Linfty almost periodicity
AlmostPeriodicity.linfty_almost_periodicity
Project documentation
An almost-periodicity theorem. Under the small-growth hypothesis , and for nonempty finite sets , there is a set of translations with . Every changes the normalized convolution by at most in .
Exact Lean statement
public theorem linfty_almost_periodicity (ε : ℝ) (hε₀ : 0 < ε) (hε₁ : ε ≤ 1) (hK₂ : 2 ≤ K)
(hK : σ[A, S] ≤ K) (B C : Finset G) (hB : B.Nonempty) (hC : C.Nonempty) :
∃ T : Finset G,
K ^ (-4096 * ⌈𝓛 (#C / #B)⌉ / ε ^ 2) * #S ≤ #T ∧
∀ t ∈ T, ‖τ t (μ_[ℂ] A ∗ᵈ 𝟭_[B] ∗ᵈ μ C) - μ A ∗ᵈ 𝟭_[B] ∗ᵈ μ C‖_[∞] ≤ εFormal artifact
Lean source
public theorem linfty_almost_periodicity (ε : ℝ) (hε₀ : 0 < ε) (hε₁ : ε ≤ 1) (hK₂ : 2 ≤ K) (hK : σ[A, S] ≤ K) (B C : Finset G) (hB : B.Nonempty) (hC : C.Nonempty) : ∃ T : Finset G, K ^ (-4096 * ⌈𝓛 (#C / #B)⌉ / ε ^ 2) * #S ≤ #T ∧ ∀ t ∈ T, ‖τ t (μ_[ℂ] A ∗ᵈ 𝟭_[B] ∗ᵈ μ C) - μ A ∗ᵈ 𝟭_[B] ∗ᵈ μ C‖_[∞] ≤ ε := by let r : ℝ := min 1 (#C / #B) set m : ℝ := 𝓛 (#C / #B) have hm₀ : 0 < m := curlog_pos (by positivity) have hm₁ : 1 ≤ ⌈m⌉₊ := Nat.one_le_iff_ne_zero.2 <| by positivity obtain ⟨T, hKT, hT⟩ := almost_periodicity (ε / exp 1) (by positivity) (div_le_one_of_le₀ (hε₁.trans <| one_le_exp zero_le_one) <| by positivity) ⌈m⌉₊ (𝟭_[B]) hK₂ hK norm_cast at hT set M : ℕ := 2 * ⌈m⌉₊ have hM₀ : (M : ℝ≥0) ≠ 0 := by positivity have hM₁ : 1 < (M : ℝ≥0) := by norm_cast; simp [← Nat.succ_le_iff, M]; linarith have hM : (M : ℝ≥0).HolderConjugate _ := NNReal.HolderConjugate.conjExponent hM₁ have : (M : ℝ≥0∞).HolderConjugate _ := hM.coe_ennreal refine ⟨T, ?_, fun t ht ↦ ?_⟩ · calc _ = K ^(-(512 * 8) / ε ^ 2 * ⌈m⌉₊) * #S := by rw [mul_div_right_comm, natCast_ceil_eq_intCast_ceil hm₀.le]; norm_num _ ≤ K ^(-(512 * exp 1 ^ 2) / ε ^ 2 * ⌈m⌉₊) * #S := by gcongr · exact one_le_two.trans hK₂ calc _ ≤ (2.7182818286 : ℝ) ^ 2 := by gcongr; exact exp_one_lt_d9.le _ ≤ _ := by norm_num _ = _ := by simp [div_div_eq_mul_div, ← mul_div_right_comm, mul_right_comm, div_pow] _ ≤ _ := hKT set F : G → ℂ := τ t (μ A ∗ᵈ 𝟭_[B]) - μ A ∗ᵈ 𝟭_[B] have (x : G) := calc (τ t (μ A ∗ᵈ 𝟭_[B] ∗ᵈ μ C) - μ A ∗ᵈ 𝟭_[B] ∗ᵈ μ C : G → ℂ) x = (F ∗ᵈ μ C) x := by simp [sub_ddconv, F] _ = ∑ y, F y * μ C (x - y) := ddconv_eq_sum_sub' .. _ = ∑ y, F y * μ (x +ᵥ -C) y := by simp [neg_add_eq_sub] rw [MeasureTheory.dLinftyNorm_eq_iSup_norm] refine ciSup_le fun x ↦ ?_ calc ‖(τ t (μ A ∗ᵈ 𝟭_[B] ∗ᵈ μ C) - μ A ∗ᵈ 𝟭_[B] ∗ᵈ μ C : G → ℂ) x‖ = ‖∑ y, F y * μ (x +ᵥ -C) y‖ := by rw [this] _ ≤ ∑ y, ‖F y * μ (x +ᵥ -C) y‖ := norm_sum_le _ _ _ = ‖F * μ (x +ᵥ -C)‖_[1] := by rw [MeasureTheory.dL1Norm_eq_sum_norm]; rfl _ ≤ ‖F‖_[M] * ‖μ_[ℂ] (x +ᵥ -C)‖_[NNReal.conjExponent M] := MeasureTheory.dLpNorm_mul_le _ _ _ ≤ ε / exp 1 * #B ^ (M : ℝ)⁻¹ * ‖μ_[ℂ] (x +ᵥ -C)‖_[NNReal.conjExponent M] := by gcongr simpa [← ENNReal.coe_natCast, MeasureTheory.dLpNorm_indicator_one hM₀, F] using hT _ ht _ = ε * ((#C / #B) ^ (-(M : ℝ)⁻¹) / exp 1) := by rw [← mul_comm_div, MeasureTheory.dLpNorm_mu hM.symm.lt.le hC.neg.vadd_finset, card_vadd_finset, card_neg, hM.symm.coe.inv_sub_one, div_rpow, mul_assoc] any_goals positivity push_cast rw [rpow_neg, rpow_neg, ← div_eq_mul_inv, inv_div_inv] all_goals positivity _ ≤ ε := mul_le_of_le_one_right (by positivity) <| (div_le_one <| by positivity).2 ?_ calc (#C / #B : ℝ) ^ (-(M : ℝ)⁻¹) ≤ r ^ (-(M : ℝ)⁻¹) := rpow_le_rpow_of_nonpos (by positivity) inf_le_right <| neg_nonpos.2 <| by positivity _ ≤ r ^ (-(1 + log r⁻¹)⁻¹) := rpow_le_rpow_of_exponent_ge (by positivity) inf_le_left <| neg_le_neg <| inv_anti₀ (by positivity) <| (Nat.le_ceil _).trans <| mod_cast Nat.le_mul_of_pos_left _ (by positivity) _ ≤ r ^ (-(0 + log r⁻¹)⁻¹) := by obtain hr | hr : r = 1 ∨ r < 1 := inf_le_left.eq_or_lt · simp [hr] have : 0 < log r⁻¹ := log_pos <| (one_lt_inv₀ (by positivity)).2 hr exact rpow_le_rpow_of_exponent_ge (by positivity) inf_le_left (by gcongr; exact zero_le_one) _ = r ^ (log r)⁻¹ := by simp [inv_neg] _ ≤ exp 1 := rpow_inv_log_le_exp_one- Project
- Arithmetic Progressions Almost Periodicity
- License
- Apache-2.0
- Commit
- afafc42a5326
- Source
- APAP/Physics/AlmostPeriodicity.lean:424-493
Reuse this declaration
Bring the exact result into your workflow
The import identifies the source module. Your project still needs the pinned package dependency shown on this page.
What this badge means
This completion status comes from the project or community source. It has not yet been represented here as an independent rebuild and axiom audit.
Continue in this project
Related declarations
Add Dissociated boring Energy le
AddDissociated.boringEnergy_le
Project documentation
If a finite set is additively dissociated, then its order- additive energy is at most , where is the project's Chang constant. This is the quantitative dissociated-set estimate used in the proof of Chang's lemma.
Source project: Arithmetic Progressions Almost Periodicity
Person-level attribution pending.
Linfty almost periodicity boosted
AlmostPeriodicity.linfty_almost_periodicity_boosted
Plain-language statement
A boosted almost-periodicity estimate. Under , it finds a large set , with the stated lower bound , such that averaging the target convolution against the -fold convolution of changes it by at most in .
Source project: Arithmetic Progressions Almost Periodicity
Person-level attribution pending.
Ap in ff
ap_in_ff
Project documentation
A finite-field approximation lemma. If and each have density at least , then for any test set and there is a subspace of explicitly bounded codimension such that smoothing by the uniform measure on changes its total mass on by at most .
Source project: Arithmetic Progressions Almost Periodicity
Person-level attribution pending.