Monotonicity of the t-EV family in the correlation parameter #
The Pickands function of the t-EV copula decreases in r on (-1,1). The derivative
of the two Student-t terms is -K t f(z_t) (w²-2rw+1) with K>0, where the balance
identity (1-t) f(z_{1-t}) = t w² f(z_t) relates the two Student-t densities.
theorem
Verification.tEVArg_hasDerivAt
(ν w : ℝ)
(hν : 0 < ν)
{r : ℝ}
(hr : r ∈ Set.Ioo (-1) 1)
:
HasDerivAt (fun (q : ℝ) => tEVArg ν q w) (tEVArgDeriv ν r w) r
The Pickands expression for fixed t, as a function of the correlation.
Equations
- One or more equations did not get rendered due to their size.
Instances For
theorem
Verification.tEVPickandsExpr_antitoneOn
(ν t : ℝ)
(hν : 0 < ν)
(ht : t ∈ Set.Ioo 0 1)
:
AntitoneOn (fun (r : ℝ) => tEVPickandsExpr ν r t) (Set.Ioo (-1) 1)
theorem
Verification.tEV_pickands_eq_expr
(ν r : ℝ)
(hν : 0 < ν)
(hr : r ∈ Set.Ioo (-1) 1)
(t : ↑unitInterval)
(ht : ↑t ∈ Set.Ioo 0 1)
:
Diagonal, extremal coefficient and tails #
theorem
Verification.tEVArg_one_strictAntiOn
(ν : ℝ)
(hν : 0 < ν)
:
StrictAntiOn (fun (r : ℝ) => tEVArg ν r 1) (Set.Ioo (-1) 1)
theorem
Verification.tEV_pickands_half
(ν r : ℝ)
(hν : 0 < ν)
(hr : r ∈ Set.Ioo (-1) 1)
:
copulaPickands (tEV ν r hν ⋯) ProbabilityTheory.Copula.unitHalf = studentTCDF (ν + 1) (tEVArg ν r 1)
theorem
Verification.tEV_tails
(ν r : ℝ)
(hν : 0 < ν)
(hr : r ∈ Set.Ioo (-1) 1)
:
(tEV ν r hν ⋯).HasLowerTailDependence 0 ∧ (tEV ν r hν ⋯).HasUpperTailDependence (2 - 2 * studentTCDF (ν + 1) (tEVArg ν r 1))