OAI/Analysis/LipschitzHilbert/Trees/SourceWeightedBadPath.lean

GitHub source file

3 declarations, 0 contain sorry, 0 declared as axioms

A reference-interval tail estimate retains the complete decaying top weight. No restriction to the central part of the top is used.

  1. theorem OAI.LipschitzHilbert.weighted_small_set_le line 15
  2. theorem OAI.LipschitzHilbert.source_line_weighted_bad line 48

    Source documentation

    On every interval of top length, an unweighted relative tail becomes an equally small weighted tail, up to the fixed weight comparison factor.

    theorem source_line_weighted_bad (T : SlopeBox) (origin y H : ℝ) {K D δ : ℝ}
        (hK : 1 ≤ K) (hD : 0 ≤ D) (hH : 0 < H) (hHT : H ≤ T.length)
        (hδ : 0 ≤ δ) (bad : Set ℝ) (hbad : MeasurableSet bad)
        (hsmall : (volume.restrict (Ico origin (origin+H))).real bad ≤ δ*H) :
        IntegrableOn (bad.indicator (fun x => T.weight K D (x,y+T.slope*x)))
          (Ico origin (origin+H)) volume ∧
          (∫ x in Ico origin (origin+H), bad.indicator
            (fun x => T.weight K D (x,y+T.slope*x)) x) ≤
            (Real.exp (2*D)*δ)*(∫ x in Ico origin (origin+H), T.weight K D (x,y+T.slope*x)) := by
    …

    Mentions: OAI.LipschitzHilbert.SlopeBox

  3. theorem OAI.LipschitzHilbert.source_weighted_bad_path_area line 79

    Source documentation

    Summing the line estimates over all tiles and then all tops retains exactly the total top area and the full spatial weight.

    theorem source_weighted_bad_path_area {ι : Type*} (tops : Finset ι)
        (T : ι → SlopeBox) (bad : ι → Set (ℝ × ℝ))
        (hbad : ∀ t ∈ tops, MeasurableSet (bad t)) {K D δ : ℝ}
        (hK : 1 ≤ K) (hD : 2 < D) (hδ : 0 ≤ δ)
        (htail : ∀ t ∈ tops, ∀ᵐ y : ℝ ∂volume, ∀ k : ℤ,
          (volume.restrict (sourceLineTile (T t).length k)).real
            {x | (x,y+(T t).slope*x) ∈ bad t} ≤ δ*(T t).length) :
        Integrable (fun z => ∑ t ∈ tops, (bad t).indicator ((T t).weight K D) z)
          (volume.prod volume) ∧
          (∫ z, ∑ t ∈ tops, (bad t).indicator ((T t).weight K D) z ∂volume.prod volume) ≤
            (((Real.exp (2*D)*δ)*(2^D*∫ z, topProfile D z ∂volume.prod volume))*K^2)*
              ∑ t ∈ tops, (T t).length*(T t).width := by
    …

    Mentions: OAI.LipschitzHilbert.SlopeBox, OAI.LipschitzHilbert.sourceLineTile, OAI.LipschitzHilbert.topProfile