math / data /lean /OAI /NumberTheory /Ostmann /Arithmetic /ArithmeticPrimeNormRate.lean
introvoyz045's picture
Migrated from GitHub
4a56cd6 verified
Raw History Blame Contribute Delete
3.86 kB
import OAI.NumberTheory.Ostmann.Arithmetic.ArithmeticNormRate
import OAI.NumberTheory.Ostmann.Characters.FourierWindowDecay
import OAI.NumberTheory.Ostmann.Construction.SpectatorBulkScale
namespace OAI
/-! # The frequency budget after return to the original prime law -/
namespace Ostmann
open Filter
open scoped SchwartzMap
theorem arithmetic_prime_page_spectator_cost (n m : β„•) :
((2 : ℝ) ^ (2 ^ n * m) * 2 * 3 ^ (2 ^ n * m)) * 2 ^ (2 ^ n * m) =
2 * Real.exp (Real.log 12 * (2 ^ n : β„•) * m) := by
have hbase : (2 : ℝ) * 3 * 2 = 12 := by norm_num
calc
_ = 2 * (((2 : ℝ) * 3 * 2) ^ (2 ^ n * m)) := by
simp only [mul_pow]
ring
_ = _ := by
rw [hbase]
congr 1
rw [show Real.log 12 * (2 ^ n : β„•) * m =
((2 ^ n * m : β„•) : ℝ) * Real.log 12 by push_cast; ring,
Real.exp_nat_mul, Real.exp_log (by norm_num : (0 : ℝ) < 12)]
/-- The prime-cell mixture adds one factor two per bulk coordinate. This
changes the constant in the exponential from log six to log twelve. -/
theorem arithmetic_prime_norm_budget_rate (n : β„•) (A C Ξ΅ : ℝ)
(hA : 0 ≀ A) (hC : 0 ≀ C) (hΞ΅ : 0 < Ξ΅) :
βˆ€αΆ  m : ℝ in atTop, βˆƒ D : ℝ, 0 ≀ D ∧
(βˆ€ q : β„•, q β‰  0 β†’ (q : ℝ) ≀ Real.exp (2 * C * m) β†’ (q.divisors.card : ℝ) ≀ D) ∧
βˆ€ N V : β„•, βˆ€ Ξ” : ℝ, (N : ℝ) ≀ Real.exp (C * m) β†’
(V : ℝ) ≀ Real.exp (Ξ” + Real.sqrt (4 * m)) β†’
(A * Real.exp (-(2 ^ n : β„•) * Ξ”)) *
((2 * Real.exp (Real.log 12 * (2 ^ n : β„•) * m)) *
((8 * D ^ 4 * (1 + Real.log N) ^ 3) ^ (2 ^ n - 1) *
(2 * (V : ℝ)) ^ (2 * 2 ^ n))) ≀
Real.exp ((2 ^ n : β„•) * Ξ” + Real.log 12 * (2 ^ n : β„•) * m + Ξ΅ * m) := by
have hlog : Real.log 12 = Real.log 2 + Real.log 6 := by
rw [← Real.log_mul (by norm_num : (2 : ℝ) β‰  0) (by norm_num : (6 : ℝ) β‰  0)]
norm_num
filter_upwards [arithmetic_norm_budget_rate n A C Ξ΅ hA hC hΞ΅] with m hm
obtain ⟨D, hD, hdiv, hcost⟩ := hm
refine ⟨D, hD, hdiv, ?_⟩
intro N V Ξ” hN hV
have h := mul_le_mul_of_nonneg_left (hcost N V Ξ” hN hV)
(Real.exp_pos (Real.log 2 * (2 ^ n : β„•) * m)).le
convert h using 1 <;> rw [hlog] <;>
simp only [add_mul, Real.exp_add] <;> ring
/-- Specialization to the actual even bulk size, with the literal window and
Page/spectator constants of the prime norm estimate. -/
theorem arithmetic_prime_fourier_budget_rate (ψ : 𝓒(ℝ, β„‚)) (n k : β„•) (hk : 0 < k)
(B C Ξ΅ : ℝ) (hC : 0 ≀ C) (hΞ΅ : 0 < Ξ΅) :
βˆ€αΆ  L : ℝ in atTop, let m := spectatorBulkCount k L
βˆƒ D : ℝ, 0 ≀ D ∧
(βˆ€ q : β„•, q β‰  0 β†’ (q : ℝ) ≀ Real.exp (2 * C * m) β†’ (q.divisors.card : ℝ) ≀ D) ∧
βˆ€ N V : β„•, βˆ€ Ξ” : ℝ, (N : ℝ) ≀ Real.exp (C * m) β†’
(V : ℝ) ≀ Real.exp (Ξ” + Real.sqrt (4 * m)) β†’
(((SchwartzMap.seminorm ℝ 0 0 ψ / Real.sqrt (Real.exp Ξ”)) ^ (2 ^ n) *
B ^ (2 ^ n - 1)) ^ 2 *
(((2 : ℝ) ^ (2 ^ n * m) * 2 * 3 ^ (2 ^ n * m)) *
((8 * D ^ 4 * (1 + Real.log N) ^ 3) ^ (2 ^ n - 1) *
(2 * (V : ℝ)) ^ (2 * 2 ^ n)))) * 2 ^ (2 ^ n * m) ≀
Real.exp ((2 ^ n : β„•) * Ξ” + Real.log 12 * (2 ^ n : β„•) * m + Ξ΅ * m) := by
let A : ℝ := ((SchwartzMap.seminorm ℝ 0 0 ψ) ^ (2 ^ n) * B ^ (2 ^ n - 1)) ^ 2
have hrate := (spectatorBulkCount_tendsto k hk).eventually
(arithmetic_prime_norm_budget_rate n A C Ξ΅ (sq_nonneg _) hC hΞ΅)
filter_upwards [hrate] with L hrate
obtain ⟨D, hD, hdiv, hcost⟩ := hrate
refine ⟨D, hD, hdiv, ?_⟩
intro N V Ξ” hN hV
have h := hcost N V Ξ” hN hV
rw [← arithmetic_prime_page_spectator_cost n (spectatorBulkCount k L)] at h
rw [fourier_window_pair_exp]
exact le_trans (le_of_eq (by dsimp only [A]; ring)) h
end Ostmann
end OAI