state
stringlengths 0
159k
| srcUpToTactic
stringlengths 387
167k
| nextTactic
stringlengths 3
9k
| declUpToTactic
stringlengths 22
11.5k
| declId
stringlengths 38
95
| decl
stringlengths 16
1.89k
| file_tag
stringlengths 17
73
|
|---|---|---|---|---|---|---|
case succ
S : Type u
inst✝ : Semiring S
a b : ℕ
⊢ eval a (ascPochhammer ℕ (Nat.succ b)) = Nat.descFactorial (a + Nat.succ b - 1) (Nat.succ b) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
| rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero] | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
| Mathlib.RingTheory.Polynomial.Pochhammer.162_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
S : Type u
inst✝ : Semiring S
a b : ℕ
⊢ eval a (ascPochhammer ℕ (Nat.succ b)) = Nat.descFactorial (a + b) (Nat.succ b) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
| cases a | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
| Mathlib.RingTheory.Polynomial.Pochhammer.162_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ.zero
S : Type u
inst✝ : Semiring S
b : ℕ
⊢ eval Nat.zero (ascPochhammer ℕ (Nat.succ b)) = Nat.descFactorial (Nat.zero + b) (Nat.succ b) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· | simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul] | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· | Mathlib.RingTheory.Polynomial.Pochhammer.162_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ.succ
S : Type u
inst✝ : Semiring S
b n✝ : ℕ
⊢ eval (Nat.succ n✝) (ascPochhammer ℕ (Nat.succ b)) = Nat.descFactorial (Nat.succ n✝ + b) (Nat.succ b) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· | rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial] | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· | Mathlib.RingTheory.Polynomial.Pochhammer.162_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b | Mathlib_RingTheory_Polynomial_Pochhammer |
S : Type u
inst✝² : Semiring S
n : ℕ
inst✝¹ : NoZeroDivisors S
inst✝ : Nontrivial S
⊢ natDegree (ascPochhammer S n) = n | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
| induction' n with n hn | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
| Mathlib.RingTheory.Polynomial.Pochhammer.174_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
S : Type u
inst✝² : Semiring S
inst✝¹ : NoZeroDivisors S
inst✝ : Nontrivial S
⊢ natDegree (ascPochhammer S Nat.zero) = Nat.zero | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· | simp | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· | Mathlib.RingTheory.Polynomial.Pochhammer.174_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
S : Type u
inst✝² : Semiring S
inst✝¹ : NoZeroDivisors S
inst✝ : Nontrivial S
n : ℕ
hn : natDegree (ascPochhammer S n) = n
⊢ natDegree (ascPochhammer S (Nat.succ n)) = Nat.succ n | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· | have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S) | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.174_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
S : Type u
inst✝² : Semiring S
inst✝¹ : NoZeroDivisors S
inst✝ : Nontrivial S
n : ℕ
hn : natDegree (ascPochhammer S n) = n
this : natDegree (X + ↑n) = 1
⊢ natDegree (ascPochhammer S (Nat.succ n)) = Nat.succ n | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
| rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this] | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
| Mathlib.RingTheory.Polynomial.Pochhammer.174_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
S : Type u
inst✝² : Semiring S
inst✝¹ : NoZeroDivisors S
inst✝ : Nontrivial S
n : ℕ
hn : natDegree (ascPochhammer S n) = n
this : natDegree (X + ↑n) = 1
⊢ ascPochhammer S n ≠ 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
| cases n | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
| Mathlib.RingTheory.Polynomial.Pochhammer.174_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
S : Type u
inst✝² : Semiring S
inst✝¹ : NoZeroDivisors S
inst✝ : Nontrivial S
hn : natDegree (ascPochhammer S Nat.zero) = Nat.zero
this : natDegree (X + ↑Nat.zero) = 1
⊢ ascPochhammer S Nat.zero ≠ 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· | simp | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· | Mathlib.RingTheory.Polynomial.Pochhammer.174_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
S : Type u
inst✝² : Semiring S
inst✝¹ : NoZeroDivisors S
inst✝ : Nontrivial S
n✝ : ℕ
hn : natDegree (ascPochhammer S (Nat.succ n✝)) = Nat.succ n✝
this : natDegree (X + ↑(Nat.succ n✝)) = 1
⊢ ascPochhammer S (Nat.succ n✝) ≠ 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· | refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.174_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
S : Type u_1
inst✝ : StrictOrderedSemiring S
n : ℕ
s : S
h : 0 < s
⊢ 0 < eval s (ascPochhammer S n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
| induction' n with n ih | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
| Mathlib.RingTheory.Polynomial.Pochhammer.192_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
S : Type u_1
inst✝ : StrictOrderedSemiring S
s : S
h : 0 < s
⊢ 0 < eval s (ascPochhammer S Nat.zero) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· | simp only [Nat.zero_eq, ascPochhammer_zero, eval_one] | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· | Mathlib.RingTheory.Polynomial.Pochhammer.192_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
S : Type u_1
inst✝ : StrictOrderedSemiring S
s : S
h : 0 < s
⊢ 0 < 1 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
| exact zero_lt_one | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
| Mathlib.RingTheory.Polynomial.Pochhammer.192_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
S : Type u_1
inst✝ : StrictOrderedSemiring S
s : S
h : 0 < s
n : ℕ
ih : 0 < eval s (ascPochhammer S n)
⊢ 0 < eval s (ascPochhammer S (Nat.succ n)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· | rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add] | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· | Mathlib.RingTheory.Polynomial.Pochhammer.192_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
S : Type u_1
inst✝ : StrictOrderedSemiring S
s : S
h : 0 < s
n : ℕ
ih : 0 < eval s (ascPochhammer S n)
⊢ 0 < eval s (ascPochhammer S n) * (s + ↑n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
| exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n))) | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
| Mathlib.RingTheory.Polynomial.Pochhammer.192_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s | Mathlib_RingTheory_Polynomial_Pochhammer |
S✝ : Type u_1
inst✝¹ : Semiring S✝
r n✝ : ℕ
S : Type u_2
inst✝ : Semiring S
n : ℕ
⊢ eval 1 (ascPochhammer S n) = ↑n ! | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
| rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial] | @[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
| Mathlib.RingTheory.Polynomial.Pochhammer.209_0.yf6mY7NVFIgfXWQ | @[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) | Mathlib_RingTheory_Polynomial_Pochhammer |
S✝ : Type u_1
inst✝¹ : Semiring S✝
r✝ n✝ : ℕ
S : Type u_2
inst✝ : Semiring S
r n : ℕ
⊢ ↑r ! * eval (↑r + 1) (ascPochhammer S n) = ↑(r + n)! | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
| rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial] | theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
| Mathlib.RingTheory.Polynomial.Pochhammer.215_0.yf6mY7NVFIgfXWQ | theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! | Mathlib_RingTheory_Polynomial_Pochhammer |
S : Type u_1
inst✝ : Semiring S
r✝ n r : ℕ
⊢ 0 * eval (0 + 1) (ascPochhammer ℕ r) = (0 + r) * eval 0 (ascPochhammer ℕ r) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
| by_cases h : r = 0 | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
| Mathlib.RingTheory.Polynomial.Pochhammer.220_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm] | Mathlib_RingTheory_Polynomial_Pochhammer |
case pos
S : Type u_1
inst✝ : Semiring S
r✝ n r : ℕ
h : r = 0
⊢ 0 * eval (0 + 1) (ascPochhammer ℕ r) = (0 + r) * eval 0 (ascPochhammer ℕ r) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· | simp only [h, zero_mul, zero_add] | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· | Mathlib.RingTheory.Polynomial.Pochhammer.220_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm] | Mathlib_RingTheory_Polynomial_Pochhammer |
case neg
S : Type u_1
inst✝ : Semiring S
r✝ n r : ℕ
h : ¬r = 0
⊢ 0 * eval (0 + 1) (ascPochhammer ℕ r) = (0 + r) * eval 0 (ascPochhammer ℕ r) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· | simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero] | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· | Mathlib.RingTheory.Polynomial.Pochhammer.220_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm] | Mathlib_RingTheory_Polynomial_Pochhammer |
S : Type u_1
inst✝ : Semiring S
r✝ n r k : ℕ
⊢ (k + 1) * eval (k + 1 + 1) (ascPochhammer ℕ r) = (k + 1 + r) * eval (k + 1) (ascPochhammer ℕ r) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by | simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm] | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by | Mathlib.RingTheory.Polynomial.Pochhammer.220_0.yf6mY7NVFIgfXWQ | theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm] | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
⊢ descPochhammer R 1 = X | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by | simp [descPochhammer] | @[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by | Mathlib.RingTheory.Polynomial.Pochhammer.252_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_one : descPochhammer R 1 = X | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ descPochhammer R (n + 1) = X * comp (descPochhammer R n) (X - 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by | rw [descPochhammer] | theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by | Mathlib.RingTheory.Polynomial.Pochhammer.255_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝² : Ring R
n : ℕ
inst✝¹ : Nontrivial R
inst✝ : NoZeroDivisors R
⊢ Monic (descPochhammer R n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
| induction' n with n hn | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
| Mathlib.RingTheory.Polynomial.Pochhammer.259_0.yf6mY7NVFIgfXWQ | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
R : Type u
inst✝² : Ring R
inst✝¹ : Nontrivial R
inst✝ : NoZeroDivisors R
⊢ Monic (descPochhammer R Nat.zero) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· | simp | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· | Mathlib.RingTheory.Polynomial.Pochhammer.259_0.yf6mY7NVFIgfXWQ | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝² : Ring R
inst✝¹ : Nontrivial R
inst✝ : NoZeroDivisors R
n : ℕ
hn : Monic (descPochhammer R n)
⊢ Monic (descPochhammer R (Nat.succ n)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· | have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1 | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.259_0.yf6mY7NVFIgfXWQ | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝² : Ring R
inst✝¹ : Nontrivial R
inst✝ : NoZeroDivisors R
n : ℕ
hn : Monic (descPochhammer R n)
h : leadingCoeff (X - 1) = 1
⊢ Monic (descPochhammer R (Nat.succ n)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
| have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R) | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
| Mathlib.RingTheory.Polynomial.Pochhammer.259_0.yf6mY7NVFIgfXWQ | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝² : Ring R
inst✝¹ : Nontrivial R
inst✝ : NoZeroDivisors R
n : ℕ
hn : Monic (descPochhammer R n)
h : leadingCoeff (X - 1) = 1
this : natDegree (X - 1) ≠ 0
⊢ Monic (descPochhammer R (Nat.succ n)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
| rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow] | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
| Mathlib.RingTheory.Polynomial.Pochhammer.259_0.yf6mY7NVFIgfXWQ | theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝¹ : Ring R
T : Type v
inst✝ : Ring T
f : R →+* T
n : ℕ
⊢ map f (descPochhammer R n) = descPochhammer T n | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
| induction' n with n ih | @[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
| Mathlib.RingTheory.Polynomial.Pochhammer.272_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
R : Type u
inst✝¹ : Ring R
T : Type v
inst✝ : Ring T
f : R →+* T
⊢ map f (descPochhammer R Nat.zero) = descPochhammer T Nat.zero | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· | simp | @[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· | Mathlib.RingTheory.Polynomial.Pochhammer.272_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝¹ : Ring R
T : Type v
inst✝ : Ring T
f : R →+* T
n : ℕ
ih : map f (descPochhammer R n) = descPochhammer T n
⊢ map f (descPochhammer R (Nat.succ n)) = descPochhammer T (Nat.succ n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· | simp [ih, descPochhammer_succ_left, map_comp] | @[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.272_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
k : ℤ
⊢ ↑(eval k (descPochhammer ℤ n)) = eval (↑k) (descPochhammer R n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
| rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)] | @[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
| Mathlib.RingTheory.Polynomial.Pochhammer.280_0.yf6mY7NVFIgfXWQ | @[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
k : ℤ
⊢ (algebraMap ℤ R) (eval k (descPochhammer ℤ n)) = eval₂ (algebraMap ℤ R) (↑k) (descPochhammer ℤ n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
| simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id] | @[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
| Mathlib.RingTheory.Polynomial.Pochhammer.280_0.yf6mY7NVFIgfXWQ | @[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ eval 0 (descPochhammer R n) = if n = 0 then 1 else 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
| cases n | theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
| Mathlib.RingTheory.Polynomial.Pochhammer.286_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
R : Type u
inst✝ : Ring R
⊢ eval 0 (descPochhammer R Nat.zero) = if Nat.zero = 0 then 1 else 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· | simp | theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· | Mathlib.RingTheory.Polynomial.Pochhammer.286_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝ : Ring R
n✝ : ℕ
⊢ eval 0 (descPochhammer R (Nat.succ n✝)) = if Nat.succ n✝ = 0 then 1 else 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· | simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left] | theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.286_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
⊢ eval 0 (descPochhammer R 0) = 1 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by | simp | theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by | Mathlib.RingTheory.Polynomial.Pochhammer.292_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
h : n ≠ 0
⊢ eval 0 (descPochhammer R n) = 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
| simp [descPochhammer_eval_zero, h] | @[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
| Mathlib.RingTheory.Polynomial.Pochhammer.294_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ descPochhammer R (n + 1) = descPochhammer R n * (X - ↑n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
| suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X])) | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n)
⊢ descPochhammer R (n + 1) = descPochhammer R n * (X - ↑n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· | apply_fun Polynomial.map (algebraMap ℤ R) at h | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· | Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
h : map (algebraMap ℤ R) (descPochhammer ℤ (n + 1)) = map (algebraMap ℤ R) (descPochhammer ℤ n * (X - ↑n))
⊢ descPochhammer R (n + 1) = descPochhammer R n * (X - ↑n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
| simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
case h
R : Type u
inst✝ : Ring R
n : ℕ
⊢ descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
| induction' n with n ih | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
case h.zero
R : Type u
inst✝ : Ring R
⊢ descPochhammer ℤ (Nat.zero + 1) = descPochhammer ℤ Nat.zero * (X - ↑Nat.zero) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· | simp [descPochhammer] | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· | Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
case h.succ
R : Type u
inst✝ : Ring R
n : ℕ
ih : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n)
⊢ descPochhammer ℤ (Nat.succ n + 1) = descPochhammer ℤ (Nat.succ n) * (X - ↑(Nat.succ n)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· | conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp] | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· | Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
ih : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n)
| descPochhammer ℤ (Nat.succ n + 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
| rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp] | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
ih : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n)
| descPochhammer ℤ (Nat.succ n + 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
| rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp] | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
ih : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n)
| descPochhammer ℤ (Nat.succ n + 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
| rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp] | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
case h.succ
R : Type u
inst✝ : Ring R
n : ℕ
ih : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n)
⊢ descPochhammer ℤ (n + 1) * (X - 1 - ↑n) = descPochhammer ℤ (Nat.succ n) * (X - ↑(Nat.succ n)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
| nth_rw 1 [Nat.succ_eq_add_one] | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
case h.succ
R : Type u
inst✝ : Ring R
n : ℕ
ih : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - ↑n)
⊢ descPochhammer ℤ (n + 1) * (X - 1 - ↑n) = descPochhammer ℤ (n + 1) * (X - ↑(Nat.succ n)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
| rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub] | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
| Mathlib.RingTheory.Polynomial.Pochhammer.298_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝² : Ring R
n : ℕ
inst✝¹ : NoZeroDivisors R
inst✝ : Nontrivial R
⊢ natDegree (descPochhammer R n) = n | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
| induction' n with n hn | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
| Mathlib.RingTheory.Polynomial.Pochhammer.312_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
R : Type u
inst✝² : Ring R
inst✝¹ : NoZeroDivisors R
inst✝ : Nontrivial R
⊢ natDegree (descPochhammer R Nat.zero) = Nat.zero | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· | simp | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· | Mathlib.RingTheory.Polynomial.Pochhammer.312_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝² : Ring R
inst✝¹ : NoZeroDivisors R
inst✝ : Nontrivial R
n : ℕ
hn : natDegree (descPochhammer R n) = n
⊢ natDegree (descPochhammer R (Nat.succ n)) = Nat.succ n | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· | have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R) | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.312_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝² : Ring R
inst✝¹ : NoZeroDivisors R
inst✝ : Nontrivial R
n : ℕ
hn : natDegree (descPochhammer R n) = n
this : natDegree (X - ↑n) = 1
⊢ natDegree (descPochhammer R (Nat.succ n)) = Nat.succ n | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
| rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this] | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
| Mathlib.RingTheory.Polynomial.Pochhammer.312_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝² : Ring R
inst✝¹ : NoZeroDivisors R
inst✝ : Nontrivial R
n : ℕ
hn : natDegree (descPochhammer R n) = n
this : natDegree (X - ↑n) = 1
⊢ descPochhammer R n ≠ 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
| cases n | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
| Mathlib.RingTheory.Polynomial.Pochhammer.312_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
R : Type u
inst✝² : Ring R
inst✝¹ : NoZeroDivisors R
inst✝ : Nontrivial R
hn : natDegree (descPochhammer R Nat.zero) = Nat.zero
this : natDegree (X - ↑Nat.zero) = 1
⊢ descPochhammer R Nat.zero ≠ 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· | simp | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· | Mathlib.RingTheory.Polynomial.Pochhammer.312_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝² : Ring R
inst✝¹ : NoZeroDivisors R
inst✝ : Nontrivial R
n✝ : ℕ
hn : natDegree (descPochhammer R (Nat.succ n✝)) = Nat.succ n✝
this : natDegree (X - ↑(Nat.succ n✝)) = 1
⊢ descPochhammer R (Nat.succ n✝) ≠ 0 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· | refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.312_0.yf6mY7NVFIgfXWQ | @[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝¹ : Ring R
S : Type u_1
inst✝ : Ring S
n : ℕ
k : S
⊢ eval k (descPochhammer S (n + 1)) = eval k (descPochhammer S n) * (k - ↑n) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
| rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub] | theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
| Mathlib.RingTheory.Polynomial.Pochhammer.324_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ comp (descPochhammer R (n + 1)) (X - 1) = descPochhammer R (n + 1) - (↑n + 1) • comp (descPochhammer R n) (X - 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
| suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
| Mathlib.RingTheory.Polynomial.Pochhammer.329_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
this : comp (descPochhammer ℤ (n + 1)) (X - 1) = descPochhammer ℤ (n + 1) - (↑n + 1) * comp (descPochhammer ℤ n) (X - 1)
⊢ comp (descPochhammer R (n + 1)) (X - 1) = descPochhammer R (n + 1) - (↑n + 1) • comp (descPochhammer R n) (X - 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by | simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by | Mathlib.RingTheory.Polynomial.Pochhammer.329_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ comp (descPochhammer ℤ (n + 1)) (X - 1) = descPochhammer ℤ (n + 1) - (↑n + 1) * comp (descPochhammer ℤ n) (X - 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
| nth_rw 2 [descPochhammer_succ_left] | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
| Mathlib.RingTheory.Polynomial.Pochhammer.329_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ comp (descPochhammer ℤ (n + 1)) (X - 1) =
X * comp (descPochhammer ℤ n) (X - 1) - (↑n + 1) * comp (descPochhammer ℤ n) (X - 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
| rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp] | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
| Mathlib.RingTheory.Polynomial.Pochhammer.329_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ (X - 1 - ↑n) * comp (descPochhammer ℤ n) (X - 1) = (X - (↑n + 1)) * comp (descPochhammer ℤ n) (X - 1) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
| ring | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
| Mathlib.RingTheory.Polynomial.Pochhammer.329_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n m : ℕ
⊢ descPochhammer R n * comp (descPochhammer R m) (X - ↑n) = descPochhammer R (n + m) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
| induction' m with m ih | theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
| Mathlib.RingTheory.Polynomial.Pochhammer.339_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) | Mathlib_RingTheory_Polynomial_Pochhammer |
case zero
R : Type u
inst✝ : Ring R
n : ℕ
⊢ descPochhammer R n * comp (descPochhammer R Nat.zero) (X - ↑n) = descPochhammer R (n + Nat.zero) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· | simp | theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· | Mathlib.RingTheory.Polynomial.Pochhammer.339_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) | Mathlib_RingTheory_Polynomial_Pochhammer |
case succ
R : Type u
inst✝ : Ring R
n m : ℕ
ih : descPochhammer R n * comp (descPochhammer R m) (X - ↑n) = descPochhammer R (n + m)
⊢ descPochhammer R n * comp (descPochhammer R (Nat.succ m)) (X - ↑n) = descPochhammer R (n + Nat.succ m) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· | rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub] | theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· | Mathlib.RingTheory.Polynomial.Pochhammer.339_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ eval (↑n) (descPochhammer ℤ 0) = ↑(Nat.descFactorial n 0) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
| rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero] | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n : ℕ
⊢ 1 = ↑1 | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
| rfl | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n t : ℕ
⊢ eval (↑n) (descPochhammer ℤ (t + 1)) = ↑(Nat.descFactorial n (t + 1)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
| rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t] | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n t : ℕ
⊢ ↑(Nat.descFactorial n t) * eval (↑n) (X - ↑t) = ↑(Nat.descFactorial n (t + 1)) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
| simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt] | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n t : ℕ
⊢ ↑(Nat.descFactorial n t) * (↑n - ↑t) = ↑(n - t) * ↑(Nat.descFactorial n t) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
| rw [mul_comm] | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n t : ℕ
⊢ (↑n - ↑t) * ↑(Nat.descFactorial n t) = ↑(n - t) * ↑(Nat.descFactorial n t) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
| simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt] | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
n t : ℕ
⊢ ↑n - ↑t = ↑(n - t) ∨ n < t | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
| by_cases h : n < t | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
case pos
R : Type u
inst✝ : Ring R
n t : ℕ
h : n < t
⊢ ↑n - ↑t = ↑(n - t) ∨ n < t | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· | tauto | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· | Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
case neg
R : Type u
inst✝ : Ring R
n t : ℕ
h : ¬n < t
⊢ ↑n - ↑t = ↑(n - t) ∨ n < t | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· | left | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· | Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
case neg.h
R : Type u
inst✝ : Ring R
n t : ℕ
h : ¬n < t
⊢ ↑n - ↑t = ↑(n - t) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
| exact (Int.ofNat_sub <| not_lt.mp h).symm | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
| Mathlib.RingTheory.Polynomial.Pochhammer.346_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm | Mathlib_RingTheory_Polynomial_Pochhammer |
R : Type u
inst✝ : Ring R
a b : ℕ
⊢ eval (↑a + ↑b) (descPochhammer ℤ b) = ↑(Nat.ascFactorial a b) | /-
Copyright (c) 2020 Scott Morrison. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Scott Morrison
-/
import Mathlib.Tactic.Abel
import Mathlib.Data.Polynomial.Degree.Definitions
import Mathlib.Data.Polynomial.Eval
import Mathlib.Data.Polynomial.Monic
import Mathlib.Data.Polynomial.RingDivision
#align_import ring_theory.polynomial.pochhammer from "leanprover-community/mathlib"@"53b216bcc1146df1c4a0a86877890ea9f1f01589"
/-!
# The Pochhammer polynomials
We define and prove some basic relations about
`ascPochhammer S n : S[X] := X * (X + 1) * ... * (X + n - 1)`
which is also known as the rising factorial and about
`descPochhammer R n : R[X] := X * (X - 1) * ... * (X - n + 1)`
which is also known as the falling factorial. Versions of this definition
that are focused on `Nat` can be found in `Data.Nat.Factorial` as `Nat.ascFactorial` and
`Nat.descFactorial`.
## Implementation
As with many other families of polynomials, even though the coefficients are always in `ℕ` or `ℤ` ,
we define the polynomial with coefficients in any `[Semiring S]` or `[Ring R]`.
## TODO
There is lots more in this direction:
* q-factorials, q-binomials, q-Pochhammer.
-/
universe u v
open Polynomial
open Polynomial
section Semiring
variable (S : Type u) [Semiring S]
/-- `ascPochhammer S n` is the polynomial `X * (X + 1) * ... * (X + n - 1)`,
with coefficients in the semiring `S`.
-/
noncomputable def ascPochhammer : ℕ → S[X]
| 0 => 1
| n + 1 => X * (ascPochhammer n).comp (X + 1)
#align pochhammer ascPochhammer
@[simp]
theorem ascPochhammer_zero : ascPochhammer S 0 = 1 :=
rfl
#align pochhammer_zero ascPochhammer_zero
@[simp]
theorem ascPochhammer_one : ascPochhammer S 1 = X := by simp [ascPochhammer]
#align pochhammer_one ascPochhammer_one
theorem ascPochhammer_succ_left (n : ℕ) :
ascPochhammer S (n + 1) = X * (ascPochhammer S n).comp (X + 1) :=
by rw [ascPochhammer]
#align pochhammer_succ_left ascPochhammer_succ_left
theorem monic_ascPochhammer (n : ℕ) [Nontrivial S] [NoZeroDivisors S] :
Monic <| ascPochhammer S n := by
induction' n with n hn
· simp
· have : leadingCoeff (X + 1 : S[X]) = 1 := leadingCoeff_X_add_C 1
rw [ascPochhammer_succ_left, Monic.def, leadingCoeff_mul,
leadingCoeff_comp (ne_zero_of_eq_one <| natDegree_X_add_C 1 : natDegree (X + 1) ≠ 0), hn,
monic_X, one_mul, one_mul, this, one_pow]
section
variable {S} {T : Type v} [Semiring T]
@[simp]
theorem ascPochhammer_map (f : S →+* T) (n : ℕ) :
(ascPochhammer S n).map f = ascPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, ascPochhammer_succ_left, map_comp]
#align pochhammer_map ascPochhammer_map
end
@[simp, norm_cast]
theorem ascPochhammer_eval_cast (n k : ℕ) :
(((ascPochhammer ℕ n).eval k : ℕ) : S) = ((ascPochhammer S n).eval k : S) := by
rw [← ascPochhammer_map (algebraMap ℕ S), eval_map, ← eq_natCast (algebraMap ℕ S),
eval₂_at_nat_cast,Nat.cast_id, eq_natCast]
#align pochhammer_eval_cast ascPochhammer_eval_cast
theorem ascPochhammer_eval_zero {n : ℕ} : (ascPochhammer S n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, ascPochhammer_succ_left]
#align pochhammer_eval_zero ascPochhammer_eval_zero
theorem ascPochhammer_zero_eval_zero : (ascPochhammer S 0).eval 0 = 1 := by simp
#align pochhammer_zero_eval_zero ascPochhammer_zero_eval_zero
@[simp]
theorem ascPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (ascPochhammer S n).eval 0 = 0 := by
simp [ascPochhammer_eval_zero, h]
#align pochhammer_ne_zero_eval_zero ascPochhammer_ne_zero_eval_zero
theorem ascPochhammer_succ_right (n : ℕ) :
ascPochhammer S (n + 1) = ascPochhammer S n * (X + (n : S[X])) := by
suffices h : ascPochhammer ℕ (n + 1) = ascPochhammer ℕ n * (X + (n : ℕ[X]))
· apply_fun Polynomial.map (algebraMap ℕ S) at h
simpa only [ascPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_nat_cast] using h
induction' n with n ih
· simp
· conv_lhs =>
rw [ascPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← ascPochhammer_succ_left, add_comp,
X_comp, nat_cast_comp, add_assoc, add_comm (1 : ℕ[X]), ← Nat.cast_succ]
#align pochhammer_succ_right ascPochhammer_succ_right
theorem ascPochhammer_succ_eval {S : Type*} [Semiring S] (n : ℕ) (k : S) :
(ascPochhammer S (n + 1)).eval k = (ascPochhammer S n).eval k * (k + n) := by
rw [ascPochhammer_succ_right, mul_add, eval_add, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_add]
#align pochhammer_succ_eval ascPochhammer_succ_eval
theorem ascPochhammer_succ_comp_X_add_one (n : ℕ) :
(ascPochhammer S (n + 1)).comp (X + 1) =
ascPochhammer S (n + 1) + (n + 1) • (ascPochhammer S n).comp (X + 1) := by
suffices (ascPochhammer ℕ (n + 1)).comp (X + 1) =
ascPochhammer ℕ (n + 1) + (n + 1) * (ascPochhammer ℕ n).comp (X + 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Nat.castRingHom S)) this
nth_rw 2 [ascPochhammer_succ_left]
rw [← add_mul, ascPochhammer_succ_right ℕ n, mul_comp, mul_comm, add_comp, X_comp, nat_cast_comp,
add_comm, ← add_assoc]
ring
set_option linter.uppercaseLean3 false in
#align pochhammer_succ_comp_X_add_one ascPochhammer_succ_comp_X_add_one
theorem ascPochhammer_mul (n m : ℕ) :
ascPochhammer S n * (ascPochhammer S m).comp (X + (n : S[X])) = ascPochhammer S (n + m) := by
induction' m with m ih
· simp
· rw [ascPochhammer_succ_right, Polynomial.mul_X_add_nat_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, ascPochhammer_succ_right, Nat.cast_add, add_assoc]
#align pochhammer_mul ascPochhammer_mul
theorem ascPochhammer_nat_eq_ascFactorial (n : ℕ) :
∀ k, (ascPochhammer ℕ k).eval (n + 1) = n.ascFactorial k
| 0 => by rw [ascPochhammer_zero, eval_one, Nat.ascFactorial_zero]
| t + 1 => by
rw [ascPochhammer_succ_right, eval_mul, ascPochhammer_nat_eq_ascFactorial n t]
simp only [eval_add, eval_X, eval_nat_cast, Nat.cast_id]
rw [Nat.ascFactorial_succ, add_right_comm, mul_comm]
#align pochhammer_nat_eq_asc_factorial ascPochhammer_nat_eq_ascFactorial
theorem ascPochhammer_nat_eq_descFactorial (a b : ℕ) :
(ascPochhammer ℕ b).eval a = (a + b - 1).descFactorial b := by
cases' b with b
· rw [Nat.descFactorial_zero, ascPochhammer_zero, Polynomial.eval_one]
rw [Nat.add_succ, Nat.succ_sub_succ, tsub_zero]
cases a
· simp only [Nat.zero_eq, ne_eq, Nat.succ_ne_zero, not_false_iff, ascPochhammer_ne_zero_eval_zero,
zero_add, Nat.descFactorial_succ, le_refl, tsub_eq_zero_of_le, zero_mul]
· rw [Nat.succ_add, ← Nat.add_succ, Nat.add_descFactorial_eq_ascFactorial,
ascPochhammer_nat_eq_ascFactorial]
#align pochhammer_nat_eq_desc_factorial ascPochhammer_nat_eq_descFactorial
@[simp]
theorem ascPochhammer_natDegree (n : ℕ) [NoZeroDivisors S] [Nontrivial S] :
(ascPochhammer S n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X + (n : S[X])) = 1 := natDegree_X_add_C (n : S)
rw [ascPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
end Semiring
section StrictOrderedSemiring
variable {S : Type*} [StrictOrderedSemiring S]
theorem ascPochhammer_pos (n : ℕ) (s : S) (h : 0 < s) : 0 < (ascPochhammer S n).eval s := by
induction' n with n ih
· simp only [Nat.zero_eq, ascPochhammer_zero, eval_one]
exact zero_lt_one
· rw [ascPochhammer_succ_right, mul_add, eval_add, ← Nat.cast_comm, eval_nat_cast_mul, eval_mul_X,
Nat.cast_comm, ← mul_add]
exact mul_pos ih (lt_of_lt_of_le h ((le_add_iff_nonneg_right _).mpr (Nat.cast_nonneg n)))
#align pochhammer_pos ascPochhammer_pos
end StrictOrderedSemiring
section Factorial
open Nat
variable (S : Type*) [Semiring S] (r n : ℕ)
@[simp]
theorem ascPochhammer_eval_one (S : Type*) [Semiring S] (n : ℕ) :
(ascPochhammer S n).eval (1 : S) = (n ! : S) := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.zero_ascFactorial]
#align pochhammer_eval_one ascPochhammer_eval_one
theorem factorial_mul_ascPochhammer (S : Type*) [Semiring S] (r n : ℕ) :
(r ! : S) * (ascPochhammer S n).eval (r + 1 : S) = (r + n)! := by
rw_mod_cast [ascPochhammer_nat_eq_ascFactorial, Nat.factorial_mul_ascFactorial]
#align factorial_mul_pochhammer factorial_mul_ascPochhammer
theorem ascPochhammer_nat_eval_succ (r : ℕ) :
∀ n : ℕ, n * (ascPochhammer ℕ r).eval (n + 1) = (n + r) * (ascPochhammer ℕ r).eval n
| 0 => by
by_cases h : r = 0
· simp only [h, zero_mul, zero_add]
· simp only [ascPochhammer_eval_zero, zero_mul, if_neg h, mul_zero]
| k + 1 => by simp only [ascPochhammer_nat_eq_ascFactorial, Nat.succ_ascFactorial, add_right_comm]
#align pochhammer_nat_eval_succ ascPochhammer_nat_eval_succ
theorem ascPochhammer_eval_succ (r n : ℕ) :
(n : S) * (ascPochhammer S r).eval (n + 1 : S) =
(n + r) * (ascPochhammer S r).eval (n : S) :=
mod_cast congr_arg Nat.cast (ascPochhammer_nat_eval_succ r n)
#align pochhammer_eval_succ ascPochhammer_eval_succ
end Factorial
section Ring
variable (R : Type u) [Ring R]
/-- `descPochhammer R n` is the polynomial `X * (X - 1) * ... * (X - n + 1)`,
with coefficients in the ring `R`.
-/
noncomputable def descPochhammer : ℕ → R[X]
| 0 => 1
| n + 1 => X * (descPochhammer n).comp (X - 1)
@[simp]
theorem descPochhammer_zero : descPochhammer R 0 = 1 :=
rfl
@[simp]
theorem descPochhammer_one : descPochhammer R 1 = X := by simp [descPochhammer]
theorem descPochhammer_succ_left (n : ℕ) :
descPochhammer R (n + 1) = X * (descPochhammer R n).comp (X - 1) :=
by rw [descPochhammer]
theorem monic_descPochhammer (n : ℕ) [Nontrivial R] [NoZeroDivisors R] :
Monic <| descPochhammer R n := by
induction' n with n hn
· simp
· have h : leadingCoeff (X - 1 : R[X]) = 1 := leadingCoeff_X_sub_C 1
have : natDegree (X - (1 : R[X])) ≠ 0 := ne_zero_of_eq_one <| natDegree_X_sub_C (1 : R)
rw [descPochhammer_succ_left, Monic.def, leadingCoeff_mul, leadingCoeff_comp this, hn, monic_X,
one_mul, one_mul, h, one_pow]
section
variable {R} {T : Type v} [Ring T]
@[simp]
theorem descPochhammer_map (f : R →+* T) (n : ℕ) :
(descPochhammer R n).map f = descPochhammer T n := by
induction' n with n ih
· simp
· simp [ih, descPochhammer_succ_left, map_comp]
end
@[simp, norm_cast]
theorem descPochhammer_eval_cast (n : ℕ) (k : ℤ) :
(((descPochhammer ℤ n).eval k : ℤ) : R) = ((descPochhammer R n).eval k : R) := by
rw [← descPochhammer_map (algebraMap ℤ R), eval_map, ← eq_intCast (algebraMap ℤ R)]
simp only [algebraMap_int_eq, eq_intCast, eval₂_at_int_cast, Nat.cast_id, eq_natCast, Int.cast_id]
theorem descPochhammer_eval_zero {n : ℕ} :
(descPochhammer R n).eval 0 = if n = 0 then 1 else 0 := by
cases n
· simp
· simp [X_mul, Nat.succ_ne_zero, descPochhammer_succ_left]
theorem descPochhammer_zero_eval_zero : (descPochhammer R 0).eval 0 = 1 := by simp
@[simp]
theorem descPochhammer_ne_zero_eval_zero {n : ℕ} (h : n ≠ 0) : (descPochhammer R n).eval 0 = 0 := by
simp [descPochhammer_eval_zero, h]
theorem descPochhammer_succ_right (n : ℕ) :
descPochhammer R (n + 1) = descPochhammer R n * (X - (n : R[X])) := by
suffices h : descPochhammer ℤ (n + 1) = descPochhammer ℤ n * (X - (n : ℤ[X]))
· apply_fun Polynomial.map (algebraMap ℤ R) at h
simpa [descPochhammer_map, Polynomial.map_mul, Polynomial.map_add, map_X,
Polynomial.map_int_cast] using h
induction' n with n ih
· simp [descPochhammer]
· conv_lhs =>
rw [descPochhammer_succ_left, ih, mul_comp, ← mul_assoc, ← descPochhammer_succ_left, sub_comp,
X_comp, nat_cast_comp]
nth_rw 1 [Nat.succ_eq_add_one]
rw [Nat.succ_eq_one_add, Nat.cast_add, Nat.cast_one, sub_add_eq_sub_sub]
@[simp]
theorem descPochhammer_natDegree (n : ℕ) [NoZeroDivisors R] [Nontrivial R] :
(descPochhammer R n).natDegree = n := by
induction' n with n hn
· simp
· have : natDegree (X - (n : R[X])) = 1 := natDegree_X_sub_C (n : R)
rw [descPochhammer_succ_right,
natDegree_mul _ (ne_zero_of_natDegree_gt <| this.symm ▸ Nat.zero_lt_one), hn, this]
cases n
· simp
· refine' ne_zero_of_natDegree_gt <| hn.symm ▸ Nat.succ_pos _
theorem descPochhammer_succ_eval {S : Type*} [Ring S] (n : ℕ) (k : S) :
(descPochhammer S (n + 1)).eval k = (descPochhammer S n).eval k * (k - n) := by
rw [descPochhammer_succ_right, mul_sub, eval_sub, eval_mul_X, ← Nat.cast_comm, ← C_eq_nat_cast,
eval_C_mul, Nat.cast_comm, ← mul_sub]
theorem descPochhammer_succ_comp_X_sub_one (n : ℕ) :
(descPochhammer R (n + 1)).comp (X - 1) =
descPochhammer R (n + 1) - (n + (1 : R[X])) • (descPochhammer R n).comp (X - 1) := by
suffices (descPochhammer ℤ (n + 1)).comp (X - 1) =
descPochhammer ℤ (n + 1) - (n + 1) * (descPochhammer ℤ n).comp (X - 1)
by simpa [map_comp] using congr_arg (Polynomial.map (Int.castRingHom R)) this
nth_rw 2 [descPochhammer_succ_left]
rw [← sub_mul, descPochhammer_succ_right ℤ n, mul_comp, mul_comm, sub_comp, X_comp, nat_cast_comp]
ring
theorem descPochhammer_mul (n m : ℕ) :
descPochhammer R n * (descPochhammer R m).comp (X - (n : R[X])) = descPochhammer R (n + m) := by
induction' m with m ih
· simp
· rw [descPochhammer_succ_right, Polynomial.mul_X_sub_int_cast_comp, ← mul_assoc, ih,
Nat.succ_eq_add_one, ← add_assoc, descPochhammer_succ_right, Nat.cast_add, sub_add_eq_sub_sub]
theorem descPochhammer_int_eq_descFactorial (n : ℕ) :
∀ k, (descPochhammer ℤ k).eval (n : ℤ) = n.descFactorial k
| 0 => by
rw [descPochhammer_zero, eval_one, Nat.descFactorial_zero]
rfl
| t + 1 => by
rw [descPochhammer_succ_right, eval_mul, descPochhammer_int_eq_descFactorial n t]
simp only [eval_sub, eval_X, eval_nat_cast, Nat.descFactorial_succ, Nat.cast_mul,
Nat.descFactorial_eq_zero_iff_lt]
rw [mul_comm]
simp only [mul_eq_mul_right_iff, Nat.cast_eq_zero, Nat.descFactorial_eq_zero_iff_lt]
by_cases h : n < t
· tauto
· left
exact (Int.ofNat_sub <| not_lt.mp h).symm
theorem descPochhammer_int_eq_ascFactorial (a b : ℕ) :
(descPochhammer ℤ b).eval (a + b : ℤ) = a.ascFactorial b := by
| rw [← Nat.cast_add, descPochhammer_int_eq_descFactorial (a + b) b,
Nat.add_descFactorial_eq_ascFactorial] | theorem descPochhammer_int_eq_ascFactorial (a b : ℕ) :
(descPochhammer ℤ b).eval (a + b : ℤ) = a.ascFactorial b := by
| Mathlib.RingTheory.Polynomial.Pochhammer.362_0.yf6mY7NVFIgfXWQ | theorem descPochhammer_int_eq_ascFactorial (a b : ℕ) :
(descPochhammer ℤ b).eval (a + b : ℤ) = a.ascFactorial b | Mathlib_RingTheory_Polynomial_Pochhammer |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝² : Preorder ι
inst✝¹ : OrderBot ι
inst✝ : InfSet ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
⊢ upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
| rw [upperCrossingTime] | theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
| Mathlib.Probability.Martingale.Upcrossing.168_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝² : Preorder ι
inst✝¹ : OrderBot ι
inst✝ : InfSet ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω✝ ω : Ω
⊢ upperCrossingTime a b f N (n + 1) ω = hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
| simp only [upperCrossingTime_succ] | theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
| Mathlib.Probability.Martingale.Upcrossing.173_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝² : Preorder ι
inst✝¹ : OrderBot ι
inst✝ : InfSet ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω✝ ω : Ω
⊢ hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
| rfl | theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
| Mathlib.Probability.Martingale.Upcrossing.173_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
⊢ upperCrossingTime a b f N n ω ≤ N | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
| cases n | theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
| Mathlib.Probability.Martingale.Upcrossing.187_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N | Mathlib_Probability_Martingale_Upcrossing |
case zero
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
m : ℕ
ω : Ω
⊢ upperCrossingTime a b f N Nat.zero ω ≤ N | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· | simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq] | theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· | Mathlib.Probability.Martingale.Upcrossing.187_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N | Mathlib_Probability_Martingale_Upcrossing |
case succ
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
m : ℕ
ω : Ω
n✝ : ℕ
⊢ upperCrossingTime a b f N (Nat.succ n✝) ω ≤ N | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· | simp only [upperCrossingTime_succ, hitting_le] | theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· | Mathlib.Probability.Martingale.Upcrossing.187_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
⊢ lowerCrossingTime a b f N n ω ≤ N | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
| simp only [lowerCrossingTime, hitting_le ω] | theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
| Mathlib.Probability.Martingale.Upcrossing.198_0.80Cpy4Qgm9i1y9y | theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
⊢ upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
| simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω] | theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
| Mathlib.Probability.Martingale.Upcrossing.202_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
⊢ lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
| rw [upperCrossingTime_succ] | theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
| Mathlib.Probability.Martingale.Upcrossing.207_0.80Cpy4Qgm9i1y9y | theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
⊢ lowerCrossingTime a b f N n ω ≤
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
| exact le_hitting lowerCrossingTime_le ω | theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
| Mathlib.Probability.Martingale.Upcrossing.207_0.80Cpy4Qgm9i1y9y | theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
hnm : n ≤ m
⊢ lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
| suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm | theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
| Mathlib.Probability.Martingale.Upcrossing.213_0.80Cpy4Qgm9i1y9y | theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
hnm : n ≤ m
this : Monotone fun n => lowerCrossingTime a b f N n ω
⊢ lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by | exact this hnm | theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by | Mathlib.Probability.Martingale.Upcrossing.213_0.80Cpy4Qgm9i1y9y | theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
hnm : n ≤ m
⊢ Monotone fun n => lowerCrossingTime a b f N n ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
| exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime | theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
| Mathlib.Probability.Martingale.Upcrossing.213_0.80Cpy4Qgm9i1y9y | theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
hnm : n ≤ m
⊢ upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
| suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm | theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
| Mathlib.Probability.Martingale.Upcrossing.220_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
hnm : n ≤ m
this : Monotone fun n => upperCrossingTime a b f N n ω
⊢ upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by | exact this hnm | theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by | Mathlib.Probability.Martingale.Upcrossing.220_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
inst✝ : ConditionallyCompleteLinearOrderBot ι
a b : ℝ
f : ι → Ω → ℝ
N : ι
n m : ℕ
ω : Ω
hnm : n ≤ m
⊢ Monotone fun n => upperCrossingTime a b f N n ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
| exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ | theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
| Mathlib.Probability.Martingale.Upcrossing.220_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
a b : ℝ
f : ℕ → Ω → ℝ
N n m : ℕ
ω : Ω
h : lowerCrossingTime a b f N n ω ≠ N
⊢ stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ
#align measure_theory.upper_crossing_time_mono MeasureTheory.upperCrossingTime_mono
end ConditionallyCompleteLinearOrderBot
variable {a b : ℝ} {f : ℕ → Ω → ℝ} {N : ℕ} {n m : ℕ} {ω : Ω}
theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
| obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl | theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
| Mathlib.Probability.Martingale.Upcrossing.231_0.80Cpy4Qgm9i1y9y | theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a | Mathlib_Probability_Martingale_Upcrossing |
case intro.intro
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
a b : ℝ
f : ℕ → Ω → ℝ
N n m : ℕ
ω : Ω
h : lowerCrossingTime a b f N n ω ≠ N
j : ℕ
hj₁ : j ∈ Set.Icc (upperCrossingTime a b f N n ω) (lowerCrossingTime a b f N n ω)
hj₂ : f j ω ∈ Set.Iic a
⊢ stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ
#align measure_theory.upper_crossing_time_mono MeasureTheory.upperCrossingTime_mono
end ConditionallyCompleteLinearOrderBot
variable {a b : ℝ} {f : ℕ → Ω → ℝ} {N : ℕ} {n m : ℕ} {ω : Ω}
theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl
| exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 lowerCrossingTime_le⟩, hj₂⟩ | theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl
| Mathlib.Probability.Martingale.Upcrossing.231_0.80Cpy4Qgm9i1y9y | theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
a b : ℝ
f : ℕ → Ω → ℝ
N n m : ℕ
ω : Ω
h : upperCrossingTime a b f N (n + 1) ω ≠ N
⊢ b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ
#align measure_theory.upper_crossing_time_mono MeasureTheory.upperCrossingTime_mono
end ConditionallyCompleteLinearOrderBot
variable {a b : ℝ} {f : ℕ → Ω → ℝ} {N : ℕ} {n m : ℕ} {ω : Ω}
theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 lowerCrossingTime_le⟩, hj₂⟩
#align measure_theory.stopped_value_lower_crossing_time MeasureTheory.stoppedValue_lowerCrossingTime
theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω := by
| obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne upperCrossingTime_le h)).1 le_rfl | theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω := by
| Mathlib.Probability.Martingale.Upcrossing.237_0.80Cpy4Qgm9i1y9y | theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω | Mathlib_Probability_Martingale_Upcrossing |
case intro.intro
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
a b : ℝ
f : ℕ → Ω → ℝ
N n m : ℕ
ω : Ω
h : upperCrossingTime a b f N (n + 1) ω ≠ N
j : ℕ
hj₁ :
j ∈
Set.Icc (lowerCrossingTimeAux a f (upperCrossingTime a b f N (Nat.add n 0) ω) N ω)
(upperCrossingTime a b f N (n + 1) ω)
hj₂ : f j ω ∈ Set.Ici b
⊢ b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ
#align measure_theory.upper_crossing_time_mono MeasureTheory.upperCrossingTime_mono
end ConditionallyCompleteLinearOrderBot
variable {a b : ℝ} {f : ℕ → Ω → ℝ} {N : ℕ} {n m : ℕ} {ω : Ω}
theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 lowerCrossingTime_le⟩, hj₂⟩
#align measure_theory.stopped_value_lower_crossing_time MeasureTheory.stoppedValue_lowerCrossingTime
theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne upperCrossingTime_le h)).1 le_rfl
| exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 (hitting_le _)⟩, hj₂⟩ | theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne upperCrossingTime_le h)).1 le_rfl
| Mathlib.Probability.Martingale.Upcrossing.237_0.80Cpy4Qgm9i1y9y | theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
a b : ℝ
f : ℕ → Ω → ℝ
N n m : ℕ
ω : Ω
hab : a < b
hn : lowerCrossingTime a b f N (n + 1) ω ≠ N
⊢ upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ
#align measure_theory.upper_crossing_time_mono MeasureTheory.upperCrossingTime_mono
end ConditionallyCompleteLinearOrderBot
variable {a b : ℝ} {f : ℕ → Ω → ℝ} {N : ℕ} {n m : ℕ} {ω : Ω}
theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 lowerCrossingTime_le⟩, hj₂⟩
#align measure_theory.stopped_value_lower_crossing_time MeasureTheory.stoppedValue_lowerCrossingTime
theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne upperCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 (hitting_le _)⟩, hj₂⟩
#align measure_theory.stopped_value_upper_crossing_time MeasureTheory.stoppedValue_upperCrossingTime
theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω := by
| refine' lt_of_le_of_ne upperCrossingTime_le_lowerCrossingTime fun h =>
not_le.2 hab <| le_trans _ (stoppedValue_lowerCrossingTime hn) | theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω := by
| Mathlib.Probability.Martingale.Upcrossing.243_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
a b : ℝ
f : ℕ → Ω → ℝ
N n m : ℕ
ω : Ω
hab : a < b
hn : lowerCrossingTime a b f N (n + 1) ω ≠ N
h : upperCrossingTime a b f N (n + 1) ω = lowerCrossingTime a b f N (n + 1) ω
⊢ b ≤ stoppedValue f (lowerCrossingTime a b f N (n + 1)) ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ
#align measure_theory.upper_crossing_time_mono MeasureTheory.upperCrossingTime_mono
end ConditionallyCompleteLinearOrderBot
variable {a b : ℝ} {f : ℕ → Ω → ℝ} {N : ℕ} {n m : ℕ} {ω : Ω}
theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 lowerCrossingTime_le⟩, hj₂⟩
#align measure_theory.stopped_value_lower_crossing_time MeasureTheory.stoppedValue_lowerCrossingTime
theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne upperCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 (hitting_le _)⟩, hj₂⟩
#align measure_theory.stopped_value_upper_crossing_time MeasureTheory.stoppedValue_upperCrossingTime
theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω := by
refine' lt_of_le_of_ne upperCrossingTime_le_lowerCrossingTime fun h =>
not_le.2 hab <| le_trans _ (stoppedValue_lowerCrossingTime hn)
| simp only [stoppedValue] | theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω := by
refine' lt_of_le_of_ne upperCrossingTime_le_lowerCrossingTime fun h =>
not_le.2 hab <| le_trans _ (stoppedValue_lowerCrossingTime hn)
| Mathlib.Probability.Martingale.Upcrossing.243_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω | Mathlib_Probability_Martingale_Upcrossing |
Ω : Type u_1
ι : Type u_2
m0 : MeasurableSpace Ω
μ : Measure Ω
a b : ℝ
f : ℕ → Ω → ℝ
N n m : ℕ
ω : Ω
hab : a < b
hn : lowerCrossingTime a b f N (n + 1) ω ≠ N
h : upperCrossingTime a b f N (n + 1) ω = lowerCrossingTime a b f N (n + 1) ω
⊢ b ≤ f (lowerCrossingTime a b f N (n + 1) ω) ω | /-
Copyright (c) 2022 Kexing Ying. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Kexing Ying
-/
import Mathlib.Data.Set.Intervals.Monotone
import Mathlib.Probability.Process.HittingTime
import Mathlib.Probability.Martingale.Basic
#align_import probability.martingale.upcrossing from "leanprover-community/mathlib"@"2c1d8ca2812b64f88992a5294ea3dba144755cd1"
/-!
# Doob's upcrossing estimate
Given a discrete real-valued submartingale $(f_n)_{n \in \mathbb{N}}$, denoting by $U_N(a, b)$ the
number of times $f_n$ crossed from below $a$ to above $b$ before time $N$, Doob's upcrossing
estimate (also known as Doob's inequality) states that
$$(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(f_N - a)^+].$$
Doob's upcrossing estimate is an important inequality and is central in proving the martingale
convergence theorems.
## Main definitions
* `MeasureTheory.upperCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing above `b` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.lowerCrossingTime a b f N n`: is the stopping time corresponding to `f`
crossing below `a` the `n`-th time before time `N` (if this does not occur then the value is
taken to be `N`).
* `MeasureTheory.upcrossingStrat a b f N`: is the predictable process which is 1 if `n` is
between a consecutive pair of lower and upper crossings and is 0 otherwise. Intuitively
one might think of the `upcrossingStrat` as the strategy of buying 1 share whenever the process
crosses below `a` for the first time after selling and selling 1 share whenever the process
crosses above `b` for the first time after buying.
* `MeasureTheory.upcrossingsBefore a b f N`: is the number of times `f` crosses from below `a` to
above `b` before time `N`.
* `MeasureTheory.upcrossings a b f`: is the number of times `f` crosses from below `a` to above
`b`. This takes value in `ℝ≥0∞` and so is allowed to be `∞`.
## Main results
* `MeasureTheory.Adapted.isStoppingTime_upperCrossingTime`: `upperCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Adapted.isStoppingTime_lowerCrossingTime`: `lowerCrossingTime` is a
stopping time whenever the process it is associated to is adapted.
* `MeasureTheory.Submartingale.mul_integral_upcrossingsBefore_le_integral_pos_part`: Doob's
upcrossing estimate.
* `MeasureTheory.Submartingale.mul_lintegral_upcrossings_le_lintegral_pos_part`: the inequality
obtained by taking the supremum on both sides of Doob's upcrossing estimate.
### References
We mostly follow the proof from [Kallenberg, *Foundations of modern probability*][kallenberg2021]
-/
open TopologicalSpace Filter
open scoped NNReal ENNReal MeasureTheory ProbabilityTheory BigOperators Topology
namespace MeasureTheory
variable {Ω ι : Type*} {m0 : MeasurableSpace Ω} {μ : Measure Ω}
/-!
## Proof outline
In this section, we will denote by $U_N(a, b)$ the number of upcrossings of $(f_n)$ from below $a$
to above $b$ before time $N$.
To define $U_N(a, b)$, we will construct two stopping times corresponding to when $(f_n)$ crosses
below $a$ and above $b$. Namely, we define
$$
\sigma_n := \inf \{n \ge \tau_n \mid f_n \le a\} \wedge N;
$$
$$
\tau_{n + 1} := \inf \{n \ge \sigma_n \mid f_n \ge b\} \wedge N.
$$
These are `lowerCrossingTime` and `upperCrossingTime` in our formalization which are defined
using `MeasureTheory.hitting` allowing us to specify a starting and ending time.
Then, we may simply define $U_N(a, b) := \sup \{n \mid \tau_n < N\}$.
Fixing $a < b \in \mathbb{R}$, we will first prove the theorem in the special case that
$0 \le f_0$ and $a \le f_N$. In particular, we will show
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[f_N].
$$
This is `MeasureTheory.integral_mul_upcrossingsBefore_le_integral` in our formalization.
To prove this, we use the fact that given a non-negative, bounded, predictable process $(C_n)$
(i.e. $(C_{n + 1})$ is adapted), $(C \bullet f)_n := \sum_{k \le n} C_{k + 1}(f_{k + 1} - f_k)$ is
a submartingale if $(f_n)$ is.
Define $C_n := \sum_{k \le n} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)$. It is easy to see that
$(1 - C_n)$ is non-negative, bounded and predictable, and hence, given a submartingale $(f_n)$,
$(1 - C) \bullet f$ is also a submartingale. Thus, by the submartingale property,
$0 \le \mathbb{E}[((1 - C) \bullet f)_0] \le \mathbb{E}[((1 - C) \bullet f)_N]$ implying
$$
\mathbb{E}[(C \bullet f)_N] \le \mathbb{E}[(1 \bullet f)_N] = \mathbb{E}[f_N] - \mathbb{E}[f_0].
$$
Furthermore,
\begin{align}
(C \bullet f)_N & =
\sum_{n \le N} \sum_{k \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} \sum_{n \le N} \mathbf{1}_{[\sigma_k, \tau_{k + 1})}(n)(f_{n + 1} - f_n)\\
& = \sum_{k \le N} (f_{\sigma_k + 1} - f_{\sigma_k} + f_{\sigma_k + 2} - f_{\sigma_k + 1}
+ \cdots + f_{\tau_{k + 1}} - f_{\tau_{k + 1} - 1})\\
& = \sum_{k \le N} (f_{\tau_{k + 1}} - f_{\sigma_k})
\ge \sum_{k < U_N(a, b)} (b - a) = (b - a) U_N(a, b)
\end{align}
where the inequality follows since for all $k < U_N(a, b)$,
$f_{\tau_{k + 1}} - f_{\sigma_k} \ge b - a$ while for all $k > U_N(a, b)$,
$f_{\tau_{k + 1}} = f_{\sigma_k} = f_N$ and
$f_{\tau_{U_N(a, b) + 1}} - f_{\sigma_{U_N(a, b)}} = f_N - a \ge 0$. Hence, we have
$$
(b - a) \mathbb{E}[U_N(a, b)] \le \mathbb{E}[(C \bullet f)_N]
\le \mathbb{E}[f_N] - \mathbb{E}[f_0] \le \mathbb{E}[f_N],
$$
as required.
To obtain the general case, we simply apply the above to $((f_n - a)^+)_n$.
-/
/-- `lowerCrossingTimeAux a f c N` is the first time `f` reached below `a` after time `c` before
time `N`. -/
noncomputable def lowerCrossingTimeAux [Preorder ι] [InfSet ι] (a : ℝ) (f : ι → Ω → ℝ) (c N : ι) :
Ω → ι :=
hitting f (Set.Iic a) c N
#align measure_theory.lower_crossing_time_aux MeasureTheory.lowerCrossingTimeAux
/-- `upperCrossingTime a b f N n` is the first time before time `N`, `f` reaches
above `b` after `f` reached below `a` for the `n - 1`-th time. -/
noncomputable def upperCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) : ℕ → Ω → ι
| 0 => ⊥
| n + 1 => fun ω =>
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω
#align measure_theory.upper_crossing_time MeasureTheory.upperCrossingTime
/-- `lowerCrossingTime a b f N n` is the first time before time `N`, `f` reaches
below `a` after `f` reached above `b` for the `n`-th time. -/
noncomputable def lowerCrossingTime [Preorder ι] [OrderBot ι] [InfSet ι] (a b : ℝ) (f : ι → Ω → ℝ)
(N : ι) (n : ℕ) : Ω → ι := fun ω => hitting f (Set.Iic a) (upperCrossingTime a b f N n ω) N ω
#align measure_theory.lower_crossing_time MeasureTheory.lowerCrossingTime
section
variable [Preorder ι] [OrderBot ι] [InfSet ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
@[simp]
theorem upperCrossingTime_zero : upperCrossingTime a b f N 0 = ⊥ :=
rfl
#align measure_theory.upper_crossing_time_zero MeasureTheory.upperCrossingTime_zero
@[simp]
theorem lowerCrossingTime_zero : lowerCrossingTime a b f N 0 = hitting f (Set.Iic a) ⊥ N :=
rfl
#align measure_theory.lower_crossing_time_zero MeasureTheory.lowerCrossingTime_zero
theorem upperCrossingTime_succ : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTimeAux a f (upperCrossingTime a b f N n ω) N ω) N ω := by
rw [upperCrossingTime]
#align measure_theory.upper_crossing_time_succ MeasureTheory.upperCrossingTime_succ
theorem upperCrossingTime_succ_eq (ω : Ω) : upperCrossingTime a b f N (n + 1) ω =
hitting f (Set.Ici b) (lowerCrossingTime a b f N n ω) N ω := by
simp only [upperCrossingTime_succ]
rfl
#align measure_theory.upper_crossing_time_succ_eq MeasureTheory.upperCrossingTime_succ_eq
end
section ConditionallyCompleteLinearOrderBot
variable [ConditionallyCompleteLinearOrderBot ι]
variable {a b : ℝ} {f : ι → Ω → ℝ} {N : ι} {n m : ℕ} {ω : Ω}
theorem upperCrossingTime_le : upperCrossingTime a b f N n ω ≤ N := by
cases n
· simp only [upperCrossingTime_zero, Pi.bot_apply, bot_le, Nat.zero_eq]
· simp only [upperCrossingTime_succ, hitting_le]
#align measure_theory.upper_crossing_time_le MeasureTheory.upperCrossingTime_le
@[simp]
theorem upperCrossingTime_zero' : upperCrossingTime a b f ⊥ n ω = ⊥ :=
eq_bot_iff.2 upperCrossingTime_le
#align measure_theory.upper_crossing_time_zero' MeasureTheory.upperCrossingTime_zero'
theorem lowerCrossingTime_le : lowerCrossingTime a b f N n ω ≤ N := by
simp only [lowerCrossingTime, hitting_le ω]
#align measure_theory.lower_crossing_time_le MeasureTheory.lowerCrossingTime_le
theorem upperCrossingTime_le_lowerCrossingTime :
upperCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N n ω := by
simp only [lowerCrossingTime, le_hitting upperCrossingTime_le ω]
#align measure_theory.upper_crossing_time_le_lower_crossing_time MeasureTheory.upperCrossingTime_le_lowerCrossingTime
theorem lowerCrossingTime_le_upperCrossingTime_succ :
lowerCrossingTime a b f N n ω ≤ upperCrossingTime a b f N (n + 1) ω := by
rw [upperCrossingTime_succ]
exact le_hitting lowerCrossingTime_le ω
#align measure_theory.lower_crossing_time_le_upper_crossing_time_succ MeasureTheory.lowerCrossingTime_le_upperCrossingTime_succ
theorem lowerCrossingTime_mono (hnm : n ≤ m) :
lowerCrossingTime a b f N n ω ≤ lowerCrossingTime a b f N m ω := by
suffices Monotone fun n => lowerCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans lowerCrossingTime_le_upperCrossingTime_succ upperCrossingTime_le_lowerCrossingTime
#align measure_theory.lower_crossing_time_mono MeasureTheory.lowerCrossingTime_mono
theorem upperCrossingTime_mono (hnm : n ≤ m) :
upperCrossingTime a b f N n ω ≤ upperCrossingTime a b f N m ω := by
suffices Monotone fun n => upperCrossingTime a b f N n ω by exact this hnm
exact monotone_nat_of_le_succ fun n =>
le_trans upperCrossingTime_le_lowerCrossingTime lowerCrossingTime_le_upperCrossingTime_succ
#align measure_theory.upper_crossing_time_mono MeasureTheory.upperCrossingTime_mono
end ConditionallyCompleteLinearOrderBot
variable {a b : ℝ} {f : ℕ → Ω → ℝ} {N : ℕ} {n m : ℕ} {ω : Ω}
theorem stoppedValue_lowerCrossingTime (h : lowerCrossingTime a b f N n ω ≠ N) :
stoppedValue f (lowerCrossingTime a b f N n) ω ≤ a := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne lowerCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 lowerCrossingTime_le⟩, hj₂⟩
#align measure_theory.stopped_value_lower_crossing_time MeasureTheory.stoppedValue_lowerCrossingTime
theorem stoppedValue_upperCrossingTime (h : upperCrossingTime a b f N (n + 1) ω ≠ N) :
b ≤ stoppedValue f (upperCrossingTime a b f N (n + 1)) ω := by
obtain ⟨j, hj₁, hj₂⟩ := (hitting_le_iff_of_lt _ (lt_of_le_of_ne upperCrossingTime_le h)).1 le_rfl
exact stoppedValue_hitting_mem ⟨j, ⟨hj₁.1, le_trans hj₁.2 (hitting_le _)⟩, hj₂⟩
#align measure_theory.stopped_value_upper_crossing_time MeasureTheory.stoppedValue_upperCrossingTime
theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω := by
refine' lt_of_le_of_ne upperCrossingTime_le_lowerCrossingTime fun h =>
not_le.2 hab <| le_trans _ (stoppedValue_lowerCrossingTime hn)
simp only [stoppedValue]
| rw [← h] | theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω := by
refine' lt_of_le_of_ne upperCrossingTime_le_lowerCrossingTime fun h =>
not_le.2 hab <| le_trans _ (stoppedValue_lowerCrossingTime hn)
simp only [stoppedValue]
| Mathlib.Probability.Martingale.Upcrossing.243_0.80Cpy4Qgm9i1y9y | theorem upperCrossingTime_lt_lowerCrossingTime (hab : a < b)
(hn : lowerCrossingTime a b f N (n + 1) ω ≠ N) :
upperCrossingTime a b f N (n + 1) ω < lowerCrossingTime a b f N (n + 1) ω | Mathlib_Probability_Martingale_Upcrossing |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.