Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 10.8.1 (Pinsker Inequality) — D(p∥q)≥12ln⁡2∥p−q∥12D(p\|q)\ge\frac{1}{2\ln 2}\|p-q\|_1^2D(p∥q)≥2ln21​∥p−q∥12​

Proved
WildeQIT.PinskerRefined.theorem_10_8_1

by iris · Oct 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

entropy-inequalitiesinformation-theoryp2o-gran-per-chapterp2o-plan-bookp2o-v1pinskerrelative-entropy

Let ppp be a probability distribution on a finite alphabet X\mathcal XX and let q:X→[0,1]q:\mathcal X\to[0,1]q:X→[0,1] satisfy ∑xq(x)≤1\sum_x q(x)\le 1∑x​q(x)≤1. Then

D(p∥q)≥12ln⁡2 ∥p−q∥12,(10.127)D(p\|q)\ge\frac{1}{2\ln 2}\,\|p-q\|_1^2, \tag{10.127}D(p∥q)≥2ln21​∥p−q∥12​,(10.127)

where DDD is the relative entropy in bits (Definition 10.5.2) and ∥p−q∥1=∑x∣p(x)−q(x)∣\|p-q\|_1=\sum_x|p(x)-q(x)|∥p−q∥1​=∑x​∣p(x)−q(x)∣ (Definition 10.7.1).

Pinsker's inequality turns a small relative entropy into a small trace distance. It strengthens the non-negativity of relative entropy, and it is the tool behind the lower bounds of Theorems 10.8.2 and 10.8.3.

Formalization Note DDD is relEntropy with values in the extended reals (+∞+\infty+∞ off support). The bound q(x)≤1q(x)\le 1q(x)≤1 is not a separate hypothesis because it follows from q≥0q\ge 0q≥0 and ∑xq(x)≤1\sum_x q(x)\le 1∑x​q(x)≤1.

Differences from the source:

  • encoding: q : X → [0, 1] with Σ_x q(x) ≤ 1 → q ≥ 0 and Σ_x q(x) ≤ 1; reason: q(x) ≤ 1 follows from these two hypotheses, so the set of admissible q is the same
Preamble
import Definitions.Def_WildeQIT_relEntropy
import Definitions.Def_WildeQIT_classicalTraceDist
Formal statement
namespace WildeQIT.PinskerRefined

open WildeQIT

/-- **Theorem 10.8.1 (Pinsker Inequality)**, Wilde, *From Classical to Quantum Shannon Theory*, arXiv:1106.1445v8, p. 311: let `p` be a probability distribution
on a finite alphabet `𝒳` and `q : 𝒳 → [0, 1]` with `∑_x q(x) ≤ 1`. Then
`D(p‖q) ≥ (1 / (2 ln 2)) ‖p − q‖₁²` (10.127).

**Formalization Note** `D` is the published `relEntropy` (bits, `EReal`, `+∞` off support,
Definition 10.5.2) and `‖p − q‖₁ = ∑_x |p(x) − q(x)|` is the published `classicalTraceDist`
(Definition 10.7.1). The book's `q(x) ≤ 1` is not a separate hypothesis: it follows from
`q ≥ 0` and `∑_x q(x) ≤ 1`. -/
theorem theorem_10_8_1 {α : Type} [Fintype α] (p : FinDist α) (q : α → ℝ)
    (hq0 : ∀ x, 0 ≤ q x) (hq1 : ∑ x, q x ≤ 1) :
    ((1 / (2 * Real.log 2) * classicalTraceDist p.prob q ^ 2 : ℝ) : EReal) ≤
      relEntropy p q := by sorry

end WildeQIT.PinskerRefined
Source
Wilde, From Classical to Quantum Shannon Theory, arXiv:1106.1445v8, Theorem 10.8.1, p. 311, Eq. (10.127)
Human review
  • Endorsed by Shuze Chen · Nov 1, 2026

    Confirmed by the moderator at approval.

  • Endorsed by iris · Nov 1, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me