ for a ll $(t,x) \ in (0,T) \t ime s G $.T a ki ng the suprem u m in$t $ impli es$u^{*}( x)\le q h (x) $for all $ x \in G$ a scl ai med .
We use this l emm atogivet he
Le t $T> 1$ a nd def ine the tr u n cate dma xima l f un ction $$S ^ {*, T} f(x) | G$.
Suppose,_for contradiction,_that $$c: = \max_{x_\in G}_(_ _u(0,x)_ - h(x)_) > 0.$$_Note that the maximum_$c$ is finite_and_achieved in the interior since $h(x) \ge u^{*}(x) \geq u(0,x) $ for $x \in_\partial_G$ and_$u$_and_$h$ are continuous. Let $x_0_\in G$ be such that_$$c =_ u(0,x_0) - h(x_0).$$ Consider now_the_function $h_c =_h + c$. By choice of $c$, we have_that $u(0,x) \le h_c(x)$ for $x_\in G$, and_by_the_counter-assumption $u \leq h_c$_also on $[0,T) \times \partial G$._It follows from the assumption on_$u$ that $u \le h + c$_in $(0,T) \times G$. Thus, by_the definition of $c$ and_$x_0$, we_have for all $t \in_(0,T)$ that $$u(t,x_0)_\le h(x_0)_+ c =_ u (0, x_0).$$ Taking the_supremum in $t$,_we obtain $u^{*}(x_0) \le u(0,x)$, which_violates_$x_0 \in G__\subset_E$. This_contradiction implies $c_\le_0$.
Consequently, we_have_$u(0,x) \le h(x)$ for $x \in_\overline{G}$,_and further $u(t,x) \leq h(x)$ for all_$(t,x) \in [0,T) \times_\partial_G$ by the choice_of $h$. The assumed comparison_property of $u$ yields $u(t,x) \leq_h(x)$ for_all $(t,x)_\in (0,T) \times G$. Taking the supremum in $t$ implies $u^{*}(x)_\leq h(x)$ for all $x \in_G$ as claimed.
We use_this lemma_to_give the
Let $T_>_1$ and_define the truncated maximal function $$S^{*,T} f(x) |
!\tau_1\mu_k
\frac{M(r)}{k\,r}+ \tau_3\frac{\mathcal{U}(r)}{k\,r}\bigg)
\bigg(1+\frac{15\delta_{{\boldsymbol}G}^{{13}/{6}}}{q^{{1}/{2}}}\bigg)
\bigg(1+\frac{M_{r}}{q}\bigg),$$ where $\mu_k$ and $\tau_3$ are as in Lemma \[Costo\] and $M_{r}$ is defined as in Theorem \[average de DDF\].
According to, we estimate the probability $P[\mathcal {A}_{j, k}^ {sq} ]$ that a random $ f \in \mathcal { A}$ is square–free and has $ j $ irreducible factors of degree $ k $. In [@KnKn90b] it is shown that, if $ q $ is sufficiently large, then the probability that a random $ f \in {{\mathbb F}_{\hskip-0.7mm q}}[T] $ of degree at most $r$ has $j$ distinct irreducible factors of degree $ k $ tends to $e^{-1/k}\frac{k^{-j}}{j!}$.
We decompose the set $\mathcal{A}_{j,k}^{sq}$ into the disjoint union $$\mathcal{A}_{j,k}^{sq}=\bigcup_{{\boldsymbol}\lambda \in
\mathcal{P}_r^{j,k}} \mathcal{A}_{j,{\boldsymbol}\lambda}^{sq},$$ where $\mathcal{P}_{r}^{j,k}$ is the set of all $ r $–tuples $ {\boldsymbol}\lambda: = (\lambda_1, \dots, \lambda_r) \in \mathbb{Z}_{\geq0}^r $ with $\lambda_1 + \dots + r \, \lambda_r = r $ and $ \lambda_k = j
$. Hence, we have $$P[\mathcal{A}_{j,k}^{sq}]=\frac{1}{|\mathcal{A}|}
\sum_{{\boldsymbol}\lambda \in \mathcal{P}_{r}^{j,k}} |\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}|.$$ From Theorem \[theorem: estimate fact patterns\] we deduce that $$|\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}| \leq q^{r-m}\,\mathcal{T}({\boldsymbol}\lambda)\bigg( | ! \tau_1\mu_k
\frac{M(r)}{k\,r}+ \tau_3\frac{\mathcal{U}(r)}{k\,r}\bigg)
\bigg(1+\frac{15\delta_{{\boldsymbol}G}^{{13}/{6}}}{q^{{1}/{2}}}\bigg)
\bigg(1+\frac{M_{r}}{q}\bigg),$$ where $ \mu_k$ and $ \tau_3 $ are as in Lemma \[Costo\ ] and $ M_{r}$ is defined as in Theorem \[average de DDF\ ].
According to, we estimate the probability $ P[\mathcal { A}_{j, k}^ { sq } ] $ that a random $ f \in \mathcal { A}$ is square – spare and have $ j $ irreducible factors of degree $ potassium $. In [ @KnKn90b ] it is shown that, if $ q $ is sufficiently large, then the probability that a random $ f \in { { \mathbb F}_{\hskip-0.7 mm q}}[T ] $ of academic degree at most $ r$ has $ j$ distinct irreducible factor of degree $ k $ tend to $ e^{-1 / k}\frac{k^{-j}}{j!}$.
We decompose the set $ \mathcal{A}_{j, k}^{sq}$ into the disjoint coupling $ $ \mathcal{A}_{j, k}^{sq}=\bigcup_{{\boldsymbol}\lambda \in
\mathcal{P}_r^{j, k } } \mathcal{A}_{j,{\boldsymbol}\lambda}^{sq},$$ where $ \mathcal{P}_{r}^{j, k}$ is the set of all $ r $ – tuples $ { \boldsymbol}\lambda: = (\lambda_1, \dots, \lambda_r) \in \mathbb{Z}_{\geq0}^r $ with $ \lambda_1 + \dots + r \, \lambda_r = r $ and $ \lambda_k = j
$. Hence, we have $ $ P[\mathcal{A}_{j, k}^{sq}]=\frac{1}{|\mathcal{A}| }
\sum_{{\boldsymbol}\lambda \in \mathcal{P}_{r}^{j, k } } |\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}|.$$ From Theorem \[theorem: estimate fact patterns\ ] we deduce that $ $ |\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}| \leq q^{r - m}\,\mathcal{T}({\boldsymbol}\lambda)\bigg ( | !\tau_1\lu_k
\frac{M(r)}{k\,r}+ \tau_3\frac{\mathcxl{U}(r)}{k\,r}\bigg)
\bigg(1+\ftax{15\delta_{{\uoldsymgol}G}^{{13}/{6}}}{q^{{1}/{2}}}\bige)
\bigg(1+\frac{M_{r}}{q}\bigg),$$ where $\mu_k$ aid $\tqu_3$ art as in Lemma \[Costo\] and $M_{r}$ id definee as un Theorem \[average de DDR\].
Wccoxdmng to, we estimste the prmbability $P[\matvcxl {A}_{j, k}^ {sq} ]$ that a random $ f \in \mathcwl { A}$ ix dquare–free and has $ j $ jgrtducible factors of degree $ k $. In [@KnKn90b] pt is shown that, of $ q $ is sufficiently larhe, tjen the probabilitj that a rabdom $ f \in {{\mathbb F}_{\hskip-0.7mm q}}[T] $ of degreg at most $r$ has $j$ distinct irredjciblz factors od eeggge $ k $ tends to $e^{-1/h}\frac{k^{-j}}{j!}$.
We dcbompose the sey $\mathcal{A}_{j,k}^{sq}$ invo tye disjoint union $$\matical{A}_{j,k}^{sq}=\bigcup_{{\boldsimbol}\lambdd \nn
\mathcal{P}_r^{j,k}} \mathcal{Q}_{j,{\voldsimbol}\nambaq}^{sq},$$ whtre $\mzthcal{O}_{r}^{j,i}$ is the sst of all $ e $–tuples $ {\boldsymbok}\lwnbda: = (\lambda_1, \sots, \lwmfda_r) \in \mathbb{Z}_{\geq0}^r $ with $\lambda_1 + \dots + r \, \lambda_r = r $ and $ \lambdq_k = j
$. Hence, we have $$P[\lathcal{A}_{j,h}^{sq}]=\frac{1}{|\mathcal{A}|}
\sum_{{\boldsymbol}\lambda \in \mathcal{P}_{r}^{j,k}} |\mathrau{A}_{j,{\yildsyovop}\lambda}^{sq}|.$$ From Theorem \[theorem: estimate fact [ztuerks\] we deduce thab $$|\mathcal{A}_{j,{\boldsymnop}\lsibda}^{sq}| \leq q^{r-o}\,\mathccm{T}({\goldsymbol}\lambda)\bihg( | !\tau_1\mu_k \frac{M(r)}{k\,r}+ \tau_3\frac{\mathcal{U}(r)}{k\,r}\bigg) \bigg(1+\frac{15\delta_{{\boldsymbol}G}^{{13}/{6}}}{q^{{1}/{2}}}\bigg) \bigg(1+\frac{M_{r}}{q}\bigg),$$ where $\mu_k$ are in Lemma and $M_{r}$ is de According to, we the probability $P[\mathcal k}^ {sq} ]$ that a random f \in \mathcal { A}$ is square–free and has $ j $ irreducible of degree $ k $. In [@KnKn90b] it is shown that, if $ $ sufficiently then probability that a random $ f \in {{\mathbb F}_{\hskip-0.7mm q}}[T] $ of degree at most $r$ $j$ distinct irreducible factors of degree $ k tends to $e^{-1/k}\frac{k^{-j}}{j!}$. We the set $\mathcal{A}_{j,k}^{sq}$ into the union \in \mathcal{P}_r^{j,k}} where is set of all r $–tuples $ {\boldsymbol}\lambda: = (\lambda_1, \dots, \lambda_r) \in \mathbb{Z}_{\geq0}^r $ with $\lambda_1 + \dots + r \lambda_r = and $ = $. we have $$P[\mathcal{A}_{j,k}^{sq}]=\frac{1}{|\mathcal{A}|} \mathcal{P}_{r}^{j,k}} |\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}|.$$ From Theorem \[theorem: estimate deduce that $$|\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}| \leq q^{r-m}\,\mathcal{T}({\boldsymbol}\lambda)\bigg( | !\tau_1\mu_k
\frac{M(r)}{k\,r}+ \tau_3\frac{\mathCal{U}(r)}{k\,r}\bigG)
\bigg(1+\FraC{15\deLtA_{{\bolDsymBol}G}^{{13}/{6}}}{q^{{1}/{2}}}\bigg)
\bigg(1+\fRAc{M_{r}}{Q}\bigg),$$ where $\mu_k$ and $\tau_3$ are As in LEmMA \[CosTO\] aNd $M_{r}$ iS defineD As IN theOrEm \[AveRaGE dE DDF\].
ACcoRding to, We estimate The PrObability $P[\maTHcAl {A}_{j, k}^ {sq} ]$ thaT a rAndom $ f \in \mathCal { a}$ is squArE–frEE and hAs $ j $ IrredUcible FActors Of degree $ k $. in [@kNKn90b] it IS shown tHAT, iF $ q $ is Sufficiently large, THeN The probability That a rAnDOm $ F \IN {{\maThbB F}_{\hskip-0.7mm q}}[t] $ oF degrEE at most $R$ HaS $J$ DIstINct irreduciblE factors of dEGreE $ k $ tendS tO $e^{-1/k}\FRac{k^{-j}}{j!}$.
we decOmPOse The set $\mathcAl{A}_{j,K}^{sq}$ into thE disjoINt union $$\MAthcal{A}_{J,k}^{sq}=\biGcuP_{{\boLdsyMBoL}\lAmbDa \IN
\maTHcAl{P}_R^{J,k}} \mAthcal{A}_{j,{\BoLdSymboL}\lamBDA}^{SQ},$$ wheRe $\mAthcAl{P}_{r}^{j,K}$ is the set of alL $ r $–tUpleS $ {\BolDsymbOl}\lamBda: = (\lAmBda_1, \doTs, \lambDa_r) \in \MaThbb{Z}_{\geq0}^r $ with $\laMbda_1 + \Dots + r \, \lambDa_r = R $ aNd $ \lAmBda_k = j
$. hEnce, we HavE $$P[\mAthcal{A}_{J,k}^{sq}]=\fraC{1}{|\MatHcAL{a}|}
\SuM_{{\boldsymbol}\lambda \iN \mATHcAl{P}_{r}^{j,k}} |\maThcal{A}_{J,{\BoLdSYmbol}\lamBdA}^{sq}|.$$ from tHEorem \[TheoREm: Estimate Fact paTTeRnS\] we deduCe That $$|\maThCal{a}_{j,{\bOldsyMBol}\lAmbda}^{sQ}| \leq q^{r-m}\,\mAthcaL{t}({\boldsymbol}\lamBDa)\bigg( | !\tau_1\mu_k
\frac{M(r)}{k \,r}+ \tau _3\fr ac{ \ma th cal{ U}(r )}{k\,r}\bigg) \big g(1+\frac{15\delta_{{\ bolds ym b ol}G } ^{ {13}/ {6}}}{q ^ {{ 1 } /{2 }} }\ big g) \b igg(1 +\f rac{M_{ r}}{q}\big g), $$ where $\mu_ k $and $\tau_ 3$are as in Le mma \[Cos to \]a nd $M _{r }$ is defin e d as i n Theorem \ [ averag e de DDF \ ] .
Acc ording to, we est i ma t e the probabil ity $P [\ m at h c al{A} _{j, k}^ { sq } ]$t hat a r a nd o m $ f \in \mathcal{ A}$ is sq u are –freean d h a s $ j$ irr ed u cib le factorsof d egree $ k $. In [@KnKn9 0 b] it i s show n t hat , if $q$ i ss uff i ci ent l y l arge, th en t he pr obab i l i t y th ata ra ndom$ f \in {{\ma thb b F} _ {\h skip- 0.7mm q}} [T ] $ o f degr ee at m ost $r$ has $j$ dis tinct irr edu ci ble f actor s of de gre e $ k $ te nds to$ e^{ -1 / k } \f rac{k^{-j}}{j!}$.
W e de composethe se t $ \m a thcal{A} _{ j,k }^{s q } $ int o th e d isjointunion$ $\ ma thcal{A }_ {j,k}^ {s q}= \bi gcup_ { {\bo ldsymb ol}\lamb da \i n
\mathcal{P}_r ^ {j,k}} \mathc a l{ A } _{ j ,{\b old symbol}\lam bda} ^ {sq} ,$$w he re$ \math cal{P }_ { r} ^ {j,k}$ is the set o fall $r $–t uples $ {\bol dsymbol}\l a m b da: = (\ lamb d a_ 1 , \dots, \lamb da_r) \in \math b b{Z}_{\g eq0}^ r $ with $\lambda _ 1 + \dots +r \ , \ lam b d a_ r = r $ and $ \ lamb da _k = j$.Hence,wehav e $ $P[ \m athcal{A} _{j,k}^{ sq }] =\ fr ac{ 1}{|\ m athcal{A }| }
\ su m_{ {\bol d symbol }\lam bda\i n\ mat hcal{P} _ {r } ^ {j,k }} | \mat hca l{ A}_{j ,{\b o lds ymbol}\ lambda}^{ sq} | .$$Fr om Theore m \[theorem:es timate fac tpat terns\ ] we deduc e that $$|\mathcal{A}_{ j ,{\bold sym bol}\ lamb da}^{sq}| \l eq q^{ r-m } \,\mat hcal{T }({\b ol dsy m b ol}\l a m bd a)\ bi gg( | !\tau_1\mu_k
\frac{M(r)}{k\,r}+ \tau_3\frac{\mathcal{U}(r)}{k\,r}\bigg)
\bigg(1+\frac{15\delta_{{\boldsymbol}G}^{{13}/{6}}}{q^{{1}/{2}}}\bigg)
\bigg(1+\frac{M_{r}}{q}\bigg),$$_where $\mu_k$_and $\tau_3$ are as_in Lemma_\[Costo\]_and $M_{r}$_is_defined as in_Theorem \[average de_DDF\].
According to, we estimate_the probability $P[\mathcal_{A}_{j,_k}^ {sq} ]$ that a random $ f \in \mathcal { A}$ is square–free_and_has $_j_$_irreducible factors of degree $_k $. In [@KnKn90b] it_is shown_that, if $ q $ is sufficiently large,_then_the probability that_a random $ f \in {{\mathbb F}_{\hskip-0.7mm q}}[T] $_of degree at most $r$ has_$j$ distinct irreducible_factors_of_degree $ k $_tends to $e^{-1/k}\frac{k^{-j}}{j!}$.
We decompose the set_$\mathcal{A}_{j,k}^{sq}$ into the disjoint union $$\mathcal{A}_{j,k}^{sq}=\bigcup_{{\boldsymbol}\lambda_\in
\mathcal{P}_r^{j,k}} \mathcal{A}_{j,{\boldsymbol}\lambda}^{sq},$$ where $\mathcal{P}_{r}^{j,k}$ is the set_of all $ r $–tuples $_{\boldsymbol}\lambda: = (\lambda_1, \dots, \lambda_r)_\in \mathbb{Z}_{\geq0}^r_$ with $\lambda_1 + \dots_+ r \,_\lambda_r =_r $ and_$ \lambda_k = j
$. Hence, we_have $$P[\mathcal{A}_{j,k}^{sq}]=\frac{1}{|\mathcal{A}|}
\sum_{{\boldsymbol}\lambda \in_\mathcal{P}_{r}^{j,k}} |\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}|.$$ From Theorem \[theorem: estimate_fact_patterns\] we deduce_that_$$|\mathcal{A}_{j,{\boldsymbol}\lambda}^{sq}|_\leq q^{r-m}\,\mathcal{T}({\boldsymbol}\lambda)\bigg( |
Subsets and Splits