Conditional kernels as witnesses of stochastic increasingness #
theorem
ProbabilityTheory.Copula.cdf_eq_integral_conditionalCDF
(C : Copula 2)
(u v : ↑unitInterval)
:
Disintegration along the first coordinate recovers the copula CDF.
theorem
ProbabilityTheory.Copula.cdf_eq_integral_kernel
(C : Copula 2)
(κ : Kernel ↑unitInterval ↑unitInterval)
(hκ : ⇑C.conditionalKernel =ᵐ[MeasureTheory.volume] ⇑κ)
(u v : ↑unitInterval)
:
theorem
ProbabilityTheory.Copula.integral_Iic_add_Ioc_unit
{f : ↑unitInterval → ℝ}
(hf : MeasureTheory.Integrable f MeasureTheory.volume)
{a b : ↑unitInterval}
(hab : a ≤ b)
:
theorem
ProbabilityTheory.Copula.integral_Iic_concave_of_antitone
{f : ↑unitInterval → ℝ}
(hf : MeasureTheory.Integrable f MeasureTheory.volume)
(ha : Antitone f)
(a b c : ↑unitInterval)
(hab : a ≤ b)
(hbc : b ≤ c)
:
The indefinite integral of an antitone function satisfies the concavity chord inequality.
theorem
ProbabilityTheory.Copula.isSI_of_kernel
(C : Copula 2)
(κ : Kernel ↑unitInterval ↑unitInterval)
[IsMarkovKernel κ]
(hκ : ⇑C.conditionalKernel =ᵐ[MeasureTheory.volume] ⇑κ)
(hmono : ∀ (v : ↑unitInterval), Antitone fun (u : ↑unitInterval) => (κ u).real (Set.Iic v))
:
C.IsSI
A stochastically increasing version of the conditional kernel proves SI. The almost-everywhere identification makes this independent of null-set choices.
TP2 of a version of the conditional CDF kernel. This allows singular copulas.
Equations
- One or more equations did not get rendered due to their size.