PairCopulaAIC

Computing the AIC of a pair-copula


Purpose

The function computes the value of the AIC for a copula for a given matrix of observations u, which have to lie in the 2-dimensional unit cube, evaluated at the ML estimates. Possible pair-copula families:

       0   Indep
       1   AMH
       2   AsymFGM
       3   BB1
       4   BB6
       5   BB7
       6   BB8
       7   Clayton
       8   FGM
       9   Frank
       10  Gaussian
       11  Gumbel
       12  IteratedFGM
       13  Joe
       14  PartialFrank
       15  Plackett
       16  Tawn1
       17  Tawn2
       18  Tawn
       19  t

Usage

    [AIC,ParamHat] = PairCopulaAIC(family,u1,u2)
  Rotated pair-copulas
    [AIC,ParamHat] = PairCopulaAIC(family,u1,u2,roatation)

Inputs

   family          = The copula family.
   u1              = A (n x 1) dimensional vector of values lying in [0,1].
   u2              = A (n x 1) dimensional vector of values lying in [0,1].
   rotation        = The degree of rotation, i.e., either 90, 180 or
                     270. No rotation is achieved by letting the
                     rotation argument empty or by choosing 0
                     rotation.

Outputs

  AIC              = The value of the AIC evaluated at the ML-estimator.
  ParamHat         = The ML-estimate for the parameter vector.