A normalized positive Laplace tilt is a probability measure
ProvedAvramDividend.Classical.normalized_laplace_tilt_is_probabilityavram-dividendbounded-variationlaplace-tiltprobability-measure
If C is the positive Laplace value at s₀, weighting a probability law by exp(-s₀ x)/C produces another probability law. This is the normalization step in the finite-cutoff Pollaczek–Khinchine derivation for target 494dbcd1-770a-49c0-986b-046103fab9c4.
Preamble
import Mathlib open MeasureTheory open scoped ENNReal
Formal statement
theorem AvramDividend.Classical.normalized_laplace_tilt_is_probability
{μ : Measure ℝ} [IsProbabilityMeasure μ]
(s₀ C : ℝ)
(hC : C = ∫ x, Real.exp (-(s₀ * x)) ∂μ)
(hCpos : 0 < C)
(hInt : Integrable (fun x : ℝ => Real.exp (-(s₀ * x))) μ) :
IsProbabilityMeasure
(μ.withDensity (fun x => ENNReal.ofReal (C⁻¹ * Real.exp (-(s₀ * x))))) := by sorrySource
Source-faithful factor of the normalized-reflected-tilt derivation in candidates/work_on/494db_laplace_ray_measure_uniqueness_draft_20261028.lean, grounded in the exact 494 target preparation at Mathlib revision 0df444a360eaa60ab8c11dca51a86af692955474. This establishes only the probability normalization identity; it does not establish the two-sided exit theorem or an authoritative parent edge.