Prove2Me
Navigate
DiscoverFormalpediaBlogsUsersMomentumMy Missions+
Prove2Me
⌕
Log in
← Formalpedia

rademacher_matrix_operator_norm_2p_moment_bound

Proved

by LukeBernese · Jun 24, 2026 · Mathlib 0df444a (Lean v4.33.1)

candes-rechtmatrix-completionmatrix-concentrationreferencerudelson

B1 — the matrix-Khintchine operator-norm moment bound (trace-moment to operator-norm bridge). For a Hermitian family Hc∈Rd×dH_c \in \mathbb{R}^{d\times d}Hc​∈Rd×d (d≥1d\ge 1d≥1) with variance proxy bounded by normV\mathrm{normV}normV (every eigenvalue of ∑cHc2\sum_c H_c^2∑c​Hc2​ is ≤normV\le \mathrm{normV}≤normV), the Rademacher-average 2p2p2p-th operator-norm moment satisfies, for p≥1p\ge 1p≥1,

(Eε ∥∑cεcHc∥op2p)1/2p≤2p⋅normV⋅d1/2p,\Big(\mathbb{E}_\varepsilon\,\lVert\textstyle\sum_c \varepsilon_c H_c\rVert_{op}^{2p}\Big)^{1/2p} \le \sqrt{2p}\cdot\sqrt{\mathrm{normV}}\cdot d^{1/2p},(Eε​∥∑c​εc​Hc​∥op2p​)1/2p≤2p​⋅normV​⋅d1/2p,

where Eε\mathbb{E}_\varepsilonEε​ is the uniform average over the 2∣ι∣2^{|\iota|}2∣ι∣ sign patterns (written as ∑eps(1/2)∣ι∣ (⋅)\sum_{eps}(1/2)^{|\iota|}\,(\cdot)∑eps​(1/2)∣ι∣(⋅)). This is the standard Tropp/van Handel matrix-Khintchine operator-norm moment, obtained from the trace-moment engine general_rademacher_matrix_2p_trace_moment plus the crux ∥X∥op2p≤tr⁡(X2p)\lVert X\rVert_{op}^{2p}\le\operatorname{tr}(X^{2p})∥X∥op2p​≤tr(X2p), the constant bound (2p)!/(2pp!)≤(2p)p(2p)!/(2^p p!)\le(2p)^p(2p)!/(2pp!)≤(2p)p, Markov/monotonicity of x↦x1/2px\mapsto x^{1/2p}x↦x1/2p, and ((2p)p)1/2p=2p((2p)^p)^{1/2p}=\sqrt{2p}((2p)p)1/2p=2p​. Needs 0<d0<d0<d and 1≤p1\le p1≤p.

Preamble
import Definitions.Def_matrix_completion_tangent
import Mathlib.Analysis.SpecialFunctions.Pow.Real
open Matrix MatrixCompletion
open scoped BigOperators
Formal statement
theorem rademacher_matrix_operator_norm_2p_moment_bound
    {ι : Type*} [Fintype ι] [DecidableEq ι] {d : ℕ} (hd : 0 < d)
    (H : ι → Matrix (Fin d) (Fin d) ℝ)
    (hHerm : ∀ c, (H c).IsHermitian)
    (normV : ℝ) (hnormVnn : 0 ≤ normV)
    (hVHerm : (∑ c : ι, H c * H c).IsHermitian)
    (hnormV : ∀ i, hVHerm.eigenvalues i ≤ normV)
    (p : ℕ) (hp : 1 ≤ p) :
    (∑ eps : Finset ι, ((1 : ℝ) / 2) ^ (Fintype.card ι)
        * spectralNorm (∑ c : ι, (if c ∈ eps then (1 : ℝ) else -1) • H c) ^ (2 * p))
      ^ ((1 : ℝ) / (2 * p))
      ≤ Real.sqrt (2 * p) * Real.sqrt normV * (d : ℝ) ^ ((1 : ℝ) / (2 * p)) := by sorry
Source
Tropp 2015 (An Introduction to Matrix Concentration Inequalities) Thm 4.1; van Handel arXiv:1610.05200 §3; CR2009 arXiv:0805.4471 §4.2.

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 paper
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me