Documentation

LeanPool.Superdiffusion.Algsuperdiff.Section5.Support.AuxiliaryScale

AuxiliaryScale #

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

The auxiliary scale of the displacement tail bound #

The tail bound on the displacement compares the maximal scale m with an auxiliary scale n, chosen so that 3^n straddles

max{ν t / (δ 3^m), (c⋆ γ^{-1})^{1/2} (t / (δ 3^m))^{1/(1-γ)}},

and the chains-of-good-cubes estimate is then applied at that n. Its hypothesis n ≤ m - Y_m(ε) is exactly what the minimal-scale condition

min{3^{2m} / (ν t), (c⋆^{-1} γ)^{1/2} ((3^m)^{2-γ} / t)^{1/(1-γ)}} ≥ S_m

delivers, through the definition S_m = δ^{-1/(1-γ)} 3^{Y_m(ε)}.

The two halves of the comparison behave differently. On the first term the powers of δ do not cancel and the argument uses δ ≤ 1: what is left over is δ^{1 - 1/(1-γ)} ≥ 1, an inequality that fails for δ > 1. On the second term every factor cancels exactly — the powers of c⋆ and γ because the two expressions are reciprocal, the powers of δ because the exponent 1/(1-γ) is the same on both sides, and the powers of three because ((3^m)^{1-γ})^{1/(1-γ)} = 3^m. This is the bookkeeping that fixes the exponent -1/(1-γ) of δ in S_m.

Main definitions #

References #

1. The two quantities #

noncomputable def Algsuperdiff.Section5.Support.auxiliaryScaleBound (nu cstar gamma delta t : ℝ) (m : ℤ) :

The quantity the auxiliary scale straddles, max{ν t / (δ 3^m), (c⋆ γ^{-1})^{1/2} (t / (δ 3^m))^{1/(1-γ)}}.

Equations
Instances For
    noncomputable def Algsuperdiff.Section5.Support.displacementMinScale (nu cstar gamma t : ℝ) (m : ℤ) :

    The left side of the minimal-scale condition, min{3^{2m} / (ν t), (c⋆^{-1} γ)^{1/2} ((3^m)^{2-γ} / t)^{1/(1-γ)}}.

    Equations
    Instances For
      theorem Algsuperdiff.Section5.Support.auxiliaryScaleBound_pos {nu cstar gamma delta t : ℝ} (hnu : 0 < nu) (hdelta : 0 < delta) (ht : 0 < t) (m : ℤ) :
      0 < auxiliaryScaleBound nu cstar gamma delta t m

      2. The choice of the auxiliary scale #

      theorem Algsuperdiff.Section5.Support.exists_auxiliaryScale {v : ℝ} (hv : 0 < v) :
      ∃ (n : ℤ), 3 ^ n ≤ v ∧ v < 3 ^ (n + 1)

      Every positive quantity is straddled by two consecutive powers of three. The base-three integer logarithm is such a scale, which is the choice the displacement tail bound makes.

      3. The comparison with the scale of the chains estimate #

      theorem Algsuperdiff.Section5.Support.auxiliaryScale_le_sub_of_displacementScale_le {nu cstar gamma delta t : ℝ} {m : ℤ} {N : ℕ} {n : ℤ} (hnu : 0 < nu) (ht : 0 < t) (hdelta : 0 < delta) (hdelta1 : delta ≤ 1) (hcstar : 0 < cstar) (hgamma : 0 < gamma) (hgamma1 : gamma < 1) (hn : 3 ^ n ≤ auxiliaryScaleBound nu cstar gamma delta t m) (hmin : displacementScale delta gamma N ≤ displacementMinScale nu cstar gamma t m) :
      n ≤ m - ↑N

      The minimal-scale condition supplies the hypothesis of the chains estimate. If 3^n is below the straddled quantity and the scale S_m = δ^{-1/(1-γ)} 3^N is below the minimal-scale quantity, then n ≤ m - N.