openAI-math-dataset / lean /OAI /NumberTheory /TwoPoint /PretentiousDistance.lean
roseking's picture
Add lean source and binary files (part 118)
180bc03 verified
Raw History Blame Contribute Delete
9.09 kB
import Mathlib.Analysis.Complex.Trigonometric
import Mathlib.NumberTheory.DirichletCharacter.Bounds
import Mathlib.Tactic
namespace OAI
/-!
# The pretentious distance used in ordinary two-point correlations
These definitions use the prime cutoff and the twist-height cutoff at the same
integer scale. The finite-prime perturbation estimate below is Lemma
`lem:finite-prime-stability` of *Ordinary two-point correlations of
multiplicative functions* (24 September 2026). It is elementary and does not
assume any analytic input.
-/
namespace TwoPointCorrelations
open scoped ComplexConjugate
open Filter Finset
/-- The primes not exceeding an integer cutoff. -/
def primesUpTo (N : ℕ) : Finset ℕ := (Finset.range (N + 1)).filter Nat.Prime
/-- The square of the pretentious distance; the second sequence can be any twist. -/
noncomputable def squaredDistance (f g : ℕ → ℂ) (N : ℕ) : ℝ :=
∑ p ∈ primesUpTo N, (1 - (f p * conj (g p)).re) / (p : ℝ)
/-- The sequence `χ(n) n^(it)`, with the value at zero immaterial to the prime sum. -/
noncomputable def characterTwist {q : ℕ} (χ : DirichletCharacter ℂ q)
(t : ℝ) (n : ℕ) : ℂ :=
χ (n : ZMod q) * Complex.exp (((t * Real.log (n : ℝ) : ℝ) : ℂ) * Complex.I)
/-- Uniform nonpretentiousness against every fixed Dirichlet character.
The quantifier over all real lower bounds expresses divergence of the squared
distance uniformly over `|t| ≤ N`. As the distance is nonnegative, this is
exactly the divergence of the distance appearing in the manuscript.
-/
def UniformlyNonpretentious (f : ℕ → ℂ) : Prop :=
∀ (q : ℕ), 0 < q → ∀ (χ : DirichletCharacter ℂ q) (K : ℝ),
∀ᶠ N : ℕ in atTop, ∀ t : ℝ, |t| ≤ (N : ℝ) →
K ≤ squaredDistance f (characterTwist χ t) N
lemma characterTwist_norm_le_one {q : ℕ} (χ : DirichletCharacter ℂ q)
(t : ℝ) (n : ℕ) : ‖characterTwist χ t n‖ ≤ 1 := by
simpa only [characterTwist, norm_mul, Complex.norm_exp_ofReal_mul_I, mul_one] using
χ.norm_le_one (n : ZMod q)
lemma squaredDistance_nonneg (f g : ℕ → ℂ) (N : ℕ)
(hf : ∀ n, ‖f n‖ ≤ 1) (hg : ∀ n, ‖g n‖ ≤ 1) :
0 ≤ squaredDistance f g N := by
apply Finset.sum_nonneg
intro p hp
apply div_nonneg
· have hnorm : ‖f p * conj (g p)‖ ≤ 1 := by
calc
‖f p * conj (g p)‖ = ‖f p‖ * ‖g p‖ := by simp
_ ≤ 1 * 1 := mul_le_mul (hf p) (hg p) (norm_nonneg _) zero_le_one
_ = 1 := by norm_num
exact sub_nonneg.mpr ((Complex.re_le_norm _).trans hnorm)
· positivity
lemma squaredDistance_conj (f g : ℕ → ℂ) (N : ℕ) :
squaredDistance (fun n => conj (f n)) (fun n => conj (g n)) N =
squaredDistance f g N := by
unfold squaredDistance
apply Finset.sum_congr rfl
intro p hp
congr 2
rw [← map_mul]
simp
/-- Conjugating a character and reversing the height conjugates its twist. -/
lemma characterTwist_conj {q : ℕ} (χ : DirichletCharacter ℂ q)
(t : ℝ) (n : ℕ) :
characterTwist (χ.ringHomComp (starRingEnd ℂ)) (-t) n =
conj (characterTwist χ t n) := by
simp only [characterTwist, MulChar.ringHomComp_apply, map_mul,
← Complex.exp_conj]
congr 2
simp only [neg_mul, Complex.ofReal_neg, Complex.conj_ofReal, Complex.conj_I, mul_neg]
/-- Uniform nonpretentiousness is invariant under complex conjugation. -/
lemma UniformlyNonpretentious.conj {f : ℕ → ℂ}
(hf : UniformlyNonpretentious f) :
UniformlyNonpretentious (fun n => conj (f n)) := by
intro q hq χ K
filter_upwards [hf q hq (χ.ringHomComp (starRingEnd ℂ)) K] with N hN
intro t ht
have h := hN (-t) (by simpa using ht)
have heq : characterTwist (χ.ringHomComp (starRingEnd ℂ)) (-t) =
fun n => conj (characterTwist χ t n) := by
funext n
exact characterTwist_conj χ t n
rw [heq] at h
simpa only [← squaredDistance_conj f (fun n => conj (characterTwist χ t n)) N,
Complex.conj_conj] using h
/-- Altering a fixed finite set of prime values changes the square of the
pretentious distance by at most twice their reciprocal sum. -/
lemma squaredDistance_finite_prime_change (f b g : ℕ → ℂ) (N : ℕ)
(P : Finset ℕ) (hf : ∀ n, 0 < n → ‖f n‖ ≤ 1) (hb : ∀ n, 0 < n → ‖b n‖ ≤ 1)
(hg : ∀ n, 0 < n → ‖g n‖ ≤ 1)
(heq : ∀ p, Nat.Prime p → p ∉ P → b p = f p) :
|squaredDistance b g N - squaredDistance f g N| ≤
2 * ∑ p ∈ P, (1 : ℝ) / (p : ℝ) := by
let S := primesUpTo N
have hterm (p : ℕ) (hp : 0 < p) :
|(1 - (b p * conj (g p)).re) / (p : ℝ) -
(1 - (f p * conj (g p)).re) / (p : ℝ)| ≤ 2 / (p : ℝ) := by
have hdiff : ‖f p - b p‖ ≤ 2 := by
exact (norm_sub_le _ _).trans (by linarith [hf p hp, hb p hp])
have hprod : ‖(f p - b p) * conj (g p)‖ ≤ 2 := by
calc
‖(f p - b p) * conj (g p)‖ = ‖f p - b p‖ * ‖g p‖ := by simp
_ ≤ 2 * 1 := mul_le_mul hdiff (hg p hp) (norm_nonneg _) (by norm_num)
_ = 2 := by norm_num
have hre : |(f p * conj (g p)).re - (b p * conj (g p)).re| ≤ 2 := by
calc
|(f p * conj (g p)).re - (b p * conj (g p)).re| =
|((f p - b p) * conj (g p)).re| := by rw [sub_mul, Complex.sub_re]
_ ≤ ‖(f p - b p) * conj (g p)‖ := Complex.abs_re_le_norm _
_ ≤ 2 := hprod
rw [← sub_div, show (1 - (b p * conj (g p)).re) -
(1 - (f p * conj (g p)).re) =
(f p * conj (g p)).re - (b p * conj (g p)).re by ring,
abs_div, show |(p : ℝ)| = (p : ℝ) from abs_of_nonneg (Nat.cast_nonneg p)]
exact div_le_div_of_nonneg_right hre (Nat.cast_nonneg p)
have htermP (p : ℕ) (hp : p ∈ S) :
|(1 - (b p * conj (g p)).re) / (p : ℝ) -
(1 - (f p * conj (g p)).re) / (p : ℝ)| ≤
if p ∈ P then 2 / (p : ℝ) else 0 := by
split_ifs with hP
· exact hterm p ((Finset.mem_filter.mp hp).2.pos)
· have hprime : Nat.Prime p := (Finset.mem_filter.mp hp).2
rw [heq p hprime hP, sub_self, abs_zero]
calc
|squaredDistance b g N - squaredDistance f g N| =
|∑ p ∈ S, ((1 - (b p * conj (g p)).re) / (p : ℝ) -
(1 - (f p * conj (g p)).re) / (p : ℝ))| := by
rw [Finset.sum_sub_distrib]
rfl
_ ≤ ∑ p ∈ S, |(1 - (b p * conj (g p)).re) / (p : ℝ) -
(1 - (f p * conj (g p)).re) / (p : ℝ)| := Finset.abs_sum_le_sum_abs _ _
_ ≤ ∑ p ∈ S, if p ∈ P then 2 / (p : ℝ) else 0 := Finset.sum_le_sum htermP
_ = ∑ p ∈ S.filter (fun p => p ∈ P), 2 / (p : ℝ) := by rw [Finset.sum_filter]
_ ≤ ∑ p ∈ P, 2 / (p : ℝ) := by
apply Finset.sum_le_sum_of_subset_of_nonneg
· intro p hp
exact (Finset.mem_filter.mp hp).2
· intro p hp hnot
positivity
_ = 2 * ∑ p ∈ P, (1 : ℝ) / (p : ℝ) := by
rw [Finset.mul_sum]
apply Finset.sum_congr rfl
intro p hp
ring
/-- The bound is uniform over the changing twist and finite modifications. -/
lemma UniformlyNonpretentious.finite_prime_change {f b : ℕ → ℂ}
(hfnp : UniformlyNonpretentious f) (P : Finset ℕ)
(hf : ∀ n, 0 < n → ‖f n‖ ≤ 1) (hb : ∀ n, 0 < n → ‖b n‖ ≤ 1)
(heq : ∀ p, Nat.Prime p → p ∉ P → b p = f p) :
UniformlyNonpretentious b := by
intro q hq χ K
filter_upwards [hfnp q hq χ (K + 2 * ∑ p ∈ P, (1 : ℝ) / (p : ℝ))] with N hN
intro t ht
have hdist := squaredDistance_finite_prime_change f b (characterTwist χ t) N
P hf hb (fun n _ => characterTwist_norm_le_one χ t n) heq
have hlow := hN t ht
have := (abs_le.mp hdist).1
linarith
/-- One lower bound works for every member of a fixed finite character family,
and for every bounded sequence whose prime values are changed only in `P`.
This is the uniformity needed when taking the short-interval theorem at a fixed
auxiliary graph scale. -/
lemma UniformlyNonpretentious.finite_family {ι : Type*} [Finite ι]
{f : ℕ → ℂ} (hfnp : UniformlyNonpretentious f) (hf : ∀ n, 0 < n → ‖f n‖ ≤ 1)
(P : Finset ℕ) (q : ι → ℕ) (hq : ∀ i, 0 < q i)
(χ : ∀ i, DirichletCharacter ℂ (q i)) (K : ℝ) :
∀ᶠ N : ℕ in atTop, ∀ i (b : ℕ → ℂ),
(∀ n, 0 < n → ‖b n‖ ≤ 1) →
(∀ p, Nat.Prime p → p ∉ P → b p = f p) →
∀ t : ℝ, |t| ≤ (N : ℝ) → K ≤ squaredDistance b (characterTwist (χ i) t) N := by
have hfamily : ∀ᶠ N : ℕ in atTop, ∀ i,
∀ t : ℝ, |t| ≤ (N : ℝ) →
K + 2 * ∑ p ∈ P, (1 : ℝ) / (p : ℝ) ≤
squaredDistance f (characterTwist (χ i) t) N := by
rw [Filter.eventually_all]
intro i
exact hfnp (q i) (hq i) (χ i) _
filter_upwards [hfamily] with N hN
intro i b hb heq t ht
have hdist := squaredDistance_finite_prime_change f b (characterTwist (χ i) t) N
P hf hb (fun n _ => characterTwist_norm_le_one (χ i) t n) heq
have hlow := hN i t ht
have := (abs_le.mp hdist).1
linarith
end TwoPointCorrelations
end OAI