Documentation

Verification.RowConvexConvergence

← Mathematical handbook

Continuous tests of partition-averaged conditional distributions #

theorem Verification.continuous_unit_affine_modulus (φ : ℝ → ℝ) (hc : Continuous φ) {ε : ℝ} (hε : 0 < ε) :
∃ (K : ℝ), 0 ≤ K ∧ ∀ x ∈ Set.Icc 0 1, ∀ y ∈ Set.Icc 0 1, |φ x - φ y| ≤ ε + K * |x - y|

A continuous test on the unit interval has an affine error modulus with arbitrarily small constant term.

Equations
Instances For
    theorem Verification.rowMeanTest_error {n : ℕ} (P : ProbabilityTheory.Copula.IntervalPartition n) (C : ProbabilityTheory.Copula 2) (v : ↑unitInterval) (φ : ℝ → ℝ) (hc : Continuous φ) (ε K : ℝ) (hm : ∀ x ∈ Set.Icc 0 1, ∀ y ∈ Set.Icc 0 1, |φ x - φ y| ≤ ε + K * |x - y|) :
    |rowMeanTest P C v φ - ∫ (u : ↑unitInterval), φ (C.conditionalCDF u v)| ≤ ε + K * partitionAverageError P fun (u : ↑unitInterval) => C.conditionalCDF u v