Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proposition 8.3.5 — Generalized Klein inequality

Proved
TroppMatrixConcentration.ch8_generalized_klein

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

convexitymatrix-analysismatrix-concentration

Let III be an interval of the real line, and let fi,gi:I→Rf_i,g_i:I\to\mathbb Rfi​,gi​:I→R be a finite family of scalar functions such that ∑ifi(a)gi(h)≥0\sum_i f_i(a)g_i(h)\ge0∑i​fi​(a)gi​(h)≥0 for every a,h∈Ia,h\in Ia,h∈I. If A,HA,HA,H are Hermitian complex matrices of the same positive dimension with all eigenvalues in III, then

∑iRe⁡tr⁡[fi(A)gi(H)]≥0.\sum_i\operatorname{Re}\operatorname{tr}[f_i(A)g_i(H)]\ge0.i∑​Retr[fi​(A)gi​(H)]≥0.

No scalar continuity assumption is required, because the matrix functional calculus evaluates each function on a finite spectrum. This transfers scalar inequalities involving two variables to trace inequalities.

Preamble
import Definitions.Def_TroppMatrixConcentration_ch8_entropy

open scoped Matrix.Norms.L2Operator ComplexOrder

Formal statement
namespace TroppMatrixConcentration

theorem ch8_generalized_klein {d N : ℕ} [NeZero d]
    (I : Set ℝ) (hI : Convex ℝ I) (f g : Fin N → ℝ → ℝ)
    (hfg : ∀ a ∈ I, ∀ h ∈ I, 0 ≤ ∑ i, f i a * g i h)
    (A H : Matrix (Fin d) (Fin d) ℂ) (hA : A.IsHermitian) (hH : H.IsHermitian)
    (hAI : spectrum ℝ A ⊆ I) (hHI : spectrum ℝ H ⊆ I) :
    0 ≤ ∑ i, (Matrix.trace (ch8_matrixFunction (f i) A *
      ch8_matrixFunction (g i) H)).re := 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 8.3.5, printed p. 126.
Read-back

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

For every positive natural number ddd, every natural number NNN (including N=0N=0N=0), every convex subset I⊆RI\subseteq\mathbb RI⊆R, and every two families of functions fi,gi:R→Rf_i,g_i:\mathbb R\to\mathbb Rfi​,gi​:R→R indexed by i=0,…,N−1i=0,\ldots,N-1i=0,…,N−1, suppose that ∑i=0N−1fi(a)gi(h)≥0\sum_{i=0}^{N-1}f_i(a)g_i(h)\geq 0∑i=0N−1​fi​(a)gi​(h)≥0 for every a∈Ia\in Ia∈I and every h∈Ih\in Ih∈I. For any two Hermitian complex d×dd\times dd×d matrices AAA and HHH whose real spectra are contained in III, the following inequality holds: 0≤∑i=0N−1Re⁡tr⁡(fi(A)gi(H))\displaystyle 0\leq\sum_{i=0}^{N-1}\operatorname{Re}\operatorname{tr}\bigl(f_i(A)g_i(H)\bigr)0≤i=0∑N−1​Retr(fi​(A)gi​(H)). Here each matrix function is defined using the real continuous functional calculus: for a Hermitian matrix BBB with spectral decomposition B=∑λλPλB=\sum_\lambda\lambda P_\lambdaB=∑λ​λPλ​, the notation u(B)u(B)u(B) means ∑λu(λ)Pλ\sum_\lambda u(\lambda)P_\lambda∑λ​u(λ)Pλ​. No continuity, differentiability, or convexity of the scalar functions is assumed. The matrices need not commute. When N=0N=0N=0, both scalar and matrix sums are empty and equal to zero. Although convex sets may in general be empty, the spectral containment assumptions for positive-size Hermitian matrices exclude an empty III.

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