The crossing counting set functions are Borel measures
ProvedExcursionCoupling.crossing_counting_functions_are_measuresLet be finite Borel measures on and let be the difference of their cumulative distribution functions. For a level let
be the sets of increasing, respectively decreasing, points of the completed graph of . Juillet's crossing measures are obtained by integrating, over all levels, the counting measure of the crossings at that level:
The assertion is that these two set functions really are Borel measures on : there exist measures whose value on every Borel set is given by the displayed integrals.
For each fixed level the set function is a counting measure, hence countably additive; the content of the statement is therefore the measurability in of the crossing counts, which lets one exchange the integral with a countable sum by monotone convergence.
Formalization note. The existential packaging is deliberate: it says exactly that the lower Lebesgue integrals below define genuine measures, without committing to a particular construction of them.
import Definitions.Def_excursion_coupling open MeasureTheory Set Function
namespace ExcursionCoupling
theorem crossing_counting_functions_are_measures
(μ ν : Measure ℝ) [IsFiniteMeasure μ] [IsFiniteMeasure ν] :
∃ ζpos ζneg : Measure ℝ,
(∀ A : Set ℝ, MeasurableSet A →
ζpos A = ∫⁻ h : ℝ, ({x ∈ A | (x, h) ∈ posPoints (Fsigma μ ν)}.encard.toENNReal)) ∧
(∀ A : Set ℝ, MeasurableSet A →
ζneg A = ∫⁻ h : ℝ, ({x ∈ A | (x, h) ∈ negPoints (Fsigma μ ν)}.encard.toENNReal)) := by
sorry
end ExcursionCoupling