AFTD Auto-Formalizing Theoretical Domains

CommunicationComplexity.integral_klFun_rnDeriv_eq_kl_integral

theoremverifiedimported Lean

Two integral forms of the Kullback–Leibler divergence. Let $\mu$ and $\nu$ be probability measures on a measurable space $\Omega$ with $\mu$ absolutely continuous with respect to $\nu$, and write $\frac{d\mu}{d\nu}$ for the (real-valued) Radon–Nikodym density of $\mu$ with respect to $\nu$. Assume that the log-likelihood ratio $x \mapsto \log\frac{d\mu}{d\nu}(x)$ is $\mu$-integrable. Then \[ \int_\Omega \varphi\!\left(\frac{d\mu}{d\nu}(x)\right) d\nu(x) \;=\; \int_\Omega \log\frac{d\mu}{d\nu}(x)\, d\mu(x), \] where $\varphi(t) = t\log t + 1 - t$.

Statement

set_option relaxedAutoImplicit false in
set_option autoImplicit false in
open MeasureTheory in
open ProbabilityTheory in
open scoped ENNReal in
theorem CommunicationComplexity.integral_klFun_rnDeriv_eq_kl_integral
    {Ω : Type*} [MeasurableSpace Ω]
    (μ ν : Measure Ω) [IsProbabilityMeasure μ] [IsProbabilityMeasure ν]
    (h_ac : μ ≪ ν)
    (h_int : Integrable (llr μ ν) μ) :
    (∫ x, InformationTheory.klFun (rnDensity μ ν x) ∂ν) =
      ∫ x, llr μ ν x ∂μ
Source
TCSlib, `CommunicationComplexity.integral_klFun_rnDeriv_eq_kl_integral`
Lean proof
By Lucy Horowitz, Timothe Kasriel, Mihir Singhal, from github.com/Shilun-Allan-Li/tcslib at c8a02591b1e5 (Copyright (c) 2026 Lucy Horowitz, Timothe Kasriel, and Mihir Singhal. All rights reserved, Apache-2.0). Compiled here against leanprover/lean4:v4.33.1, Mathlib v4.33.1; written for leanprover/lean4:v4.25.0.
Verified
07 Oct 2026
Axioms
Classical.choiceQuot.soundpropext
Built on 1
Used by 1

Lean source view module on GitHub

set_option relaxedAutoImplicit false inset_option autoImplicit false inopen MeasureTheory inopen ProbabilityTheory inopen scoped ENNReal in/-- If `μ ≪ ν` and the log-likelihood ratio is `μ`-integrable, then `∫ klFun (dμ/dν) dν = ∫ llr μ ν dμ`, the real-valued KL divergence (Mathlib's `integral_klFun_rnDeriv`). -/theorem CommunicationComplexity.integral_klFun_rnDeriv_eq_kl_integral    {Ω : Type*} [MeasurableSpace Ω]    (μ ν : Measure Ω) [IsProbabilityMeasure μ] [IsProbabilityMeasure ν]    (h_ac : μ ≪ ν)    (h_int : Integrable (llr μ ν) μ) :    (∫ x, InformationTheory.klFun (rnDensity μ ν x) ∂ν) =      ∫ x, llr μ ν x ∂μ := by  have h := InformationTheory.integral_klFun_rnDeriv h_ac h_int  simpa [rnDensity] using h
See it in the graphDiscuss this resultChallenge it