Lemma 18 — finite measures correspond bijectively to consistent arrays
ProvedLambdaCoalescent.Rates.lemma_18_bijectioncoalescentexchangeable-partitionlambda-coalescentmarkov-chainp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1
Every finite nonnegative Borel measure on gives a consistent nonnegative merger-rate array by formula (1). Conversely, every consistent nonnegative array has exactly one such representing measure:
Both directions and uniqueness express the bijection stated after equation (22) in Lemma 18. The zero measure and zero array are included.
Formalization Note Measures are represented on with zero mass outside . Finiteness is explicit to ensure the integral denotes the ordinary finite-measure integral.
Preamble
import Definitions.Def_LambdaCoalescent_Rates_Setting
Formal statement
namespace LambdaCoalescent.Rates
open MeasureTheory
/-- Lemma 18, p. 1882: formula (1) is a bijection. -/
theorem lemma_18_bijection :
(∀ Λ : Measure ℝ, IsFiniteMeasure Λ → Λ (Set.Icc (0 : ℝ) 1)ᶜ = 0 →
Consistent (lamOf Λ)) ∧
(∀ lam : ℕ → ℕ → ℝ,
(∀ b k, 2 ≤ k → k ≤ b → 0 ≤ lam b k) →
Consistent lam →
∃! Λ : Measure ℝ,
IsFiniteMeasure Λ ∧ Λ (Set.Icc (0 : ℝ) 1)ᶜ = 0 ∧
∀ b k, 2 ≤ k → k ≤ b → lam b k = lamOf Λ b k) := by sorry
end LambdaCoalescent.Rates
Source
Pitman, Coalescents with multiple collisions, Ann. Probab. 27 (1999), p. 1882, Lemma 18, second sentence; eq. (1)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.