Documentation

LeanPool.HighContrastHomogenization.Provider.Response.ProfileRecentCellLp

High-contrast homogenization: Provider.Response.ProfileRecentCellLp #

Imported from the Apache-2.0 HighContrastHomogenization development at commit 7a13dbcd8d6609264a713373f5c69ceeac870472.

The probabilistic recent-cell bound #

The centered maximum has exactly the terminal centered history as its Q-moment. Monotonicity of probability-space L^p seminorms then turns the pointwise recent-cell estimate into its printed L² bound.

The centered terminal maximum is an a.e. measurable statistic.

theorem HCPolySupport.HighContrast.Response.eLpNorm_profileCenteredMaximum_eq {d : ℕ} {P : MeasureTheory.Measure (CoeffSpace d)} {q : Mat d} {rhoMax Q : ℝ} (hQ : 0 < Q) (jStar t : ℤ) (hmeas : MeasureTheory.AEStronglyMeasurable (profileCenteredMaximum P rhoMax q jStar t) P) :
MeasureTheory.eLpNorm (profileCenteredMaximum P rhoMax q jStar t) (ENNReal.ofReal Q) P = centeredHistory P Q rhoMax q jStar t ^ Q⁻¹

The L^Q seminorm of the centered maximum is the Q-th root of its defining history.

On a probability space the centered maximum has L² seminorm at most the Q-th root of the centered history.

Multiplication by a finite nonnegative scalar costs at most that scalar in an extended-real L^p seminorm.