Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§6.2, p. 41 — the basis-pursuit x-update is (I−AT(AAT)−1A)(zk−uk)+AT(AAT)−1b(I-A^T(AA^T)^{-1}A)(z^k-u^k)+A^T(AA^T)^{-1}b(I−AT(AAT)−1A)(zk−uk)+AT(AAT)−1b

Proved
BoydADMM.L1.basis_pursuit_x_update

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

admmbasis-pursuitp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-bookp2o-v1projection

Let A∈Rm×nA\in\mathbb R^{m\times n}A∈Rm×n with AATAA^TAAT invertible, b∈Rmb\in\mathbb R^mb∈Rm, and z,u∈Rnz,u\in\mathbb R^nz,u∈Rn. Then

x⋆=(I−AT(AAT)−1A)(z−u)+AT(AAT)−1bx^\star=\bigl(I-A^T(AA^T)^{-1}A\bigr)(z-u)+A^T(AA^T)^{-1}bx⋆=(I−AT(AAT)−1A)(z−u)+AT(AAT)−1b

is the Euclidean projection of z−uz-uz−u onto the affine set {x∈Rn∣Ax=b}\{x\in\mathbb R^n\mid Ax=b\}{x∈Rn∣Ax=b}: Ax⋆=bAx^\star=bAx⋆=b, and ∥x⋆−(z−u)∥2<∥x−(z−u)∥2\|x^\star-(z-u)\|_2<\|x-(z-u)\|_2∥x⋆−(z−u)∥2​<∥x−(z−u)∥2​ for every other xxx with Ax=bAx=bAx=b.

With z=zkz=z^kz=zk and u=uku=u^ku=uk this is the x-update xk+1=Π(zk−uk)x^{k+1}=\Pi(z^k-u^k)xk+1=Π(zk−uk) of ADMM for basis pursuit, minimize ∥x∥1\|x\|_1∥x∥1​ subject to Ax=bAx=bAx=b: a minimum Euclidean norm problem with linear constraints, in closed form.

Formalization Note The book assumes m<nm<nm<n and uses (AAT)−1(AA^T)^{-1}(AAT)−1 without stating that it exists; the statement assumes det⁡(AAT)≠0\det(AA^T)\ne0det(AAT)=0 (AAA has full row rank), which is what the formula needs. Uniqueness of the nearest point is part of the conclusion. Vectors are EuclideanSpace, matrices act by Matrix.toEuclideanLin.

Preamble
import Mathlib
import Definitions.Def_BoydADMM_L1_Basic

open Matrix
Formal statement
namespace BoydADMM.L1

/-- §6.2, p. 41: if `AAᵀ` is invertible, then
`(I − Aᵀ(AAᵀ)⁻¹A)(z − u) + Aᵀ(AAᵀ)⁻¹b` is the Euclidean projection `Π(z − u)` of `z − u` onto
`{x ∈ ℝⁿ | Ax = b}`: the unique point of that set nearest to `z − u`. -/
theorem basis_pursuit_x_update {m n : ℕ} (A : Matrix (Fin m) (Fin n) ℝ)
    (b : EuclideanSpace ℝ (Fin m)) (hA : (A * Aᵀ).det ≠ 0) (z u : EuclideanSpace ℝ (Fin n)) :
    IsUniqueMinimizerOn (fun x : EuclideanSpace ℝ (Fin n) => ‖x - (z - u)‖)
      {x | Matrix.toEuclideanLin A x = b}
      (Matrix.toEuclideanLin (1 - Aᵀ * (A * Aᵀ)⁻¹ * A) (z - u) +
        Matrix.toEuclideanLin (Aᵀ * (A * Aᵀ)⁻¹) b) := by sorry

end BoydADMM.L1
Source
Boyd, Parikh, Chu, Peleato, Eckstein, Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers, Found. Trends Mach. Learn. 3(1) (2011), p. 41, §6.2
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 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