integral poissonChaosProd mul
ProvedLevyStochCalc.Poisson.integral_poissonChaosProd_mulcontext-reviewlevystochcalcsource-import
The source theorem LevyStochCalc.Poisson.integral_poissonChaosProd_mul is extracted from LevyStochCalc.Poisson.SimpleChaos. The original Lean proposition is retained verbatim, and its proof is tracked separately for remote certification.
Preamble
import Definitions.Def_LevyStochCalc_Poisson_SimpleChaos_F001
/-
Copyright (c) 2026 Christian Garry. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Christian Garry
-/
/-!
# Chaos elements of a finite disjoint family of regions
For a Poisson random measure `N` and a finite family `B : Fin p → Set (ℝ × E)` of regions,
`poissonChaosProd N α B` is the product over the family of the Charlier elements
`poissonChaosStep N (α k) (B k)`, one factor per index, of the degree prescribed by the
multi-degree `α : Fin p → ℕ`. When the regions are pairwise disjoint their counts are jointly
independent, so the one-region orthogonality of `Poisson/ChaosStep.lean` multiplies out: the
elements of distinct multi-degrees are orthogonal under `P`, the element of multi-degree `α` has
second moment given by the norming `∏_k α_k ! λ_k ^ α_k` with `λ_k = ν̂(B k)`, the elements of
nonzero multi-degree are centred, and every element whose multi-degree is not `Pi.single m 1` is
orthogonal to the compensated count `Ñ(B m)` of the family.
Pairwise disjointness of the family is a hypothesis throughout, and it is what the independence
of the counts rests on; overlapping regions are not covered. Each factor is a polynomial in the
one count `N(B k)`, so the elements here are polynomials in the `p` counts of the family.
## Main statements
* `LevyStochCalc.Poisson.iIndepFun_count_fin` — the counts of a finite pairwise disjoint family
of measurable regions are jointly independent.
* `LevyStochCalc.Poisson.poissonChaosProd_single_one` — the element of multi-degree
`Pi.single m 1` is the compensated count `Ñ(B m)`.
* `LevyStochCalc.Poisson.integrable_poissonChaosProd_mul`,
`LevyStochCalc.Poisson.integrable_poissonChaosProd` — an element of a pairwise disjoint family
of regions of finite intensity, and a product of two of them, are integrable.
* `LevyStochCalc.Poisson.memLp_two_poissonChaosProd` — such an element lies in `L²`.
* `LevyStochCalc.Poisson.integral_poissonChaosProd_mul` —
`E[C_α C_β] = δ_(α β) ∏_k α_k ! λ_k ^ α_k`.
* `LevyStochCalc.Poisson.integral_poissonChaosProd_eq_zero` — `E[C_α] = 0` for `α ≠ 0`.
* `LevyStochCalc.Poisson.integral_poissonChaosProd_mul_compensated_eq_zero` —
`E[C_α Ñ(B m)] = 0` for `α ≠ Pi.single m 1`.
-/
open MeasureTheory ProbabilityTheory
open scoped ENNReal NNReal BigOperators
namespace LevyStochCalc.Poisson
universe u v w
variable {Ω : Type u} [MeasurableSpace Ω] {E : Type v} [MeasurableSpace E]
{P : Measure Ω} [IsProbabilityMeasure P] {ν : Measure E} [SigmaFinite ν]
/-! ### Joint independence of the counts of a finite disjoint family -/
/-! ### The chaos element of a multi-degree -/
/-! ### Integrability -/
/-! ### Orthogonality across multi-degrees -/
Formal statement
theorem _root_.LevyStochCalc.Poisson.integral_poissonChaosProd_mul (N : PoissonRandomMeasure.{u, v, w} P ν) {p : ℕ}
{B : Fin p → Set (ℝ × E)} (hB : ∀ k, MeasurableSet (B k))
(hd : Pairwise fun k l => Disjoint (B k) (B l))
(hfin : ∀ k, referenceIntensity ν (B k) ≠ ⊤) (α β : Fin p → ℕ) :
∫ ω, poissonChaosProd N α B ω * poissonChaosProd N β B ω ∂P
= if α = β then ∏ k, ((α k).factorial : ℝ) * (referenceIntensity ν (B k)).toReal ^ α k
else 0 := by sorry
Source