Lemma 1 (Glasserman) — d/dθ E X(θ) = E dX(θ)/dθ for a Lipschitz, a.s. differentiable random function
ProvedRLPBidPrice.Unbiased.lemma1_glassermanLet be a probability space and a real random function of a real parameter , each a (strongly) measurable random variable, with integrable. Suppose satisfies the Lipschitz condition: there is a constant with
and that is differentiable at for -almost every . Then is differentiable at and
This is the interchange lemma the paper uses to pass from almost-sure differentiability of the LP value to differentiability of its expectation.
Formalization Note. The Lipschitz condition is read with universally quantified (a global condition; by symmetry it bounds ). The implicit standing assumptions of "a random function" with an expectation — measurability of each and integrability (at ; the Lipschitz bound then gives it everywhere) — are explicit hypotheses. The conclusion is stated with HasDerivAt, which asserts both that the derivative exists and that it equals the displayed expectation.
import Mathlib import Definitions.Def_RLPBidPrice_Unbiased_Model open MeasureTheory Matrix
namespace RLPBidPrice.Unbiased
theorem lemma1_glasserman {Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω)
[IsProbabilityMeasure P] (X : ℝ → Ω → ℝ) (θ₀ : ℝ)
(hmeas : ∀ θ, AEStronglyMeasurable (X θ) P)
(hint : Integrable (X θ₀) P)
(K : ℝ) (hK : 0 ≤ K)
(hlip : ∀ ω θ h, 0 ≤ h → |X (θ + h) ω - X θ ω| ≤ K * h)
(hdiff : ∀ᵐ ω ∂P, DifferentiableAt ℝ (fun θ => X θ ω) θ₀) :
HasDerivAt (fun θ => ∫ ω, X θ ω ∂P) (∫ ω, deriv (fun θ => X θ ω) θ₀ ∂P) θ₀ := by sorry
end RLPBidPrice.Unbiased
Read-back
What the Lean code literally says, in plain math · claude-opus-5-5
Let be a probability space. Let , written , be a real-valued family indexed by , and fix . Assume:
- for every , is almost-everywhere strongly measurable with respect to ;
- is -integrable;
- there is a constant such that, for every (not merely almost every), every and every ,
Because ranges over all reals, this says each is -Lipschitz on ;
- for -almost every , is differentiable at .
Conclusion: the function has a derivative at , equal to
In this integrand, the derivative is taken to be at any where it does not exist. By the last hypothesis that happens only on a null set.
Degenerate cases.
- cannot be empty, because is a probability measure.
- If , each is constant in and the asserted derivative is .
- If the pointwise-derivative function were not integrable, the right-hand integral would take the default value . The statement makes no separate integrability assertion about it.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.