Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 4.12 (first assertion), p. 27 — for an atom δ with δ + γ ⪯_E ν, γ ⪯_E ν − S^ν(δ)

Proved
MartOT.Shadow.lemma_4_12_ext

by tianyipeng · Nov 5, 2026 · Mathlib 0df444a (Lean v4.33.1)

convex-orderextended-convex-ordershadow

Let M\mathcal MM be the finite Borel measures on R\mathbb RR with finite first moment, and write μ⪯Eν\mu\preceq_E\nuμ⪯E​ν (extended convex order) if ∫φ dμ≤∫φ dν\int\varphi\,d\mu\le\int\varphi\,d\nu∫φdμ≤∫φdν for every nonnegative convex φ:R→R\varphi:\mathbb R\to\mathbb Rφ:R→R. For μ⪯Eν\mu\preceq_E\nuμ⪯E​ν, the shadow Sν(μ)S^\nu(\mu)Sν(μ) is the measure η≤ν\eta\le\nuη≤ν with μ⪯Cη\mu\preceq_C\etaμ⪯C​η that is smallest in the convex order among all such measures (Lemma 4.6).

Let δ=α δx\delta=\alpha\,\delta_xδ=αδx​ be an atom of mass α≥0\alpha\ge0α≥0 at x∈Rx\in\mathbb Rx∈R, and let γ,ν∈M\gamma,\nu\in\mathcal Mγ,ν∈M with δ+γ⪯Eν\delta+\gamma\preceq_E\nuδ+γ⪯E​ν. Then

γ⪯Eν−Sν(δ).\gamma\preceq_E\nu-S^\nu(\delta).γ⪯E​ν−Sν(δ).

This is the first assertion of Lemma 4.12: after the shadow of a single atom is removed from ν\nuν, the rest of the measure still dominates γ\gammaγ in the extended convex order, so the shadow Sν−Sν(δ)(γ)S^{\nu-S^\nu(\delta)}(\gamma)Sν−Sν(δ)(γ) exists. It is the step that makes the decomposition Sν(δ+γ)=Sν(δ)+Sν−Sν(δ)(γ)S^\nu(\delta+\gamma)=S^\nu(\delta)+S^{\nu-S^\nu(\delta)}(\gamma)Sν(δ+γ)=Sν(δ)+Sν−Sν(δ)(γ) of Lemma 4.12 meaningful.

Formalization note. The shadow is encoded by the predicate IsShadow ν δ η1, properties (i)–(iii) of Lemma 4.6, and the claim is stated for every such η1\eta_1η1​. The difference ν−η1\nu-\eta_1ν−η1​ is Mathlib's truncated subtraction of measures, which is the paper's difference because η1≤ν\eta_1\le\nuη1​≤ν.

Preamble
import Mathlib
import Definitions.Def_MartOT_Var_Setting
Formal statement
namespace MartOT.Shadow

open MeasureTheory NNReal

theorem lemma_4_12_ext (α : ℝ≥0) (x : ℝ) (γ ν : Measure ℝ) (hγ : MartOT.Var.InM γ)
    (hν : MartOT.Var.InM ν)
    (h : MartOT.Var.ExtConvexLE ((α : ENNReal) • Measure.dirac x + γ) ν) :
    ∀ η1 : Measure ℝ, MartOT.Var.IsShadow ν ((α : ENNReal) • Measure.dirac x) η1 →
      MartOT.Var.ExtConvexLE γ (ν - η1) := by sorry

end MartOT.Shadow
Source
Beiglböck–Juillet, On a problem of optimal transport under marginal martingale constraints, arXiv:1208.1509v2, Lemma 4.12 (first assertion: γ ⪯_E ν − S^ν(δ)), p. 27

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