Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 2.4 — every CW-minimum x* of (P) satisfies |∇ᵢf(x*)| ≤ L₂(f)M_s(x*) on I₀(x*) and ∇ᵢf(x*) = 0 on I₁(x*)

Proved
SparseNLO.CWOpt.theorem_2_4

by mikedeng1 · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

nonconvex-optimizationoptimality-conditionsp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1sparse-optimization

Let f:Rn→Rf:\mathbb R^n\to\mathbb Rf:Rn→R be continuously differentiable and bounded below (Assumption 1), with L(f)L(f)L(f)-Lipschitz gradient (Assumption 2), and let 0<s<n0<s<n0<s<n be an integer. Let L2L_2L2​ be a constant satisfying the local Lipschitz condition (2.15), for instance the local Lipschitz constant L2(f)L_2(f)L2​(f). If x∗x^*x∗ is a coordinate-wise minimum of

min⁡ f(x)s.t.∥x∥0≤s,\min\ f(x)\quad\text{s.t.}\quad\|x\|_0\le s,min f(x)s.t.∥x∥0​≤s,

then

∣∇if(x∗)∣ {≤L2 Ms(x∗),i∈I0(x∗),=0,i∈I1(x∗),(2.16)|\nabla_i f(x^*)|\ \begin{cases}\le L_2\,M_s(x^*), & i\in I_0(x^*),\\ =0, & i\in I_1(x^*),\end{cases}\qquad(2.16)∣∇i​f(x∗)∣ {≤L2​Ms​(x∗),=0,​i∈I0​(x∗),i∈I1​(x∗),​(2.16)

where Ms(x∗)M_s(x^*)Ms​(x∗) is the sss-th largest absolute value of the components of x∗x^*x∗. That is, x∗x^*x∗ is an L2(f)L_2(f)L2​(f)-stationary point.

The theorem places coordinate-wise minimality above LLL-stationarity in the hierarchy of necessary optimality conditions for sparsity constrained problems, with the local constant L2(f)≤L(f)L_2(f)\le L(f)L2​(f)≤L(f) in place of the global Lipschitz constant. Since every optimal solution of (P) is a CW-minimum, every optimal solution is an L2(f)L_2(f)L2​(f)-stationary point.

Formalization Note The conclusion is the displayed condition (2.16). Its identification with L2(f)L_2(f)L2​(f)-stationarity in the sense of Definition 2.3 goes through Lemma 2.2 and is not restated here. L2L_2L2​ is universally quantified among constants satisfying (2.15). Assumption 2 is kept as on the page, although (2.15) is the hypothesis the proof uses. When ∥x∗∥0<s\|x^*\|_0<s∥x∗∥0​<s, Ms(x∗)=0M_s(x^*)=0Ms​(x∗)=0 and the first line says ∇if(x∗)=0\nabla_i f(x^*)=0∇i​f(x∗)=0 on I0(x∗)I_0(x^*)I0​(x∗).

Preamble
import Mathlib
import Definitions.Def_SparseNLO_CWOpt_Setting
Formal statement
namespace SparseNLO.CWOpt

/-- Theorem 2.4 (p. 12): under Assumption 2, every CW-minimum `x*` of (P) satisfies (2.16):
`|∇_i f(x*)| ≤ L2 · M_s(x*)` for `i ∈ I₀(x*)` and `∇_i f(x*) = 0` for `i ∈ I₁(x*)`,
where `L2` is any constant satisfying the local Lipschitz condition (2.15). -/
theorem theorem_2_4 {n s : ℕ} (f : EuclideanSpace ℝ (Fin n) → ℝ) (hf : ContDiff ℝ 1 f)
    (hbdd : ∃ γ : ℝ, ∀ x, γ ≤ f x) (hs : 0 < s) (hsn : s < n)
    (Lf : NNReal) (hLf : LipschitzWith Lf (gradient f))
    (L2 : ℝ) (hL2 : LocalLipschitz f L2)
    (xstar : EuclideanSpace ℝ (Fin n)) (hcw : IsCWMin f s xstar) :
    (∀ i ∈ I0 xstar, |gradient f xstar i| ≤ L2 * Ms s xstar) ∧
      (∀ i ∈ I1 xstar, gradient f xstar i = 0) := by sorry

end SparseNLO.CWOpt
Source
Beck, Eldar, Sparsity Constrained Nonlinear Optimization: Optimality Conditions and Algorithms, SIAM J. Optim. (2013), doi:10.1137/120869778 (source: arXiv:1203.4580v1), p. 12, Theorem 2.4, (2.16)
Human review
  • Endorsed by Shuze Chen · Oct 8, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 8, 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