Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

(5.2.2), p. 65 — expected squared norm of a random row-and-column submatrix

Proved
IntroMatrixConc.Submatrix.eq_5_2_2

by mikedeng1 · Oct 9, 2026 · Mathlib 0df444a (Lean v4.33.1)

matrix-chernoffmatrix-concentrationp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-bookp2o-v1submatrix

Let B∈Cd×nB\in\mathbb C^{d\times n}B∈Cd×n, with d,n≥1d,n\ge1d,n≥1. Independently select its rows with probability p/dp/dp/d and its columns with probability r/nr/nr/n, where 0≤p≤d0\le p\le d0≤p≤d and 0≤r≤n0\le r\le n0≤r≤n. Let PPP and RRR be the diagonal zero-one projectors recording the selections, and set Z=PBRZ=PBRZ=PBR. Then

E∥Z∥2≤3pdrn∥B∥2+2plog⁡ddmax⁡k∥b:k∥2+2rlog⁡nnmax⁡j∥bj:∥2+(log⁡d)(log⁡n)max⁡j,k∣bjk∣2.\begin{aligned} \mathbb E\|Z\|^2\le{}&3\frac pd\frac rn\|B\|^2 +2\frac{p\log d}{d}\max_k\|b_{:k}\|^2\\ &+2\frac{r\log n}{n}\max_j\|b_{j:}\|^2 +(\log d)(\log n)\max_{j,k}|b_{jk}|^2. \end{aligned}E∥Z∥2≤​3dp​nr​∥B∥2+2dplogd​kmax​∥b:k​∥2+2nrlogn​jmax​∥bj:​∥2+(logd)(logn)j,kmax​∣bjk​∣2.​

The estimate separates a term involving the whole matrix from fluctuations controlled by its largest column, row, and entry.

Formalization Note The indicators for all rows and columns are one jointly independent measurable family. The book's simplified constants 333 and 222 are used exactly. Integrability of ∥Z∥2\|Z\|^2∥Z∥2 is asserted explicitly under the bounded Bernoulli model; d,n≥1d,n\ge1d,n≥1 keep every maximum and logarithm in its intended domain.

Preamble
import Definitions.Def_IntroMatrixConc_Submatrix_Defs

open MeasureTheory ProbabilityTheory TroppMatrixConcentration
open scoped Matrix.Norms.L2Operator
Formal statement
namespace IntroMatrixConc.Submatrix

/-- Tropp (5.2.2), p. 65: the row-and-column random submatrix expectation bound. -/
theorem eq_5_2_2 {Ω : Type*} [MeasurableSpace Ω]
    (μ : Measure Ω) [IsProbabilityMeasure μ] {d n : ℕ} [NeZero d] [NeZero n]
    (B : Matrix (Fin d) (Fin n) ℂ) (δ : Fin d → Ω → ℝ)
    (ξ : Fin n → Ω → ℝ) (p r : ℝ)
    (hp0 : 0 ≤ p) (hpd : p ≤ (d : ℝ))
    (hr0 : 0 ≤ r) (hrn : r ≤ (n : ℝ))
    (hδMeas : ∀ j, Measurable (δ j)) (hξMeas : ∀ k, Measurable (ξ k))
    (hIndep : iIndepFun (Sum.elim δ ξ) μ)
    (hδ : ∀ j, IsBernoulli μ (δ j) (p / (d : ℝ)))
    (hξ : ∀ k, IsBernoulli μ (ξ k) (r / (n : ℝ))) :
    Integrable (fun ω => spectralNorm (selectedSubmatrix B δ ξ ω) ^ 2) μ ∧
    (∫ ω, spectralNorm (selectedSubmatrix B δ ξ ω) ^ 2 ∂μ) ≤
      (3 : ℝ) * (p / (d : ℝ)) * (r / (n : ℝ)) * spectralNorm B ^ 2 +
      2 * (p * Real.log d / (d : ℝ)) * maxColumnSq B +
      2 * (r * Real.log n / (n : ℝ)) * maxRowSq B +
      Real.log d * Real.log n * maxEntrySq B := by sorry

end IntroMatrixConc.Submatrix
Source
Tropp, arXiv:1501.01571v1, (5.2.2), p. 65; simplification note, p. 66
Human review
  • Endorsed by Shuze Chen · Oct 9, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 9, 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