Documentation

LeanPool.Superdiffusion.Algsuperdiff.Section24.Sensitivity.Provider.Lambda.PathGronwall

PathGronwall #

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

The pure-flux response along the canonical perturbation path #

Source: ABK26. The conditional lambda_{s,q} sensitivity estimate compounds a per-cube estimate along the segment a + t h, t ∈ [0,1]. This module supplies the two facts about that segment that the compounding needs, at a general Chapter 2 domain and for the pure-flux loading p = 0:

Nothing in this module asserts differentiability of any multiscale aggregate; only the single-cube scalar response is differentiated, exactly as required by the aggregation argument.

Every declaration is an internal helper for the Section 2.4 sensitivity providers.

The perturbation path is a semigroup #

The D_h flux form #

The D_h quadratic form at the pure-flux loading P = (0, -q). This is the bottom-right entry of D_h tested against q, i.e. the slope of the pure-flux response along the perturbation path.

Equations
  • One or more equations did not get rendered due to their size.
Instances For

    Differentiability of the pure-flux response along the whole path #

    The pure-flux response is differentiable at every point of the perturbation path, with derivative one half of the D_h flux form at the current base point. This is the derivative theorem transported along the path semigroup.

    The Grönwall step #

    theorem Algsuperdiff.Section24.Sensitivity.Provider.Lambda.le_exp_mul_of_hasDerivAt_le {y y' : ℝ → ℝ} {K t₁ t₂ : ℝ} (hderiv : ∀ (τ : ℝ), HasDerivAt y (y' τ) τ) (hbound : ∀ (τ : ℝ), t₁ ≤ τ → τ ≤ t₂ → y' τ ≤ K * y τ) (ht : t₁ ≤ t₂) :
    y t₂ ≤ Real.exp (K * (t₂ - t₁)) * y t₁

    Grönwall on an interval. A nonnegative differentiable function whose derivative is bounded by K times its value grows at most exponentially.