Lower orthant, upper orthant, and concordance orders #
These predicates use the copula convention: smaller lower orthant probabilities
mean a smaller copula. No global LE instance selects one of the different orders.
Pointwise comparison of distribution functions.
Equations
- C.LowerOrthantLE D = ∀ (u : Fin d → ↑unitInterval), C.cdf u ≤ D.cdf u
Instances For
Pointwise comparison of upper orthant probabilities.
Equations
- C.UpperOrthantLE D = ∀ (u : Fin d → ↑unitInterval), C.survival u ≤ D.survival u
Instances For
Concordance compares both lower and upper orthants.
Equations
- C.ConcordanceLE D = (C.LowerOrthantLE D ∧ C.UpperOrthantLE D)
Instances For
theorem
ProbabilityTheory.Copula.LowerOrthantLE.trans
{d : ℕ}
{C D E : Copula d}
(h : C.LowerOrthantLE D)
(k : D.LowerOrthantLE E)
:
C.LowerOrthantLE E
theorem
ProbabilityTheory.Copula.LowerOrthantLE.antisymm
{d : ℕ}
{C D : Copula d}
(h : C.LowerOrthantLE D)
(k : D.LowerOrthantLE C)
:
theorem
ProbabilityTheory.Copula.UpperOrthantLE.trans
{d : ℕ}
{C D E : Copula d}
(h : C.UpperOrthantLE D)
(k : D.UpperOrthantLE E)
:
C.UpperOrthantLE E
theorem
ProbabilityTheory.Copula.UpperOrthantLE.antisymm
{d : ℕ}
{C D : Copula d}
(h : C.UpperOrthantLE D)
(k : D.UpperOrthantLE C)
:
theorem
ProbabilityTheory.Copula.ConcordanceLE.trans
{d : ℕ}
{C D E : Copula d}
(h : C.ConcordanceLE D)
(k : D.ConcordanceLE E)
:
C.ConcordanceLE E
theorem
ProbabilityTheory.Copula.ConcordanceLE.antisymm
{d : ℕ}
{C D : Copula d}
(h : C.ConcordanceLE D)
(k : D.ConcordanceLE C)
:
In dimension two the two orthant comparisons coincide.
theorem
ProbabilityTheory.Copula.LowerOrthantLE.mix
{d : ℕ}
{C D E F : Copula d}
(h : C.LowerOrthantLE D)
(k : E.LowerOrthantLE F)
(a : ↑unitInterval)
:
(C.mix E a).LowerOrthantLE (D.mix F a)
theorem
ProbabilityTheory.Copula.LowerOrthantLE.mix_weight
{d : ℕ}
{C D : Copula d}
(h : C.LowerOrthantLE D)
{a b : ↑unitInterval}
(hab : a ≤ b)
:
(D.mix C a).LowerOrthantLE (D.mix C b)
Increasing the weight of the larger copula increases the mixture.