Brownian inverse-Gaussian first-passage increment vector has product law (correct sample binder)
ProvedKyprianou.Ch01.inverseGaussianFirstPassageJointIncrementProductLawV2For the inverse-Gaussian Brownian first-passage clock, the joint law of the level-indexed passage-time increments along every finite nondecreasing barrier partition is the product of the individual increment laws. This is the genuine stochastic factorisation needed for independent increments. A sound proof uses strong Markov at successive finite first-passage stopping times and deterministic first-passage additivity after spatial shift. The pinned Mathlib iIndepFun_iff_map_fun_eq_pi_map lemma proves the original independent-increments axiom directly once this joint law and measurable coordinates are established. The n=0 and zero-length partition segments require no stochastic argument. V1 publication failed because the lambda 'fun ω k : Fin n' assigned type Fin n to both binders. V2 uses distinct types Ω and Fin n.
import Mathlib import Definitions.Def_Kyprianou_Ch01_McKeanBrownianStopping import Definitions.Def_Kyprianou_Ch01_ExplicitLevyExamples import Definitions.Def_LeanLevy_Processes_StochasticProcess open MeasureTheory ProbabilityTheory Set Filter open scoped NNReal ENNReal set_option autoImplicit false
theorem Kyprianou.Ch01.inverseGaussianFirstPassageJointIncrementProductLawV2
{Ω : Type*} [MeasurableSpace Ω]
(P : Measure Ω) [IsProbabilityMeasure P]
(X : ℝ≥0 → Ω → ℝ)
(b : ℝ) (hb : 0 < b)
(hB : Kyprianou.Ch01.IsLinearBrownian P X 1 b)
(hMeas : ∀ t : ℝ≥0, Measurable (X t))
(hContinuous : ∀ ω, Continuous (fun t : ℝ≥0 => X t ω)) :
let T : ℝ≥0 → Ω → ℝ :=
fun s ω =>
if s = 0 then (0 : ℝ)
else (match Kyprianou.Ch01.firstUpperPassage X (s : ℝ) ω with
| none => (0 : ℝ)
| some t => (t : ℝ))
∀ (n : ℕ) (t : Fin (n + 1) → ℝ≥0),
Monotone t →
P.map (fun (ω : Ω) (k : Fin n) =>
ProbabilityTheory.increment T (t k.castSucc) (t k.succ) ω) =
Measure.pi (fun k : Fin n =>
P.map (ProbabilityTheory.increment T (t k.castSucc) (t k.succ))) := by sorry