Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 5.1, p. 488 — under B1–B3, C1 or C2 and a bounded level set, essentially cyclic BCD is defined, bounded, and its cluster points are coordinatewise minima

Proved
TsengBCD.Hemivariate.theorem_5_1

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

block-coordinate-descentconvergencecoordinatewise-minimumhemivariatep2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1quasiconvex

Let f=f0+∑k=1Nfkf=f_0+\sum_{k=1}^N f_kf=f0​+∑k=1N​fk​ satisfy Assumptions B1–B3, and let f0f_0f0​ satisfy either Assumption C1 or C2. Let x0∈dom⁡fx^0\in\operatorname{dom}fx0∈domf and assume that the level set

{x:f(x)≤f(x0)}\{x: f(x)\le f(x^0)\}{x:f(x)≤f(x0)}

is bounded. Then the sequence {xr}\{x^r\}{xr} generated by the BCD method (2)–(3) from x0x^0x0 using the essentially cyclic rule is defined, bounded, and every cluster point of {xr}\{x^r\}{xr} is a coordinatewise minimum point of fff.

This is the main convergence theorem of §5. It extends a result of Auslender for strongly convex f0f_0f0​ on a product of closed convex sets to the nonconvex, nondifferentiable setting where only quasiconvexity and hemivariance in each block are assumed.

Formalization Note. "Defined" is stated as: for every sequence of block choices there is a BCD run from x0x^0x0 (this is stronger than asking it only for essentially cyclic choices). "Bounded" and the cluster-point conclusion are stated for every BCD run from x0x^0x0 whose block choices follow the essentially cyclic rule. f0,fkf_0,f_kf0​,fk​ never take the value −∞-\infty−∞; blocks are 0-based; x0∈dom⁡fx^0\in\operatorname{dom}fx0∈domf is the initialization of the method.

Preamble
import Mathlib
import Definitions.Def_TsengBCD_Hemivariate_Setting
Formal statement
namespace TsengBCD.Hemivariate

open Filter Topology

theorem theorem_5_1 {N : ℕ} [NeZero N] {n : Fin N → ℕ} (f0 : TsengBCD.Stationary.X n → EReal)
    (fk : (k : Fin N) → EuclideanSpace ℝ (Fin (n k)) → EReal)
    (h0bot : ∀ x, f0 x ≠ ⊥) (hkbot : ∀ k y, fk k y ≠ ⊥)
    (hB : AssumptionB f0 fk) (hC : AssumptionC1 f0 ∨ AssumptionC2 f0)
    (x0 : TsengBCD.Stationary.X n) (hx0 : TsengBCD.Stationary.fSum f0 fk x0 ≠ ⊤)
    (hbdd : Bornology.IsBounded (TsengBCD.Stationary.levelSet (TsengBCD.Stationary.fSum f0 fk) x0)) :
    (∀ s : ℕ → Fin N, ∃ x : ℕ → TsengBCD.Stationary.X n, x 0 = x0 ∧ TsengBCD.Stationary.IsBCDRun (TsengBCD.Stationary.fSum f0 fk) s x) ∧
      ∀ (s : ℕ → Fin N) (x : ℕ → TsengBCD.Stationary.X n), TsengBCD.Stationary.IsEssCyclic s → x 0 = x0 → TsengBCD.Stationary.IsBCDRun (TsengBCD.Stationary.fSum f0 fk) s x →
        Bornology.IsBounded (Set.range x) ∧
          ∀ z, MapClusterPt z atTop x → TsengBCD.Stationary.IsCoordMin (TsengBCD.Stationary.fSum f0 fk) z := by sorry

end TsengBCD.Hemivariate
Source
Tseng, Convergence of a block coordinate descent method for nondifferentiable minimization, J. Optim. Theory Appl. 109 (2001), p. 488, Theorem 5.1
Human review
  • Endorsed by Shuze Chen · Oct 7, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · 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