Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§10.6, p. 38 — total variation is at most half the Pearson statistic

Proved
RobustSAA.Discrete.chi_tv_bound

by mikedeng1 · Oct 8, 2026 · Mathlib 0df444a (Lean v4.33.1)

finite-supportgoodness-of-fitp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1uniform-consistency

Let sss be any finite sample on the known support, let p^N\widehat p_Np​N​ be its empirical frequencies, and let F0F_0F0​ be any hypothetical distribution with probability vector p0p_0p0​. Pearson's statistic satisfies

dTV(p^N,p0)≤XN(F0)2.d_{\mathrm{TV}}(\widehat p_N,p_0)\leq\frac{X_N(F_0)}{2}.dTV​(p​N​,p0​)≤2XN​(F0​)​.

This is the paper's Cauchy–Schwarz estimate, which turns membership in a Pearson confidence region into a shrinking total-variation bound.

Formalization Note The two sides are compared in [0,+∞][0,+\infty][0,+∞] so that a zero hypothetical probability paired with a positive empirical frequency gives XN=+∞X_N=+\inftyXN​=+∞; no unjustified positive-probability assumption is added.

Preamble
import Mathlib
import Definitions.Def_RobustSAA_Discrete_Setting
Formal statement
namespace RobustSAA.Discrete

open MeasureTheory

theorem chi_tv_bound {n N : ℕ} (F₀ : ProbabilityMeasure (Fin n))
    (s : Fin N → Fin n) :
    ENNReal.ofReal (dTV (phat s) (pvec F₀)) ≤ chiStat F₀ s / 2 := by sorry

end RobustSAA.Discrete
Source
Bertsimas, Gupta, Kallus, Robust Sample Average Approximation, arXiv:1408.4445v3, §10.6, p. 38, second display
Human review
  • Endorsed by Shuze Chen · Oct 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 8, 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