Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 6 — ‖(d̃, δ̃)‖ ≤ κ‖d_I(x)‖ for P Lipschitz on dom P, κ depending only on the Lipschitz constant

Proved
TsengCGD.ErrorBound.lemma6

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

error-boundnonsmooth-optimizationp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1

Assume the standing assumptions of problem (1): c>0c > 0c>0, PPP proper convex lower semicontinuous, fff continuously differentiable on an open set containing dom⁡P\operatorname{dom}PdomP. Let dI(x)d_I(x)dI​(x) be the residual (13) and, for ξ=P(x)\xi = P(x)ξ=P(x), let (d~,δ~)(\tilde d,\tilde\delta)(d~,δ~) be an optimal solution of the subproblem

min⁡(d,δ){∇f(x)⊤d+12∥d∥2+12δ2+cδ  ∣  (x+d, ξ+δ)∈epi⁡P}.(42)\min_{(d,\delta)} \Big\{ \nabla f(x)^\top d + \tfrac12\|d\|^2 + \tfrac12\delta^2 + c\delta \;\Big|\; (x+d,\ \xi+\delta) \in \operatorname{epi}P \Big\}. \qquad (42)(d,δ)min​{∇f(x)⊤d+21​∥d∥2+21​δ2+cδ​(x+d, ξ+δ)∈epiP}.(42)

For every K≥0K \ge 0K≥0 there is a scalar κ>0\kappa > 0κ>0, depending only on KKK, such that whenever PPP is Lipschitz continuous on dom⁡P\operatorname{dom}PdomP with constant KKK, then for every x∈dom⁡Px \in \operatorname{dom}Px∈domP

∥(d~,δ~)∥≤κ ∥dI(x)∥.\big\|(\tilde d,\tilde\delta)\big\| \le \kappa\,\|d_I(x)\|.​(d~,δ~)​≤κ∥dI​(x)∥.

The lemma compares the residual of the reformulation (41), a smooth problem over epi⁡P\operatorname{epi}PepiP, with the residual dId_IdI​ of the original problem; it is what lets error bounds known for constrained smooth problems be transferred to (1).

Formalization Note κ\kappaκ is quantified before the dimension nnn and the data f,P,cf, P, cf,P,c, so it is a function of KKK alone, as the paper says. The bound is asserted for every optimal solution of (42) (predicate IsEpiSubSol); the paper notes the solution is unique. ∥(d,δ)∥=∥d∥2+δ2\|(d,\delta)\| = \sqrt{\|d\|^2 + \delta^2}∥(d,δ)∥=∥d∥2+δ2​.

Preamble
import Mathlib
import Definitions.Def_TsengCGD_Linear_Basic
import Definitions.Def_TsengCGD_ErrorBound_Conditions
Formal statement
namespace TsengCGD.ErrorBound

theorem lemma6 :
    ∀ K : ℝ, 0 ≤ K → ∃ κ > 0, ∀ {n : ℕ} (f : TsengCGD.Global.Vec n → ℝ) (D : Set (TsengCGD.Global.Vec n)) (P : TsengCGD.Global.Vec n → ℝ) (c : ℝ),
      TsengCGD.Global.Standing f D P c → PLipOn D P K →
        ∀ x ∈ D, ∀ (dt : TsengCGD.Global.Vec n) (δt : ℝ), IsEpiSubSol f D P c x (P x) dt δt →
          pairNorm dt δt ≤ κ * ‖TsengCGD.Global.dI f D P c x‖ := by sorry

end TsengCGD.ErrorBound
Source
Tseng and Yun, A coordinate gradient descent method for nonsmooth separable minimization, Math. Program. Ser. B 117 (2009), p. 410, Lemma 6 (with (42))
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