R/RVinePar2Beta.R

In VineCopula: Statistical Inference of Vine Copulas

#' Blomqvist's Beta Values of an R-Vine Copula Model
#'
#' This function computes the values of Blomqvist's beta corresponding to the
#' parameters of an R-vine copula model.
#'
#'
#' @param RVM An [RVineMatrix()] object. \cr Note that the Student's
#' t-copula is not allowed since the CDF of the t-copula is not implemented
#' (see [BiCopCDF()] and [BiCopPar2Beta()]).
#' @param check.pars logical; default is `TRUE`; if `FALSE`, checks
#' for family/parameter-consistency are omitted (should only be used with care).
#'
#' @return Matrix with the same structure as the family and parameter matrices
#' of the [RVineMatrix()] object `RVM` where the entries are
#' values of Blomqvist's beta corresponding to the families and parameters of
#' the R-vine copula model given by `RVM`.
#'
#' @author Ulf Schepsmeier
#'
#' @seealso [RVineMatrix()], [BiCopPar2Beta()]
#'
#' @examples
#' # define 5-dimensional R-vine tree structure matrix
#' Matrix <- c(5, 2, 3, 1, 4,
#'             0, 2, 3, 4, 1,
#'             0, 0, 3, 4, 1,
#'             0, 0, 0, 4, 1,
#'             0, 0, 0, 0, 1)
#' Matrix <- matrix(Matrix, 5, 5)
#'
#' # define R-vine pair-copula family matrix
#' family <- c(0, 1, 3, 4, 4,
#'             0, 0, 3, 4, 1,
#'             0, 0, 0, 4, 1,
#'             0, 0, 0, 0, 3,
#'             0, 0, 0, 0, 0)
#' family <- matrix(family, 5, 5)
#'
#' # define R-vine pair-copula parameter matrix
#' par <- c(0, 0.2, 0.9, 1.5, 3.9,
#'          0, 0, 1.1, 1.6, 0.9,
#'          0, 0, 0, 1.9, 0.5,
#'          0, 0, 0, 0, 4.8,
#'          0, 0, 0, 0, 0)
#' par <- matrix(par, 5, 5)
#'
#' # define second R-vine pair-copula parameter matrix
#' par2 <- matrix(0, 5, 5)
#'
#' # define RVineMatrix object
#' RVM <- RVineMatrix(Matrix = Matrix, family = family,
#'                    par = par, par2 = par2,
#'                    names = c("V1", "V2", "V3", "V4", "V5"))
#'
#' # compute the Blomqvist's beta values
#' BlomBeta <- RVinePar2Beta(RVM)
#'
RVinePar2Beta <- function(RVM, check.pars = TRUE) {

    if (is(RVM)[1] != "RVineMatrix")
        stop("'RVM' has to be an RVineMatrix object.")

    taus <- RVM$par
    n <- dim(RVM)

    for (i in 2:n) {
        for (j in 1:(i - 1)) {
            taus[i, j] <- BiCopPar2Beta(RVM$family[i, j],
                                        RVM$par[i, j],
                                        RVM$par2[i, j],
                                        check.pars = check.pars)
        }
    }

    return(taus)
}