Documentation

LeanPool.Superdiffusion.Algsuperdiff.Section4.Provider.Regularity.StepSixBoundaryEndpoint

StepSixBoundaryEndpoint #

Ported from scottnarmstrong/Superdiffusion at 7f20476385d4af418f110303085ad0740c9eba46. Module imports, support namespace and local option directives adapted for Lean Pool.

1. The composition, with the flat leg held out #

theorem Algsuperdiff.Section4.Provider.Regularity.holderBoundary_compose {C E R A oscLo oscHi dataG dataHs Fl : ℝ} (hC : 0 ≤ C) (hE : 0 ≤ E) (hER : E ≤ R) (hoscHi : 0 ≤ oscHi) (hdataG : 0 ≤ dataG) (hdataHs : 0 ≤ dataHs) (habs : E * Fl ≤ A) (hiter : oscLo ≤ C * E * (oscHi + C * dataG) + C * E * (dataHs + Fl)) :
oscLo ≤ (C + C * C) * R * (oscHi + dataG + dataHs) + C * A

Step 6 with a held-out flat leg. The prefactor is replaced by R on the geometric part, while the flat part Fl is absorbed separately (habs), because absorption consumes the exponential itself.

2. The print's boundary Step 6: the flat leg ε-funded, α-free #

theorem Algsuperdiff.Section4.Provider.Regularity.stepSixBoundary_holderBound_epsFunded (d : ℕ) {Cedos Cann Citer C alpha : ℝ} {k : ℕ} {n m : ℤ} {oscLo oscHi dataG Khleg Hinf ind Se : ℝ} (hC : 0 ≤ C) (hCiter : 0 ≤ Citer) (halpha : alpha ≤ 1) (hnm : n ≤ m) (hoscHi : 0 ≤ oscHi) (hdataG : 0 ≤ dataG) (hKhleg : 0 ≤ Khleg) (hHinf : 0 ≤ Hinf) (hind : 0 ≤ ind) (hSe0 : 0 ≤ Se) (hSe : Se ≤ 2 * stepOneDelta (stepOneC1 d Cedos Cann Citer k) alpha * (↑m - ↑n)) (hiter : oscLo ≤ C * Real.exp ((stepOneC1 d Cedos Cann Citer k)⁻¹ * Citer * stepSixExponent alpha n m) * (oscHi + C * dataG) + C * Real.exp ((stepOneC1 d Cedos Cann Citer k)⁻¹ * Citer * stepSixExponent alpha n m) * ((Khleg + Se * Hinf) * ind)) :
oscLo ≤ (C + C * C) * Real.rpow 3 (1 / 2 * stepSixExponent alpha n m) * (oscHi + dataG + Khleg * ind) + C * (4 / Real.log 3 * Real.rpow 3 (1 / 2 * stepSixExponent alpha n m) * (Hinf * ind))

e.oscillation.Holder.bound on the boundary branch, as the print builds it.

The flat ∇h leg enters δ_j with the coefficient ε_j, so its window sum is S_ε ≤ 2 δ (m-n) = 2 C₁^{-1}(1-α)(m-n); the absorption t e^t ≤ e^{2t} then delivers it at the printed exponent with the ABSOLUTE constant 4/log 3:

   oscLo ≤ (C+C²) 3^{(1/2)(1-α)(m-n)} ( oscHi + dataG + K_h-leg · 1 )
             + C (4/log 3) 3^{(1/2)(1-α)(m-n)} ( ‖∇h‖_{L^∞} · 1 ) .

No α, γ or δ survives in either constant.