OAI/Analysis/PlaneSpectrum/PhaseCollision.lean

GitHub source file

19 declarations, 0 contain sorry, 0 declared as axioms

  1. def OAI.EuclideanFiveColor.PhaseCollision.repeatedPairs line 26

    Source documentation

    Ordered distinct pairs in one assigned bin.

    def repeatedPairs (s : Finset ι) (f : ι → κ) : Finset (ι × ι) := by
    …
  2. lemma OAI.EuclideanFiveColor.PhaseCollision.sameBin_card line 31
  3. lemma OAI.EuclideanFiveColor.PhaseCollision.sameBin_card_eq line 48
  4. theorem OAI.EuclideanFiveColor.PhaseCollision.repeated_bin_lower_bound line 68
  5. def OAI.EuclideanFiveColor.PhaseCollision.closePairs line 87

    Source documentation

    Ordered distinct phase pairs within a fixed distance.

    def closePairs (v : ι → V) (r : ℝ) : Finset (ι × ι) := by
    …
  6. theorem OAI.EuclideanFiveColor.PhaseCollision.close_pair_lower_bound line 91
  7. theorem OAI.EuclideanFiveColor.PhaseCollision.normalized_close_pair_lower_bound line 119

    Source documentation

    The normalized source estimate q² / #bins - 1/#indices.

    theorem normalized_close_pair_lower_bound [Nonempty ι] [Nonempty κ]
        (s : Finset ι) (f : ι → κ) (v : ι → V) (w : κ → V) (r : ℝ)
        (h : ∀ i ∈ s, dist (v i) (w (f i)) ≤ r) :
        ((s.card : ℝ) / Fintype.card ι) ^ 2 / Fintype.card κ - 1 / Fintype.card ι ≤
          (closePairs v (2 * r)).card / (Fintype.card ι : ℝ) ^ 2 := by
    …

    Mentions: OAI.EuclideanFiveColor.PhaseCollision.closePairs

  8. theorem OAI.EuclideanFiveColor.PhaseCollision.average_cover_forces_collision line 142

    Source documentation

    Larger than 95 percent average phase coverage forces a colliding pair. The bins and assignments may depend on the sampled point; no measurable choice of assignments is required.

    theorem average_cover_forces_collision (μ : Measure Ω) [IsProbabilityMeasure μ]
        (s : Ω → Finset ι) (f : Ω → ι → κ) (v : Ω → ι → V) (w : Ω → κ → V)
        (r : ℝ)
        (hcov : ∀ᵐ x ∂μ, ∀ i ∈ s x, dist (v x i) (w x (f x i)) ≤ r)
        (hm : ∀ i j, Measurable (fun x => dist (v x i) (v x j)))
        (hq : Measurable (fun x => ((s x).card : ℝ) / Fintype.card ι))
        (hlarge : 10 * Fintype.card κ < Fintype.card ι)
        (havg : (19/20 : ℝ) < ∫ x, ((s x).card : ℝ) / Fintype.card ι ∂μ) :
        ∃ i j : ι, i ≠ j ∧
          1 / (4 * (Fintype.card κ : ℝ)) ≤ μ.real {x | dist (v x i) (v x j) ≤ 2*r} := by
    …
  9. def OAI.EuclideanFiveColor.AverageErrors.goodFraction line 259

    Source documentation

    The fraction of indices whose nonnegative error is below a threshold.

    def goodFraction (e : ι → Ω → ℝ) (t : ℝ) (x : Ω) : ℝ := by
    …
  10. lemma OAI.EuclideanFiveColor.AverageErrors.goodFraction_eq_sum line 264
  11. lemma OAI.EuclideanFiveColor.AverageErrors.measurable_goodFraction line 270
  12. lemma OAI.EuclideanFiveColor.AverageErrors.goodFraction_nonneg line 279
  13. lemma OAI.EuclideanFiveColor.AverageErrors.goodFraction_le_one line 283
  14. lemma OAI.EuclideanFiveColor.AverageErrors.integrable_goodFraction line 289
  15. theorem OAI.EuclideanFiveColor.AverageErrors.average_goodFraction line 298

    Source documentation

    Less than 5% average error gives more than 95% good indices on average.

    theorem average_goodFraction (μ : Measure Ω) [IsProbabilityMeasure μ]
        (e : ι → Ω → ℝ) (t : ℝ) (ht : 0 < t)
        (he : ∀ i, Measurable (e i)) (hi : ∀ i, Integrable (e i) μ)
        (hn : ∀ᵐ x ∂μ, ∀ i, 0 ≤ e i x)
        (havg : ∀ i, ∫ x, e i x ∂μ < t/20) :
        (19/20 : ℝ) < ∫ x, goodFraction e t x ∂μ := by
    …

    Mentions: OAI.EuclideanFiveColor.AverageErrors.goodFraction

  16. theorem OAI.EuclideanFiveColor.PhaseCollision.average_exists_cover_forces_collision line 352

    Source documentation

    The assignment itself need not be measurable and may be chosen separately at every sample. Only the phase distances and fraction are measured.

    theorem average_exists_cover_forces_collision (μ : Measure Ω) [IsProbabilityMeasure μ]
        (s : Ω → Finset ι) (v : Ω → ι → V) (w : Ω → κ → V) (r : ℝ)
        (hcov : ∀ᵐ x ∂μ, ∀ i ∈ s x, ∃ j, dist (v x i) (w x j) ≤ r)
        (hm : ∀ i j, Measurable (fun x => dist (v x i) (v x j)))
        (hq : Measurable (fun x => ((s x).card : ℝ) / Fintype.card ι))
        (hlarge : 10 * Fintype.card κ < Fintype.card ι)
        (havg : (19/20 : ℝ) < ∫ x, ((s x).card : ℝ) / Fintype.card ι ∂μ) :
        ∃ i j : ι, i ≠ j ∧
          1 / (4 * (Fintype.card κ : ℝ)) ≤ μ.real {x | dist (v x i) (v x j) ≤ 2*r} := by
    …
  17. lemma OAI.EuclideanFiveColor.PhaseCollision.circle_product_dist line 373
  18. theorem OAI.EuclideanFiveColor.EncodedCollision.collision_from_close_encodings line 396

    Source documentation

    This is the source's collision contradiction on one selected predecessor fiber. The lists need not be measurably enumerated. Measurability is required only of their symmetric encodings and the actual sampled phases.

    theorem collision_from_close_encodings (μ : Measure Ω) [IsProbabilityMeasure μ]
        (m : ℕ) (hm : 0 < m) (ρ : ℝ) (hρ : 0 < ρ)
        (zp zm : Ω → ι → Fin m → Circle) (ap am : ι → Ω → Circle) (j₀ : ι)
        (hap : ∀ j, Measurable (ap j)) (ham : ∀ j, Measurable (am j))
        (hDp : ∀ j, Measurable (fun x => dist (encoding (zp x j)) (encoding (zp x j₀))))
        (hDm : ∀ j, Measurable (fun x => dist (encoding (zm x j)) (encoding (zm x j₀))))
        (hmem : ∀ᵐ x ∂μ, ∀ j, (∃ r, ap j x = zp x j r) ∧ (∃ r, am j x = zm x j r))
        (hi : ∀ j, Integrable (fun x =>
           dist (encoding (zp x j)) (encoding (zp x j₀))^2 +
           dist (encoding (zm x j)) (encoding (zm x j₀))^2) μ)
        (he : ∀ j, (∫ x,
           dist (encoding (zp x j)) (encoding (zp x j₀))^2 +
    …
  19. theorem OAI.EuclideanFiveColor.EncodedCollision.collision_from_measurable_encodings line 463

    Source documentation

    This is the source's collision contradiction on one selected predecessor fiber. The lists need not be measurably enumerated. Measurability is required only of their symmetric encodings and the actual sampled phases.

    theorem collision_from_measurable_encodings (μ : Measure Ω) [IsProbabilityMeasure μ]
        (m : ℕ) (hm : 0 < m) (ρ : ℝ) (hρ : 0 < ρ)
        (zp zm : Ω → ι → Fin m → Circle) (ap am : ι → Ω → Circle) (j₀ : ι)
        (Hp Hm : ι → Ω → Fin m → ℂ)
        (henc : ∀ᵐ x ∂μ, ∀ j, encoding (zp x j) = Hp j x ∧ encoding (zm x j) = Hm j x)
        (hap : ∀ j, Measurable (ap j)) (ham : ∀ j, Measurable (am j))
        (hDp : ∀ j, Measurable (fun x => dist (Hp j x) (Hp j₀ x)))
        (hDm : ∀ j, Measurable (fun x => dist (Hm j x) (Hm j₀ x)))
        (hmem : ∀ᵐ x ∂μ, ∀ j, (∃ r, ap j x = zp x j r) ∧ (∃ r, am j x = zm x j r))
        (hi : ∀ j, Integrable (fun x =>
           dist (Hp j x) (Hp j₀ x)^2 +
           dist (Hm j x) (Hm j₀ x)^2) μ)
    …