Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 3.2.2 — Expectation bounds for eigenvalues

Proved
TroppMatrixConcentration.ch3_laplace_expectations

by tc · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

matrix-concentrationrandom-matrices

For a measurable, Bochner-integrable, almost-surely Hermitian random matrix YYY on a probability space, whenever eθYe^{\theta Y}eθY is Bochner integrable,

Eλmax⁡(Y)≤log⁡Etr⁡eθYθ(θ>0),\mathbb E\lambda_{\max}(Y)\le\frac{\log\mathbb E\operatorname{tr}e^{\theta Y}}{\theta}\quad(\theta>0),Eλmax​(Y)≤θlogEtreθY​(θ>0), Eλmin⁡(Y)≥log⁡Etr⁡eθYθ(θ<0).\mathbb E\lambda_{\min}(Y)\ge\frac{\log\mathbb E\operatorname{tr}e^{\theta Y}}{\theta}\quad(\theta<0).Eλmin​(Y)≥θlogEtreθY​(θ<0).

The integrability hypotheses make the expectations finite. The pointwise parameter formulation supports the source's infimum and supremum optimizations over admissible parameters.

Preamble
import Definitions.Def_TroppMatrixConcentration_probability

open MeasureTheory ProbabilityTheory
open scoped Matrix.Norms.L2Operator

Formal statement
namespace TroppMatrixConcentration

theorem ch3_laplace_expectations {Ω : Type*} [MeasurableSpace Ω]
    (μ : Measure Ω) [IsProbabilityMeasure μ] {d : ℕ} [NeZero d]
    (Y : Ω → Matrix (Fin d) (Fin d) ℂ) (θ : ℝ)
    (hMeas : Measurable Y) (hHerm : ∀ᵐ ω ∂μ, (Y ω).IsHermitian)
    (hInt : Integrable Y μ)
    (hExp : Integrable (fun ω => matrixExp (θ • Y ω)) μ) :
    (0 < θ → (∫ ω, lambdaMax (Y ω) ∂μ) ≤
      Real.log (∫ ω, traceExp (θ • Y ω) ∂μ) / θ) ∧
    (θ < 0 → Real.log (∫ ω, traceExp (θ • Y ω) ∂μ) / θ ≤
      (∫ ω, lambdaMin (Y ω) ∂μ)) := by sorry

end TroppMatrixConcentration
Source
Joel A. Tropp, An Introduction to Matrix Concentration Inequalities, arXiv:1501.01571v1 (7 January 2015); https://arxiv.org/abs/1501.01571v1; Proposition 3.2.2, printed p. 33.
Read-back

What the Lean code literally says, in plain math · gpt-6

For every measurable space Ω\OmegaΩ, every probability measure μ\muμ on it, every natural number d≠0d\ne 0d=0, every measurable function Y:Ω→Cd×dY:\Omega\to\mathbb C^{d\times d}Y:Ω→Cd×d (with the Borel measurable structure on matrices), and every real number θ\thetaθ, suppose that Y(ω)Y(\omega)Y(ω) is Hermitian for μ\muμ-almost every ω\omegaω, that YYY is Bochner integrable with respect to μ\muμ, and that ω↦exp⁡(θY(ω))\omega\mapsto\exp(\theta Y(\omega))ω↦exp(θY(ω)) is also Bochner integrable with respect to μ\muμ. Here matrices carry the Euclidean operator norm, scalar multiplication by θ\thetaθ means multiplication by the real scalar, and exp⁡(A)\exp(A)exp(A) is the matrix exponential defined by its power series. For any matrix AAA, put λmax⁡(A)=sup⁡σR(A)\lambda_{\max}(A)=\sup\sigma_{\mathbb R}(A)λmax​(A)=supσR​(A) and λmin⁡(A)=inf⁡σR(A)\lambda_{\min}(A)=\inf\sigma_{\mathbb R}(A)λmin​(A)=infσR​(A), where σR(A)\sigma_{\mathbb R}(A)σR​(A) is the set of real numbers rrr for which rI−ArI-ArI−A is not invertible; these use the total real supremum and infimum operations, including their value 000 on the empty set. Put T=∫ΩRe⁡(tr⁡(exp⁡(θY(ω)))) dμ(ω)T=\int_\Omega\operatorname{Re}\bigl(\operatorname{tr}(\exp(\theta Y(\omega)))\bigr)\,d\mu(\omega)T=∫Ω​Re(tr(exp(θY(ω))))dμ(ω). Then both implications hold: if 0<θ0<\theta0<θ, then ∫Ωλmax⁡(Y(ω)) dμ(ω)≤(log⁡T)/θ\int_\Omega\lambda_{\max}(Y(\omega))\,d\mu(\omega)\le (\log T)/\theta∫Ω​λmax​(Y(ω))dμ(ω)≤(logT)/θ; and if θ<0\theta<0θ<0, then (log⁡T)/θ≤∫Ωλmin⁡(Y(ω)) dμ(ω)(\log T)/\theta\le\int_\Omega\lambda_{\min}(Y(\omega))\,d\mu(\omega)(logT)/θ≤∫Ω​λmin​(Y(ω))dμ(ω). All displayed integrals are Bochner integrals. The Hermitian requirement permits an exceptional set of measure zero. The dimension-zero case is excluded, whereas θ=0\theta=0θ=0 is permitted and makes both implications vacuous; the formal real logarithm and division are total, with log⁡0=0\log 0=0log0=0 and division by zero equal to 000.

Human review
  • Endorsed by Shuze Chen · Oct 7, 2026

    Confirmed by the moderator at approval.

  • Endorsed by tc · Oct 7, 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