File size: 56,420 Bytes
c61ccee
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
import sys

import torch
from torch._C import _add_docstr, _fft  # type: ignore[attr-defined]
from torch._torch_docs import factory_common_args, common_args

__all__ = ['fft', 'ifft', 'fft2', 'ifft2', 'fftn', 'ifftn',
           'rfft', 'irfft', 'rfft2', 'irfft2', 'rfftn', 'irfftn',
           'hfft', 'ihfft', 'fftfreq', 'rfftfreq', 'fftshift', 'ifftshift',
           'Tensor']

Tensor = torch.Tensor

# Note: This not only adds the doc strings for the spectral ops, but
# connects the torch.fft Python namespace to the torch._C._fft builtins.

fft = _add_docstr(_fft.fft_fft, r"""

fft(input, n=None, dim=-1, norm=None, *, out=None) -> Tensor



Computes the one dimensional discrete Fourier transform of :attr:`input`.



Note:

    The Fourier domain representation of any real signal satisfies the

    Hermitian property: `X[i] = conj(X[-i])`. This function always returns both

    the positive and negative frequency terms even though, for real inputs, the

    negative frequencies are redundant. :func:`~torch.fft.rfft` returns the

    more compact one-sided representation where only the positive frequencies

    are returned.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimension.



Args:

    input (Tensor): the input tensor

    n (int, optional): Signal length. If given, the input will either be zero-padded

        or trimmed to this length before computing the FFT.

    dim (int, optional): The dimension along which to take the one dimensional FFT.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.fft`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the FFT orthonormal)



        Calling the backward transform (:func:`~torch.fft.ifft`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ifft`

        the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    >>> t = torch.arange(4)

    >>> t

    tensor([0, 1, 2, 3])

    >>> torch.fft.fft(t)

    tensor([ 6.+0.j, -2.+2.j, -2.+0.j, -2.-2.j])



    >>> t = torch.tensor([0.+1.j, 2.+3.j, 4.+5.j, 6.+7.j])

    >>> torch.fft.fft(t)

    tensor([12.+16.j, -8.+0.j, -4.-4.j,  0.-8.j])

""".format(**common_args))

ifft = _add_docstr(_fft.fft_ifft, r"""

ifft(input, n=None, dim=-1, norm=None, *, out=None) -> Tensor



Computes the one dimensional inverse discrete Fourier transform of :attr:`input`.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimension.



Args:

    input (Tensor): the input tensor

    n (int, optional): Signal length. If given, the input will either be zero-padded

        or trimmed to this length before computing the IFFT.

    dim (int, optional): The dimension along which to take the one dimensional IFFT.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.ifft`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the IFFT orthonormal)



        Calling the forward transform (:func:`~torch.fft.fft`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ifft`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> t = torch.tensor([ 6.+0.j, -2.+2.j, -2.+0.j, -2.-2.j])

    >>> torch.fft.ifft(t)

    tensor([0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j])

""".format(**common_args))

fft2 = _add_docstr(_fft.fft_fft2, r"""

fft2(input, s=None, dim=(-2, -1), norm=None, *, out=None) -> Tensor



Computes the 2 dimensional discrete Fourier transform of :attr:`input`.

Equivalent to :func:`~torch.fft.fftn` but FFTs only the last two dimensions by default.



Note:

    The Fourier domain representation of any real signal satisfies the

    Hermitian property: ``X[i, j] = conj(X[-i, -j])``. This

    function always returns all positive and negative frequency terms even

    though, for real inputs, half of these values are redundant.

    :func:`~torch.fft.rfft2` returns the more compact one-sided representation

    where only the positive frequencies of the last dimension are returned.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: last two dimensions.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.fft2`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the FFT orthonormal)



        Where ``n = prod(s)`` is the logical FFT size.

        Calling the backward transform (:func:`~torch.fft.ifft2`) with the same

        normalization mode will apply an overall normalization of ``1/n``

        between the two transforms. This is required to make

        :func:`~torch.fft.ifft2` the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    >>> x = torch.rand(10, 10, dtype=torch.complex64)

    >>> fft2 = torch.fft.fft2(x)



    The discrete Fourier transform is separable, so :func:`~torch.fft.fft2`

    here is equivalent to two one-dimensional :func:`~torch.fft.fft` calls:



    >>> two_ffts = torch.fft.fft(torch.fft.fft(x, dim=0), dim=1)

    >>> torch.testing.assert_close(fft2, two_ffts, check_stride=False)



""".format(**common_args))

ifft2 = _add_docstr(_fft.fft_ifft2, r"""

ifft2(input, s=None, dim=(-2, -1), norm=None, *, out=None) -> Tensor



Computes the 2 dimensional inverse discrete Fourier transform of :attr:`input`.

Equivalent to :func:`~torch.fft.ifftn` but IFFTs only the last two dimensions by default.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the IFFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: last two dimensions.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.ifft2`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the IFFT orthonormal)



        Where ``n = prod(s)`` is the logical IFFT size.

        Calling the forward transform (:func:`~torch.fft.fft2`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ifft2`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> x = torch.rand(10, 10, dtype=torch.complex64)

    >>> ifft2 = torch.fft.ifft2(x)



    The discrete Fourier transform is separable, so :func:`~torch.fft.ifft2`

    here is equivalent to two one-dimensional :func:`~torch.fft.ifft` calls:



    >>> two_iffts = torch.fft.ifft(torch.fft.ifft(x, dim=0), dim=1)

    >>> torch.testing.assert_close(ifft2, two_iffts, check_stride=False)



""".format(**common_args))

fftn = _add_docstr(_fft.fft_fftn, r"""

fftn(input, s=None, dim=None, norm=None, *, out=None) -> Tensor



Computes the N dimensional discrete Fourier transform of :attr:`input`.



Note:

    The Fourier domain representation of any real signal satisfies the

    Hermitian property: ``X[i_1, ..., i_n] = conj(X[-i_1, ..., -i_n])``. This

    function always returns all positive and negative frequency terms even

    though, for real inputs, half of these values are redundant.

    :func:`~torch.fft.rfftn` returns the more compact one-sided representation

    where only the positive frequencies of the last dimension are returned.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: all dimensions, or the last ``len(s)`` dimensions if :attr:`s` is given.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.fftn`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the FFT orthonormal)



        Where ``n = prod(s)`` is the logical FFT size.

        Calling the backward transform (:func:`~torch.fft.ifftn`) with the same

        normalization mode will apply an overall normalization of ``1/n``

        between the two transforms. This is required to make

        :func:`~torch.fft.ifftn` the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    >>> x = torch.rand(10, 10, dtype=torch.complex64)

    >>> fftn = torch.fft.fftn(x)



    The discrete Fourier transform is separable, so :func:`~torch.fft.fftn`

    here is equivalent to two one-dimensional :func:`~torch.fft.fft` calls:



    >>> two_ffts = torch.fft.fft(torch.fft.fft(x, dim=0), dim=1)

    >>> torch.testing.assert_close(fftn, two_ffts, check_stride=False)



""".format(**common_args))

ifftn = _add_docstr(_fft.fft_ifftn, r"""

ifftn(input, s=None, dim=None, norm=None, *, out=None) -> Tensor



Computes the N dimensional inverse discrete Fourier transform of :attr:`input`.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the IFFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: all dimensions, or the last ``len(s)`` dimensions if :attr:`s` is given.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.ifftn`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the IFFT orthonormal)



        Where ``n = prod(s)`` is the logical IFFT size.

        Calling the forward transform (:func:`~torch.fft.fftn`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ifftn`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> x = torch.rand(10, 10, dtype=torch.complex64)

    >>> ifftn = torch.fft.ifftn(x)



    The discrete Fourier transform is separable, so :func:`~torch.fft.ifftn`

    here is equivalent to two one-dimensional :func:`~torch.fft.ifft` calls:



    >>> two_iffts = torch.fft.ifft(torch.fft.ifft(x, dim=0), dim=1)

    >>> torch.testing.assert_close(ifftn, two_iffts, check_stride=False)



""".format(**common_args))

rfft = _add_docstr(_fft.fft_rfft, r"""

rfft(input, n=None, dim=-1, norm=None, *, out=None) -> Tensor



Computes the one dimensional Fourier transform of real-valued :attr:`input`.



The FFT of a real signal is Hermitian-symmetric, ``X[i] = conj(X[-i])`` so

the output contains only the positive frequencies below the Nyquist frequency.

To compute the full output, use :func:`~torch.fft.fft`



Note:

    Supports torch.half on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimension.



Args:

    input (Tensor): the real input tensor

    n (int, optional): Signal length. If given, the input will either be zero-padded

        or trimmed to this length before computing the real FFT.

    dim (int, optional): The dimension along which to take the one dimensional real FFT.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.rfft`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the FFT orthonormal)



        Calling the backward transform (:func:`~torch.fft.irfft`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.irfft`

        the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    >>> t = torch.arange(4)

    >>> t

    tensor([0, 1, 2, 3])

    >>> torch.fft.rfft(t)

    tensor([ 6.+0.j, -2.+2.j, -2.+0.j])



    Compare against the full output from :func:`~torch.fft.fft`:



    >>> torch.fft.fft(t)

    tensor([ 6.+0.j, -2.+2.j, -2.+0.j, -2.-2.j])



    Notice that the symmetric element ``T[-1] == T[1].conj()`` is omitted.

    At the Nyquist frequency ``T[-2] == T[2]`` is it's own symmetric pair,

    and therefore must always be real-valued.

""".format(**common_args))

irfft = _add_docstr(_fft.fft_irfft, r"""

irfft(input, n=None, dim=-1, norm=None, *, out=None) -> Tensor



Computes the inverse of :func:`~torch.fft.rfft`.



:attr:`input` is interpreted as a one-sided Hermitian signal in the Fourier

domain, as produced by :func:`~torch.fft.rfft`. By the Hermitian property, the

output will be real-valued.



Note:

    Some input frequencies must be real-valued to satisfy the Hermitian

    property. In these cases the imaginary component will be ignored.

    For example, any imaginary component in the zero-frequency term cannot

    be represented in a real output and so will always be ignored.



Note:

    The correct interpretation of the Hermitian input depends on the length of

    the original data, as given by :attr:`n`. This is because each input shape

    could correspond to either an odd or even length signal. By default, the

    signal is assumed to be even length and odd signals will not round-trip

    properly. So, it is recommended to always pass the signal length :attr:`n`.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimension.

    With default arguments, size of the transformed dimension should be (2^n + 1) as argument

    `n` defaults to even output size = 2 * (transformed_dim_size - 1)



Args:

    input (Tensor): the input tensor representing a half-Hermitian signal

    n (int, optional): Output signal length. This determines the length of the

        output signal. If given, the input will either be zero-padded or trimmed to this

        length before computing the real IFFT.

        Defaults to even output: ``n=2*(input.size(dim) - 1)``.

    dim (int, optional): The dimension along which to take the one dimensional real IFFT.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.irfft`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the real IFFT orthonormal)



        Calling the forward transform (:func:`~torch.fft.rfft`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.irfft`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> t = torch.linspace(0, 1, 5)

    >>> t

    tensor([0.0000, 0.2500, 0.5000, 0.7500, 1.0000])

    >>> T = torch.fft.rfft(t)

    >>> T

    tensor([ 2.5000+0.0000j, -0.6250+0.8602j, -0.6250+0.2031j])



    Without specifying the output length to :func:`~torch.fft.irfft`, the output

    will not round-trip properly because the input is odd-length:



    >>> torch.fft.irfft(T)

    tensor([0.1562, 0.3511, 0.7812, 1.2114])



    So, it is recommended to always pass the signal length :attr:`n`:



    >>> roundtrip = torch.fft.irfft(T, t.numel())

    >>> torch.testing.assert_close(roundtrip, t, check_stride=False)



""".format(**common_args))

rfft2 = _add_docstr(_fft.fft_rfft2, r"""

rfft2(input, s=None, dim=(-2, -1), norm=None, *, out=None) -> Tensor



Computes the 2-dimensional discrete Fourier transform of real :attr:`input`.

Equivalent to :func:`~torch.fft.rfftn` but FFTs only the last two dimensions by default.



The FFT of a real signal is Hermitian-symmetric, ``X[i, j] = conj(X[-i, -j])``,

so the full :func:`~torch.fft.fft2` output contains redundant information.

:func:`~torch.fft.rfft2` instead omits the negative frequencies in the last

dimension.



Note:

    Supports torch.half on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the real FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: last two dimensions.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.rfft2`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the real FFT orthonormal)



        Where ``n = prod(s)`` is the logical FFT size.

        Calling the backward transform (:func:`~torch.fft.irfft2`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.irfft2`

        the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    >>> t = torch.rand(10, 10)

    >>> rfft2 = torch.fft.rfft2(t)

    >>> rfft2.size()

    torch.Size([10, 6])



    Compared against the full output from :func:`~torch.fft.fft2`, we have all

    elements up to the Nyquist frequency.



    >>> fft2 = torch.fft.fft2(t)

    >>> torch.testing.assert_close(fft2[..., :6], rfft2, check_stride=False)



    The discrete Fourier transform is separable, so :func:`~torch.fft.rfft2`

    here is equivalent to a combination of :func:`~torch.fft.fft` and

    :func:`~torch.fft.rfft`:



    >>> two_ffts = torch.fft.fft(torch.fft.rfft(t, dim=1), dim=0)

    >>> torch.testing.assert_close(rfft2, two_ffts, check_stride=False)



""".format(**common_args))

irfft2 = _add_docstr(_fft.fft_irfft2, r"""

irfft2(input, s=None, dim=(-2, -1), norm=None, *, out=None) -> Tensor



Computes the inverse of :func:`~torch.fft.rfft2`.

Equivalent to :func:`~torch.fft.irfftn` but IFFTs only the last two dimensions by default.



:attr:`input` is interpreted as a one-sided Hermitian signal in the Fourier

domain, as produced by :func:`~torch.fft.rfft2`. By the Hermitian property, the

output will be real-valued.



Note:

    Some input frequencies must be real-valued to satisfy the Hermitian

    property. In these cases the imaginary component will be ignored.

    For example, any imaginary component in the zero-frequency term cannot

    be represented in a real output and so will always be ignored.



Note:

    The correct interpretation of the Hermitian input depends on the length of

    the original data, as given by :attr:`s`. This is because each input shape

    could correspond to either an odd or even length signal. By default, the

    signal is assumed to be even length and odd signals will not round-trip

    properly. So, it is recommended to always pass the signal shape :attr:`s`.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.

    With default arguments, the size of last dimension should be (2^n + 1) as argument

    `s` defaults to even output size = 2 * (last_dim_size - 1)



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the real FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Defaults to even output in the last dimension:

        ``s[-1] = 2*(input.size(dim[-1]) - 1)``.

    dim (Tuple[int], optional): Dimensions to be transformed.

        The last dimension must be the half-Hermitian compressed dimension.

        Default: last two dimensions.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.irfft2`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the real IFFT orthonormal)



        Where ``n = prod(s)`` is the logical IFFT size.

        Calling the forward transform (:func:`~torch.fft.rfft2`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.irfft2`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> t = torch.rand(10, 9)

    >>> T = torch.fft.rfft2(t)



    Without specifying the output length to :func:`~torch.fft.irfft2`, the output

    will not round-trip properly because the input is odd-length in the last

    dimension:



    >>> torch.fft.irfft2(T).size()

    torch.Size([10, 8])



    So, it is recommended to always pass the signal shape :attr:`s`.



    >>> roundtrip = torch.fft.irfft2(T, t.size())

    >>> roundtrip.size()

    torch.Size([10, 9])

    >>> torch.testing.assert_close(roundtrip, t, check_stride=False)



""".format(**common_args))

rfftn = _add_docstr(_fft.fft_rfftn, r"""

rfftn(input, s=None, dim=None, norm=None, *, out=None) -> Tensor



Computes the N-dimensional discrete Fourier transform of real :attr:`input`.



The FFT of a real signal is Hermitian-symmetric,

``X[i_1, ..., i_n] = conj(X[-i_1, ..., -i_n])`` so the full

:func:`~torch.fft.fftn` output contains redundant information.

:func:`~torch.fft.rfftn` instead omits the negative frequencies in the

last dimension.



Note:

    Supports torch.half on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the real FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: all dimensions, or the last ``len(s)`` dimensions if :attr:`s` is given.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.rfftn`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the real FFT orthonormal)



        Where ``n = prod(s)`` is the logical FFT size.

        Calling the backward transform (:func:`~torch.fft.irfftn`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.irfftn`

        the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    >>> t = torch.rand(10, 10)

    >>> rfftn = torch.fft.rfftn(t)

    >>> rfftn.size()

    torch.Size([10, 6])



    Compared against the full output from :func:`~torch.fft.fftn`, we have all

    elements up to the Nyquist frequency.



    >>> fftn = torch.fft.fftn(t)

    >>> torch.testing.assert_close(fftn[..., :6], rfftn, check_stride=False)



    The discrete Fourier transform is separable, so :func:`~torch.fft.rfftn`

    here is equivalent to a combination of :func:`~torch.fft.fft` and

    :func:`~torch.fft.rfft`:



    >>> two_ffts = torch.fft.fft(torch.fft.rfft(t, dim=1), dim=0)

    >>> torch.testing.assert_close(rfftn, two_ffts, check_stride=False)



""".format(**common_args))

irfftn = _add_docstr(_fft.fft_irfftn, r"""

irfftn(input, s=None, dim=None, norm=None, *, out=None) -> Tensor



Computes the inverse of :func:`~torch.fft.rfftn`.



:attr:`input` is interpreted as a one-sided Hermitian signal in the Fourier

domain, as produced by :func:`~torch.fft.rfftn`. By the Hermitian property, the

output will be real-valued.



Note:

    Some input frequencies must be real-valued to satisfy the Hermitian

    property. In these cases the imaginary component will be ignored.

    For example, any imaginary component in the zero-frequency term cannot

    be represented in a real output and so will always be ignored.



Note:

    The correct interpretation of the Hermitian input depends on the length of

    the original data, as given by :attr:`s`. This is because each input shape

    could correspond to either an odd or even length signal. By default, the

    signal is assumed to be even length and odd signals will not round-trip

    properly. So, it is recommended to always pass the signal shape :attr:`s`.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.

    With default arguments, the size of last dimension should be (2^n + 1) as argument

    `s` defaults to even output size = 2 * (last_dim_size - 1)



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the real FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Defaults to even output in the last dimension:

        ``s[-1] = 2*(input.size(dim[-1]) - 1)``.

    dim (Tuple[int], optional): Dimensions to be transformed.

        The last dimension must be the half-Hermitian compressed dimension.

        Default: all dimensions, or the last ``len(s)`` dimensions if :attr:`s` is given.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.irfftn`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the real IFFT orthonormal)



        Where ``n = prod(s)`` is the logical IFFT size.

        Calling the forward transform (:func:`~torch.fft.rfftn`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.irfftn`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> t = torch.rand(10, 9)

    >>> T = torch.fft.rfftn(t)



    Without specifying the output length to :func:`~torch.fft.irfft`, the output

    will not round-trip properly because the input is odd-length in the last

    dimension:



    >>> torch.fft.irfftn(T).size()

    torch.Size([10, 8])



    So, it is recommended to always pass the signal shape :attr:`s`.



    >>> roundtrip = torch.fft.irfftn(T, t.size())

    >>> roundtrip.size()

    torch.Size([10, 9])

    >>> torch.testing.assert_close(roundtrip, t, check_stride=False)



""".format(**common_args))

hfft = _add_docstr(_fft.fft_hfft, r"""

hfft(input, n=None, dim=-1, norm=None, *, out=None) -> Tensor



Computes the one dimensional discrete Fourier transform of a Hermitian

symmetric :attr:`input` signal.



Note:



    :func:`~torch.fft.hfft`/:func:`~torch.fft.ihfft` are analogous to

    :func:`~torch.fft.rfft`/:func:`~torch.fft.irfft`. The real FFT expects

    a real signal in the time-domain and gives a Hermitian symmetry in the

    frequency-domain. The Hermitian FFT is the opposite; Hermitian symmetric in

    the time-domain and real-valued in the frequency-domain. For this reason,

    special care needs to be taken with the length argument :attr:`n`, in the

    same way as with :func:`~torch.fft.irfft`.



Note:

    Because the signal is Hermitian in the time-domain, the result will be

    real in the frequency domain. Note that some input frequencies must be

    real-valued to satisfy the Hermitian property. In these cases the imaginary

    component will be ignored. For example, any imaginary component in

    ``input[0]`` would result in one or more complex frequency terms which

    cannot be represented in a real output and so will always be ignored.



Note:

    The correct interpretation of the Hermitian input depends on the length of

    the original data, as given by :attr:`n`. This is because each input shape

    could correspond to either an odd or even length signal. By default, the

    signal is assumed to be even length and odd signals will not round-trip

    properly. So, it is recommended to always pass the signal length :attr:`n`.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimension.

    With default arguments, size of the transformed dimension should be (2^n + 1) as argument

    `n` defaults to even output size = 2 * (transformed_dim_size - 1)



Args:

    input (Tensor): the input tensor representing a half-Hermitian signal

    n (int, optional): Output signal length. This determines the length of the

        real output. If given, the input will either be zero-padded or trimmed to this

        length before computing the Hermitian FFT.

        Defaults to even output: ``n=2*(input.size(dim) - 1)``.

    dim (int, optional): The dimension along which to take the one dimensional Hermitian FFT.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.hfft`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the Hermitian FFT orthonormal)



        Calling the backward transform (:func:`~torch.fft.ihfft`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ihfft`

        the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    Taking a real-valued frequency signal and bringing it into the time domain

    gives Hermitian symmetric output:



    >>> t = torch.linspace(0, 1, 5)

    >>> t

    tensor([0.0000, 0.2500, 0.5000, 0.7500, 1.0000])

    >>> T = torch.fft.ifft(t)

    >>> T

    tensor([ 0.5000-0.0000j, -0.1250-0.1720j, -0.1250-0.0406j, -0.1250+0.0406j,

            -0.1250+0.1720j])



    Note that ``T[1] == T[-1].conj()`` and ``T[2] == T[-2].conj()`` is

    redundant. We can thus compute the forward transform without considering

    negative frequencies:



    >>> torch.fft.hfft(T[:3], n=5)

    tensor([0.0000, 0.2500, 0.5000, 0.7500, 1.0000])



    Like with :func:`~torch.fft.irfft`, the output length must be given in order

    to recover an even length output:



    >>> torch.fft.hfft(T[:3])

    tensor([0.1250, 0.2809, 0.6250, 0.9691])

""".format(**common_args))

ihfft = _add_docstr(_fft.fft_ihfft, r"""

ihfft(input, n=None, dim=-1, norm=None, *, out=None) -> Tensor



Computes the inverse of :func:`~torch.fft.hfft`.



:attr:`input` must be a real-valued signal, interpreted in the Fourier domain.

The IFFT of a real signal is Hermitian-symmetric, ``X[i] = conj(X[-i])``.

:func:`~torch.fft.ihfft` represents this in the one-sided form where only the

positive frequencies below the Nyquist frequency are included. To compute the

full output, use :func:`~torch.fft.ifft`.



Note:

    Supports torch.half on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimension.



Args:

    input (Tensor): the real input tensor

    n (int, optional): Signal length. If given, the input will either be zero-padded

        or trimmed to this length before computing the Hermitian IFFT.

    dim (int, optional): The dimension along which to take the one dimensional Hermitian IFFT.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.ihfft`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the IFFT orthonormal)



        Calling the forward transform (:func:`~torch.fft.hfft`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ihfft`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> t = torch.arange(5)

    >>> t

    tensor([0, 1, 2, 3, 4])

    >>> torch.fft.ihfft(t)

    tensor([ 2.0000-0.0000j, -0.5000-0.6882j, -0.5000-0.1625j])



    Compare against the full output from :func:`~torch.fft.ifft`:



    >>> torch.fft.ifft(t)

    tensor([ 2.0000-0.0000j, -0.5000-0.6882j, -0.5000-0.1625j, -0.5000+0.1625j,

            -0.5000+0.6882j])

""".format(**common_args))

hfft2 = _add_docstr(_fft.fft_hfft2, r"""

hfft2(input, s=None, dim=(-2, -1), norm=None, *, out=None) -> Tensor



Computes the 2-dimensional discrete Fourier transform of a Hermitian symmetric

:attr:`input` signal. Equivalent to :func:`~torch.fft.hfftn` but only

transforms the last two dimensions by default.



:attr:`input` is interpreted as a one-sided Hermitian signal in the time

domain. By the Hermitian property, the Fourier transform will be real-valued.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.

    With default arguments, the size of last dimension should be (2^n + 1) as argument

    `s` defaults to even output size = 2 * (last_dim_size - 1)



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the Hermitian FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Defaults to even output in the last dimension:

        ``s[-1] = 2*(input.size(dim[-1]) - 1)``.

    dim (Tuple[int], optional): Dimensions to be transformed.

        The last dimension must be the half-Hermitian compressed dimension.

        Default: last two dimensions.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.hfft2`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the Hermitian FFT orthonormal)



        Where ``n = prod(s)`` is the logical FFT size.

        Calling the backward transform (:func:`~torch.fft.ihfft2`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ihfft2`

        the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    Starting from a real frequency-space signal, we can generate a

    Hermitian-symmetric time-domain signal:

    >>> T = torch.rand(10, 9)

    >>> t = torch.fft.ihfft2(T)



    Without specifying the output length to :func:`~torch.fft.hfftn`, the

    output will not round-trip properly because the input is odd-length in the

    last dimension:



    >>> torch.fft.hfft2(t).size()

    torch.Size([10, 10])



    So, it is recommended to always pass the signal shape :attr:`s`.



    >>> roundtrip = torch.fft.hfft2(t, T.size())

    >>> roundtrip.size()

    torch.Size([10, 9])

    >>> torch.allclose(roundtrip, T)

    True



""".format(**common_args))

ihfft2 = _add_docstr(_fft.fft_ihfft2, r"""

ihfft2(input, s=None, dim=(-2, -1), norm=None, *, out=None) -> Tensor



Computes the 2-dimensional inverse discrete Fourier transform of real

:attr:`input`. Equivalent to :func:`~torch.fft.ihfftn` but transforms only the

two last dimensions by default.



Note:

    Supports torch.half on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the Hermitian IFFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: last two dimensions.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.ihfft2`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the Hermitian IFFT orthonormal)



        Where ``n = prod(s)`` is the logical IFFT size.

        Calling the forward transform (:func:`~torch.fft.hfft2`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ihfft2`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> T = torch.rand(10, 10)

    >>> t = torch.fft.ihfft2(t)

    >>> t.size()

    torch.Size([10, 6])



    Compared against the full output from :func:`~torch.fft.ifft2`, the

    Hermitian time-space signal takes up only half the space.



    >>> fftn = torch.fft.ifft2(t)

    >>> torch.allclose(fftn[..., :6], rfftn)

    True



    The discrete Fourier transform is separable, so :func:`~torch.fft.ihfft2`

    here is equivalent to a combination of :func:`~torch.fft.ifft` and

    :func:`~torch.fft.ihfft`:



    >>> two_ffts = torch.fft.ifft(torch.fft.ihfft(t, dim=1), dim=0)

    >>> torch.allclose(t, two_ffts)

    True



""".format(**common_args))

hfftn = _add_docstr(_fft.fft_hfftn, r"""

hfftn(input, s=None, dim=None, norm=None, *, out=None) -> Tensor



Computes the n-dimensional discrete Fourier transform of a Hermitian symmetric

:attr:`input` signal.



:attr:`input` is interpreted as a one-sided Hermitian signal in the time

domain. By the Hermitian property, the Fourier transform will be real-valued.



Note:

    :func:`~torch.fft.hfftn`/:func:`~torch.fft.ihfftn` are analogous to

    :func:`~torch.fft.rfftn`/:func:`~torch.fft.irfftn`. The real FFT expects

    a real signal in the time-domain and gives Hermitian symmetry in the

    frequency-domain. The Hermitian FFT is the opposite; Hermitian symmetric in

    the time-domain and real-valued in the frequency-domain. For this reason,

    special care needs to be taken with the shape argument :attr:`s`, in the

    same way as with :func:`~torch.fft.irfftn`.



Note:

    Some input frequencies must be real-valued to satisfy the Hermitian

    property. In these cases the imaginary component will be ignored.

    For example, any imaginary component in the zero-frequency term cannot

    be represented in a real output and so will always be ignored.



Note:

    The correct interpretation of the Hermitian input depends on the length of

    the original data, as given by :attr:`s`. This is because each input shape

    could correspond to either an odd or even length signal. By default, the

    signal is assumed to be even length and odd signals will not round-trip

    properly. It is recommended to always pass the signal shape :attr:`s`.



Note:

    Supports torch.half and torch.chalf on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.

    With default arguments, the size of last dimension should be (2^n + 1) as argument

    `s` defaults to even output size = 2 * (last_dim_size - 1)



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the real FFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Defaults to even output in the last dimension:

        ``s[-1] = 2*(input.size(dim[-1]) - 1)``.

    dim (Tuple[int], optional): Dimensions to be transformed.

        The last dimension must be the half-Hermitian compressed dimension.

        Default: all dimensions, or the last ``len(s)`` dimensions if :attr:`s` is given.

    norm (str, optional): Normalization mode. For the forward transform

        (:func:`~torch.fft.hfftn`), these correspond to:



        * ``"forward"`` - normalize by ``1/n``

        * ``"backward"`` - no normalization

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the Hermitian FFT orthonormal)



        Where ``n = prod(s)`` is the logical FFT size.

        Calling the backward transform (:func:`~torch.fft.ihfftn`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ihfftn`

        the exact inverse.



        Default is ``"backward"`` (no normalization).



Keyword args:

    {out}



Example:



    Starting from a real frequency-space signal, we can generate a

    Hermitian-symmetric time-domain signal:

    >>> T = torch.rand(10, 9)

    >>> t = torch.fft.ihfftn(T)



    Without specifying the output length to :func:`~torch.fft.hfftn`, the

    output will not round-trip properly because the input is odd-length in the

    last dimension:



    >>> torch.fft.hfftn(t).size()

    torch.Size([10, 10])



    So, it is recommended to always pass the signal shape :attr:`s`.



    >>> roundtrip = torch.fft.hfftn(t, T.size())

    >>> roundtrip.size()

    torch.Size([10, 9])

    >>> torch.allclose(roundtrip, T)

    True



""".format(**common_args))

ihfftn = _add_docstr(_fft.fft_ihfftn, r"""

ihfftn(input, s=None, dim=None, norm=None, *, out=None) -> Tensor



Computes the N-dimensional inverse discrete Fourier transform of real :attr:`input`.



:attr:`input` must be a real-valued signal, interpreted in the Fourier domain.

The n-dimensional IFFT of a real signal is Hermitian-symmetric,

``X[i, j, ...] = conj(X[-i, -j, ...])``. :func:`~torch.fft.ihfftn` represents

this in the one-sided form where only the positive frequencies below the

Nyquist frequency are included in the last signal dimension. To compute the

full output, use :func:`~torch.fft.ifftn`.



Note:

    Supports torch.half on CUDA with GPU Architecture SM53 or greater.

    However it only supports powers of 2 signal length in every transformed dimensions.



Args:

    input (Tensor): the input tensor

    s (Tuple[int], optional): Signal size in the transformed dimensions.

        If given, each dimension ``dim[i]`` will either be zero-padded or

        trimmed to the length ``s[i]`` before computing the Hermitian IFFT.

        If a length ``-1`` is specified, no padding is done in that dimension.

        Default: ``s = [input.size(d) for d in dim]``

    dim (Tuple[int], optional): Dimensions to be transformed.

        Default: all dimensions, or the last ``len(s)`` dimensions if :attr:`s` is given.

    norm (str, optional): Normalization mode. For the backward transform

        (:func:`~torch.fft.ihfftn`), these correspond to:



        * ``"forward"`` - no normalization

        * ``"backward"`` - normalize by ``1/n``

        * ``"ortho"`` - normalize by ``1/sqrt(n)`` (making the Hermitian IFFT orthonormal)



        Where ``n = prod(s)`` is the logical IFFT size.

        Calling the forward transform (:func:`~torch.fft.hfftn`) with the same

        normalization mode will apply an overall normalization of ``1/n`` between

        the two transforms. This is required to make :func:`~torch.fft.ihfftn`

        the exact inverse.



        Default is ``"backward"`` (normalize by ``1/n``).



Keyword args:

    {out}



Example:



    >>> T = torch.rand(10, 10)

    >>> ihfftn = torch.fft.ihfftn(T)

    >>> ihfftn.size()

    torch.Size([10, 6])



    Compared against the full output from :func:`~torch.fft.ifftn`, we have all

    elements up to the Nyquist frequency.



    >>> ifftn = torch.fft.ifftn(t)

    >>> torch.allclose(ifftn[..., :6], ihfftn)

    True



    The discrete Fourier transform is separable, so :func:`~torch.fft.ihfftn`

    here is equivalent to a combination of :func:`~torch.fft.ihfft` and

    :func:`~torch.fft.ifft`:



    >>> two_iffts = torch.fft.ifft(torch.fft.ihfft(t, dim=1), dim=0)

    >>> torch.allclose(ihfftn, two_iffts)

    True



""".format(**common_args))

fftfreq = _add_docstr(_fft.fft_fftfreq, r"""

fftfreq(n, d=1.0, *, out=None, dtype=None, layout=torch.strided, device=None, requires_grad=False) -> Tensor



Computes the discrete Fourier Transform sample frequencies for a signal of size :attr:`n`.



Note:

    By convention, :func:`~torch.fft.fft` returns positive frequency terms

    first, followed by the negative frequencies in reverse order, so that

    ``f[-i]`` for all :math:`0 < i \leq n/2`` in Python gives the negative

    frequency terms. For an FFT of length :attr:`n` and with inputs spaced in

    length unit :attr:`d`, the frequencies are::



        f = [0, 1, ..., (n - 1) // 2, -(n // 2), ..., -1] / (d * n)



Note:

    For even lengths, the Nyquist frequency at ``f[n/2]`` can be thought of as

    either negative or positive. :func:`~torch.fft.fftfreq` follows NumPy's

    convention of taking it to be negative.



Args:

    n (int): the FFT length

    d (float, optional): The sampling length scale.

        The spacing between individual samples of the FFT input.

        The default assumes unit spacing, dividing that result by the actual

        spacing gives the result in physical frequency units.



Keyword Args:

    {out}

    {dtype}

    {layout}

    {device}

    {requires_grad}



Example:



    >>> torch.fft.fftfreq(5)

    tensor([ 0.0000,  0.2000,  0.4000, -0.4000, -0.2000])



    For even input, we can see the Nyquist frequency at ``f[2]`` is given as

    negative:



    >>> torch.fft.fftfreq(4)

    tensor([ 0.0000,  0.2500, -0.5000, -0.2500])



""".format(**factory_common_args))

rfftfreq = _add_docstr(_fft.fft_rfftfreq, r"""

rfftfreq(n, d=1.0, *, out=None, dtype=None, layout=torch.strided, device=None, requires_grad=False) -> Tensor



Computes the sample frequencies for :func:`~torch.fft.rfft` with a signal of size :attr:`n`.



Note:

    :func:`~torch.fft.rfft` returns Hermitian one-sided output, so only the

    positive frequency terms are returned. For a real FFT of length :attr:`n`

    and with inputs spaced in length unit :attr:`d`, the frequencies are::



        f = torch.arange((n + 1) // 2) / (d * n)



Note:

    For even lengths, the Nyquist frequency at ``f[n/2]`` can be thought of as

    either negative or positive. Unlike :func:`~torch.fft.fftfreq`,

    :func:`~torch.fft.rfftfreq` always returns it as positive.



Args:

    n (int): the real FFT length

    d (float, optional): The sampling length scale.

        The spacing between individual samples of the FFT input.

        The default assumes unit spacing, dividing that result by the actual

        spacing gives the result in physical frequency units.



Keyword Args:

    {out}

    {dtype}

    {layout}

    {device}

    {requires_grad}



Example:



    >>> torch.fft.rfftfreq(5)

    tensor([0.0000, 0.2000, 0.4000])



    >>> torch.fft.rfftfreq(4)

    tensor([0.0000, 0.2500, 0.5000])



    Compared to the output from :func:`~torch.fft.fftfreq`, we see that the

    Nyquist frequency at ``f[2]`` has changed sign:

    >>> torch.fft.fftfreq(4)

    tensor([ 0.0000,  0.2500, -0.5000, -0.2500])



""".format(**factory_common_args))

fftshift = _add_docstr(_fft.fft_fftshift, r"""

fftshift(input, dim=None) -> Tensor



Reorders n-dimensional FFT data, as provided by :func:`~torch.fft.fftn`, to have

negative frequency terms first.



This performs a periodic shift of n-dimensional data such that the origin

``(0, ..., 0)`` is moved to the center of the tensor. Specifically, to

``input.shape[dim] // 2`` in each selected dimension.



Note:

    By convention, the FFT returns positive frequency terms first, followed by

    the negative frequencies in reverse order, so that ``f[-i]`` for all

    :math:`0 < i \leq n/2` in Python gives the negative frequency terms.

    :func:`~torch.fft.fftshift` rearranges all frequencies into ascending order

    from negative to positive with the zero-frequency term in the center.



Note:

    For even lengths, the Nyquist frequency at ``f[n/2]`` can be thought of as

    either negative or positive. :func:`~torch.fft.fftshift` always puts the

    Nyquist term at the 0-index. This is the same convention used by

    :func:`~torch.fft.fftfreq`.



Args:

    input (Tensor): the tensor in FFT order

    dim (int, Tuple[int], optional): The dimensions to rearrange.

        Only dimensions specified here will be rearranged, any other dimensions

        will be left in their original order.

        Default: All dimensions of :attr:`input`.



Example:



    >>> f = torch.fft.fftfreq(4)

    >>> f

    tensor([ 0.0000,  0.2500, -0.5000, -0.2500])



    >>> torch.fft.fftshift(f)

    tensor([-0.5000, -0.2500,  0.0000,  0.2500])



    Also notice that the Nyquist frequency term at ``f[2]`` was moved to the

    beginning of the tensor.



    This also works for multi-dimensional transforms:



    >>> x = torch.fft.fftfreq(5, d=1/5) + 0.1 * torch.fft.fftfreq(5, d=1/5).unsqueeze(1)

    >>> x

    tensor([[ 0.0000,  1.0000,  2.0000, -2.0000, -1.0000],

            [ 0.1000,  1.1000,  2.1000, -1.9000, -0.9000],

            [ 0.2000,  1.2000,  2.2000, -1.8000, -0.8000],

            [-0.2000,  0.8000,  1.8000, -2.2000, -1.2000],

            [-0.1000,  0.9000,  1.9000, -2.1000, -1.1000]])



    >>> torch.fft.fftshift(x)

    tensor([[-2.2000, -1.2000, -0.2000,  0.8000,  1.8000],

            [-2.1000, -1.1000, -0.1000,  0.9000,  1.9000],

            [-2.0000, -1.0000,  0.0000,  1.0000,  2.0000],

            [-1.9000, -0.9000,  0.1000,  1.1000,  2.1000],

            [-1.8000, -0.8000,  0.2000,  1.2000,  2.2000]])



    :func:`~torch.fft.fftshift` can also be useful for spatial data. If our

    data is defined on a centered grid (``[-(N//2), (N-1)//2]``) then we can

    use the standard FFT defined on an uncentered grid (``[0, N)``) by first

    applying an :func:`~torch.fft.ifftshift`.



    >>> x_centered = torch.arange(-5, 5)

    >>> x_uncentered = torch.fft.ifftshift(x_centered)

    >>> fft_uncentered = torch.fft.fft(x_uncentered)



    Similarly, we can convert the frequency domain components to centered

    convention by applying :func:`~torch.fft.fftshift`.



    >>> fft_centered = torch.fft.fftshift(fft_uncentered)



    The inverse transform, from centered Fourier space back to centered spatial

    data, can be performed by applying the inverse shifts in reverse order:



    >>> x_centered_2 = torch.fft.fftshift(torch.fft.ifft(torch.fft.ifftshift(fft_centered)))

    >>> torch.testing.assert_close(x_centered.to(torch.complex64), x_centered_2, check_stride=False)





""")

ifftshift = _add_docstr(_fft.fft_ifftshift, r"""

ifftshift(input, dim=None) -> Tensor



Inverse of :func:`~torch.fft.fftshift`.



Args:

    input (Tensor): the tensor in FFT order

    dim (int, Tuple[int], optional): The dimensions to rearrange.

        Only dimensions specified here will be rearranged, any other dimensions

        will be left in their original order.

        Default: All dimensions of :attr:`input`.



Example:



    >>> f = torch.fft.fftfreq(5)

    >>> f

    tensor([ 0.0000,  0.2000,  0.4000, -0.4000, -0.2000])



    A round-trip through :func:`~torch.fft.fftshift` and

    :func:`~torch.fft.ifftshift` gives the same result:



    >>> shifted = torch.fft.fftshift(f)

    >>> torch.fft.ifftshift(shifted)

    tensor([ 0.0000,  0.2000,  0.4000, -0.4000, -0.2000])



""")