Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 7: the coupled lower bound on the cost of πˉ\bar\piπˉ through ikr∗i^{r^*}_kikr∗​

Proved
LostSalesConstantOrder.LargeLeadTime.lemma_7

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

couplinginventorylost-salesp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

In the setting of Lemma 6 (π\piπ never orders more than QQQ, (x∗,I∗)(\mathbf x^*,\mathcal I^*)(x∗,I∗) solves (4)), let r∗=L−1∑i=1Lxi∗r^*=L^{-1}\sum_{i=1}^Lx^*_ir∗=L−1∑i=1L​xi∗​, let I1,∞r∗I^{r^*}_{1,\infty}I1,∞r∗​ be a copy of I∞r∗I^{r^*}_\inftyI∞r∗​ independent of the demands, and let ikr∗i^{r^*}_kikr∗​ be the largest index attaining Wkr∗W^{r^*}_kWkr∗​. For every period τ\tauτ, E[∑t=ττ+L−1Ctπ]\mathbb E[\sum_{t=\tau}^{\tau+L-1}C^{\pi}_t]E[∑t=ττ+L−1​Ctπ​] is at least

h(∑k=1LE[ikr∗r∗−∑i=1ikr∗Di]+∑k=1LE[∑i=k+1−ikr∗kxi∗]−r∗∑k=1LE[ikr∗])+c(E[IL+1∗]−I∗+L E[D]−∑i=1Lxi∗).h\Big(\sum_{k=1}^L\mathbb E\Big[i^{r^*}_kr^*-\sum_{i=1}^{i^{r^*}_k}D_i\Big]+\sum_{k=1}^L\mathbb E\Big[\sum_{i=k+1-i^{r^*}_k}^kx^*_i\Big]-r^*\sum_{k=1}^L\mathbb E[i^{r^*}_k]\Big)+c\Big(\mathbb E[I^*_{L+1}]-\mathcal I^*+L\,\mathbb E[D]-\sum_{i=1}^Lx^*_i\Big).h(k=1∑L​E[ikr∗​r∗−i=1∑ikr∗​​Di​]+k=1∑L​E[i=k+1−ikr∗​∑k​xi∗​]−r∗k=1∑L​E[ikr∗​])+c(E[IL+1∗​]−I∗+LE[D]−i=1∑L​xi∗​).

The coupling of VkV_kVk​ with the index ikr∗i^{r^*}_kikr∗​ puts the lower bound in the same terms as the cost of the constant-order policy πr∗\pi_{r^*}πr∗​ (Lemma 4).

Formalization Note As in Lemma 6, the statement is for every policy that never orders more than QQQ (the paper's πˉ\bar\piπˉ is one) and every window start. The expectations are over the product of the law of I∞r∗I^{r^*}_\inftyI∞r∗​ and the demand law. No hypothesis r∗<E[D]r^*<\mathbb E[D]r∗<E[D] is added (the paper states none); the integrands do not involve I1,∞r∗I^{r^*}_{1,\infty}I1,∞r∗​ except through the index ikr∗i^{r^*}_kikr∗​. The real right-hand side is mapped into [0,∞][0,\infty][0,∞] with ENNReal.ofReal.

Preamble
import Definitions.Def_LostSalesConstantOrder_LargeLeadTime_RandomWalk

open MeasureTheory ProbabilityTheory Filter
open scoped ENNReal NNReal BigOperators
Formal statement
namespace LostSalesConstantOrder.LargeLeadTime

/-- Lemma 7, p. 9. Hypotheses as in Lemma 6; `r* = L⁻¹ Σ x*_i` and `i^{r*}_k` is built from an
independent pair `(I^{r*}_{1,∞}, (D_i))`. Then `E[Σ_{t=τ}^{τ+L−1} C_t^π]` is at least
`h(Σ_k E[i^{r*}_k r* − Σ_{i=1}^{i^{r*}_k} D_i] + Σ_k E[Σ_{i=k+1−i^{r*}_k}^k x*_i] − r* Σ_k E[i^{r*}_k])
+ c(E[I*_{L+1}] − 𝓘* + L E[D] − Σ x*_i)`, all sums over `k = 1, …, L`. -/
theorem lemma_7 (P : CappedBaseStock.DemandLaw) (prm : CappedBaseStock.Parameters)
    (h3 : FiniteThirdMoment P) (hnd : NonDeterministic P)
    (π : CappedBaseStock.AdmissiblePolicy) (hπ : OrdersBounded P π (Qlevel P prm))
    (x : Fin prm.L → ℝ≥0) (I : ℝ≥0) (hmin : IsWindowMinimizer P prm x I) (t₀ : ℕ) :
    ENNReal.ofReal
        (prm.h *
            ((∑ k ∈ Finset.Icc 1 prm.L,
                ∫ p, ((ik (rStar x) k p.1 p.2 : ℝ) * (rStar x : ℝ) -
                  ∑ i ∈ Finset.Icc 1 (ik (rStar x) k p.1 p.2), (p.2 (i - 1) : ℝ)) ∂((lawSupWalk P (rStar x)).prod (CappedBaseStock.demandPathLaw P))) +
              (∑ k ∈ Finset.Icc 1 prm.L,
                ∫ p, ∑ i ∈ Finset.Icc (k + 1 - ik (rStar x) k p.1 p.2) k, (slot x (i - 1) : ℝ) ∂((lawSupWalk P (rStar x)).prod (CappedBaseStock.demandPathLaw P))) -
              (rStar x : ℝ) * ∑ k ∈ Finset.Icc 1 prm.L, ∫ p, (ik (rStar x) k p.1 p.2 : ℝ) ∂((lawSupWalk P (rStar x)).prod (CappedBaseStock.demandPathLaw P))) +
          prm.p * ((∫ d, (endInventory prm x I d : ℝ) ∂CappedBaseStock.demandPathLaw P) -
            I + prm.L * CappedBaseStock.mean P - ∑ i, (x i : ℝ))) ≤
      costWindow P prm π t₀ := by sorry

end LostSalesConstantOrder.LargeLeadTime
Source
Goldberg, Katz-Rogozhnikov, Lu, Sharma & Squillante, Asymptotic Optimality of Constant-Order Policies for Lost Sales Inventory Models with Large Lead Times, arXiv:1211.4063v2, p. 9, Lemma 7
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