Nothing
R/BiCopPar2Tau.R
In VineCopula: Statistical Inference of Vine Copulas
Defines functions
calcTau
frankTau
BiCopPar2Tau
Documented in
BiCopPar2Tau
#' Kendall's Tau Value of a Bivariate Copula
#'
#' This function computes the theoretical Kendall's tau value of a bivariate
#' copula for given parameter values.
#'
#' If the family and parameter specification is stored in a [BiCop()]
#' object `obj`, the alternative version \cr
#' \preformatted{BiCopPar2Tau(obj)} can be used.
#'
#' @param family integer; single number or vector of size `m`; defines the
#' bivariate copula family: \cr
#' `0` = independence copula \cr
#' `1` = Gaussian copula \cr
#' `2` = Student t copula (t-copula) \cr
#' `3` = Clayton copula \cr
#' `4` = Gumbel copula \cr
#' `5` = Frank copula \cr
#' `6` = Joe copula \cr
#' `7` = BB1 copula \cr
#' `8` = BB6 copula \cr
#' `9` = BB7 copula \cr
#' `10` = BB8 copula \cr
#' `13` = rotated Clayton copula (180 degrees; ``survival Clayton'') \cr
#' `14` = rotated Gumbel copula (180 degrees; ``survival Gumbel'') \cr
#' `16` = rotated Joe copula (180 degrees; ``survival Joe'') \cr
#' `17` = rotated BB1 copula (180 degrees; ``survival BB1'')\cr
#' `18` = rotated BB6 copula (180 degrees; ``survival BB6'')\cr
#' `19` = rotated BB7 copula (180 degrees; ``survival BB7'')\cr
#' `20` = rotated BB8 copula (180 degrees; ``survival BB8'')\cr
#' `23` = rotated Clayton copula (90 degrees) \cr
#' `24` = rotated Gumbel copula (90 degrees) \cr
#' `26` = rotated Joe copula (90 degrees) \cr
#' `27` = rotated BB1 copula (90 degrees) \cr
#' `28` = rotated BB6 copula (90 degrees) \cr
#' `29` = rotated BB7 copula (90 degrees) \cr
#' `30` = rotated BB8 copula (90 degrees) \cr
#' `33` = rotated Clayton copula (270 degrees) \cr
#' `34` = rotated Gumbel copula (270 degrees) \cr
#' `36` = rotated Joe copula (270 degrees) \cr
#' `37` = rotated BB1 copula (270 degrees) \cr
#' `38` = rotated BB6 copula (270 degrees) \cr
#' `39` = rotated BB7 copula (270 degrees) \cr
#' `40` = rotated BB8 copula (270 degrees) \cr
#' `104` = Tawn type 1 copula \cr
#' `114` = rotated Tawn type 1 copula (180 degrees) \cr
#' `124` = rotated Tawn type 1 copula (90 degrees) \cr
#' `134` = rotated Tawn type 1 copula (270 degrees) \cr
#' `204` = Tawn type 2 copula \cr
#' `214` = rotated Tawn type 2 copula (180 degrees) \cr
#' `224` = rotated Tawn type 2 copula (90 degrees) \cr
#' `234` = rotated Tawn type 2 copula (270 degrees) \cr
#' @param par numeric; single number or vector of size `n`; copula
#' parameter.
#' @param par2 numeric; single number or vector of size `n`; second
#' parameter for bivariate copulas with two parameters (t, BB1, BB6, BB7, BB8,
#' Tawn type 1 and type 2; default: `par2 = 0`). Note that the degrees of
#' freedom parameter of the t-copula does not need to be set, because the
#' theoretical Kendall's tau value of the t-copula is independent of this
#' choice.
#' @param obj `BiCop` object containing the family and parameter
#' specification.
#' @param check.pars logical; default is `TRUE`; if `FALSE`, checks
#' for family/parameter-consistency are omitted (should only be used with
#' care).
#' @return Theoretical value of Kendall's tau (vector) corresponding to the
#' bivariate copula `family` and parameter vector \eqn{(\theta, \delta) =}
#' `(par, par2)`.
#' \tabular{ll}{ No. (`family`) \tab Kendall's tau (`tau`) \cr
#' `1, 2` \tab \eqn{\frac{2}{\pi}\arcsin(\theta)}{2 / \pi arcsin(\theta)} \cr
#' `3, 13` \tab \eqn{\frac{\theta}{\theta+2}}{\theta / (\theta+2)} \cr
#' `4, 14` \tab \eqn{1-\frac{1}{\theta}}{1-1/\theta} \cr
#' `5` \tab \eqn{1-\frac{4}{\theta}+4\frac{D_1(\theta)}{\theta}}{1-4/\theta +
#' 4 D_1(\theta)/\theta} \cr
#' \tab with \eqn{D_1(\theta)=\int_0^\theta \frac{x/\theta}{\exp(x)-1}dx}{D_1(\theta)=
#' \int_0^\theta (x/\theta)/(exp(x)-1)dx} (Debye function) \cr
#' `6, 16` \tab \eqn{1+\frac{4}{\theta^2}\int_0^1
#' x\log(x)(1-x)^{2(1-\theta)/\theta}dx}{1+4/\theta^2\int_0^1
#' x\log(x)(1-x)^{2(1-\theta)/\theta}dx} \cr
#' `7, 17` \tab \eqn{1-\frac{2}{\delta(\theta+2)}}{1-2/(\delta(\theta+2))} \cr
#' `8, 18` \tab \eqn{1+4\int_0^1 -\log(-(1-t)^\theta+1)
#' (1-t-(1-t)^{-\theta}+(1-t)^{-\theta}t)/(\delta\theta) dt} \cr
#' `9, 19` \tab \eqn{1+4\int_0^1 ( (1-(1-t)^{\theta})^{-\delta} - 1)
#' /( -\theta\delta(1-t)^{\theta-1}(1-(1-t)^{\theta})^{-\delta-1} ) dt} \cr
#' `10, 20` \tab \eqn{1+4\int_0^1
#' -\log \left(((1-t\delta)^\theta-1)/((1-\delta)^\theta-1) \right) } \cr
#' \tab \eqn{* (1-t\delta-(1-t\delta)^{-\theta}+(1-t\delta)^{-\theta}t\delta)/(\theta\delta) dt} \cr
#' `23, 33` \tab \eqn{\frac{\theta}{2-\theta}}{\theta/(2-\theta)} \cr
#' `24, 34` \tab \eqn{-1-\frac{1}{\theta}}{-1-1/\theta} \cr
#' `26, 36` \tab \eqn{-1-\frac{4}{\theta^2}\int_0^1
#' x\log(x)(1-x)^{-2(1+\theta)/\theta}dx}{-1-4/\theta^2
#' \int_0^1 x\log(x)(1-x)^{-2(1+\theta)/\theta}dx} \cr
#' `27, 37` \tab \eqn{-1-\frac{2}{\delta(2-\theta)}}{1-2/(\delta(\theta+2))} \cr
#' `28, 38` \tab \eqn{-1-4\int_0^1 -\log(-(1-t)^{-\theta}+1)
#' (1-t-(1-t)^{\theta}+(1-t)^{\theta}t)/(\delta\theta) dt} \cr
#' `29, 39` \tab \eqn{-1-4\int_0^1 ( (1-(1-t)^{-\theta})^{\delta} - 1)
#' /( -\theta\delta(1-t)^{-\theta-1}(1-(1-t)^{-\theta})^{\delta-1} ) dt} \cr
#' `30, 40` \tab \eqn{-1-4\int_0^1 -\log
#' \left( ((1+t\delta)^{-\theta}-1)/((1+\delta)^{-\theta}-1) \right)} \cr
#' \tab \eqn{* (1+t\delta-(1+t\delta)^{\theta}-(1+t\delta)^{\theta}t\delta)/(\theta\delta) dt} \cr
#' `104,114` \tab \eqn{\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)t+[(\delta(1-t))^{\theta}+t^{\theta}]^{1/\theta}} \cr
#' `204,214` \tab \eqn{\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)(1-t)+[(1-t)^{-\theta}+(\delta t)^{-\theta}]^{-1/\theta}} \cr
#' `124,134` \tab \eqn{-\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)t+[(\delta(1-t))^{-\theta}+t^{-\theta}]^{-1/\theta}} \cr
#' `224,234` \tab \eqn{-\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)(1-t)+[(1-t)^{-\theta}+(\delta t)^{-\theta}]^{-1/\theta}} \cr
#'
#' }
#'
#' @note The number `n` can be chosen arbitrarily, but must agree across
#' arguments.
#'
#' @author Ulf Schepsmeier, Tobias Erhardt
#'
#' @seealso [BiCopTau2Par()], [BiCop()]
#'
#' @references Joe, H. (1997). Multivariate Models and Dependence Concepts.
#' Chapman and Hall, London.
#'
#' Czado, C., U. Schepsmeier, and A. Min (2012). Maximum likelihood estimation
#' of mixed C-vines with application to exchange rates. Statistical Modelling,
#' 12(3), 229-255.
#'
#' @examples
#' ## Example 1: Gaussian copula
#' tau0 <- 0.5
#' rho <- BiCopTau2Par(family = 1, tau = tau0)
#' # transform back
#' tau <- BiCopPar2Tau(family = 1, par = rho)
#' tau - 2/pi*asin(rho)
#'
#' ## Example 2:
#' vpar <- seq(from = 1.1, to = 10, length.out = 100)
#' tauC <- BiCopPar2Tau(family = 3, par = vpar)
#' tauG <- BiCopPar2Tau(family = 4, par = vpar)
#' tauF <- BiCopPar2Tau(family = 5, par = vpar)
#' tauJ <- BiCopPar2Tau(family = 6, par = vpar)
#' plot(tauC ~ vpar, type = "l", ylim = c(0,1))
#' lines(tauG ~ vpar, col = 2)
#' lines(tauF ~ vpar, col = 3)
#' lines(tauJ ~ vpar, col = 4)
#'
#' ## Example 3: different copula families
#' theta <- BiCopTau2Par(family = c(3,4,6), tau = c(0.4, 0.5, 0.6))
#' BiCopPar2Tau(family = c(3,4,6), par = theta)
#'
#' \dontshow{
#' # Test BiCopPar2Tau (one parametric families)
#' theta <- BiCopTau2Par(family = 0, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 0, par = theta)
#' theta <- BiCopTau2Par(family = 1, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 1, par = theta)
#' theta <- BiCopTau2Par(family = 3, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 3, par = theta)
#' theta <- BiCopTau2Par(family = 4, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 4, par = theta)
#' theta <- BiCopTau2Par(family = 5, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 5, par = theta)
#' theta <- BiCopTau2Par(family = 6, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 6, par = theta)
#' theta <- BiCopTau2Par(family = 13, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 13, par = theta)
#' theta <- BiCopTau2Par(family = 14, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 14, par = theta)
#' theta <- BiCopTau2Par(family = 16, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 16, par = theta)
#' theta <- BiCopTau2Par(family = 23, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 23, par = theta)
#' theta <- BiCopTau2Par(family = 24, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 24, par = theta)
#' theta <- BiCopTau2Par(family = 26, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 26, par = theta)
#' theta <- BiCopTau2Par(family = 33, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 33, par = theta)
#' theta <- BiCopTau2Par(family = 34, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 34, par = theta)
#' theta <- BiCopTau2Par(family = 36, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 36, par = theta)
#' theta <- BiCopTau2Par(family = 41, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 41, par = theta)
#' theta <- BiCopTau2Par(family = 51, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 51, par = theta)
#' theta <- BiCopTau2Par(family = 61, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 61, par = theta)
#' theta <- BiCopTau2Par(family = 71, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 71, par = theta)
#' theta <- BiCopTau2Par(family = 41, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 41, par = theta)
#' theta <- BiCopTau2Par(family = 51, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 51, par = theta)
#' theta <- BiCopTau2Par(family = 61, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 61, par = theta)
#' theta <- BiCopTau2Par(family = 71, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 71, par = theta)
#'
#' # Test BiCopPar2Tau (two parametric families)
#' theta <- BiCopTau2Par(family = 2, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 2, par = theta)
#' theta <- 1:3
#' delta <- 1:3
#' BiCopPar2Tau(family = 7, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 17, par = theta, par2 = delta)
#' theta <- -(1:3)
#' delta <- -(1:3)
#' BiCopPar2Tau(family = 27, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 37, par = theta, par2 = delta)
#' theta <- 2:4
#' delta <- 1:3
#' BiCopPar2Tau(family = 8, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 18, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- -(1:3)
#' BiCopPar2Tau(family = 28, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 38, par = theta, par2 = delta)
#' theta <- 1:3
#' delta <- 1:3
#' BiCopPar2Tau(family = 9, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 19, par = theta, par2 = delta)
#' theta <- -(1:3)
#' delta <- -(1:3)
#' BiCopPar2Tau(family = 29, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 39, par = theta, par2 = delta)
#' theta <- 2:4
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 10, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 20, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- -c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 30, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 40, par = theta, par2 = delta)
#'
#' theta <- 2:4
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 104, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 114, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 124, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 134, par = theta, par2 = delta)
#'
#' theta <- 2:4
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 204, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 214, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 224, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 234, par = theta, par2 = delta)
#' }
#'
BiCopPar2Tau <- function(family, par, par2 = 0, obj = NULL, check.pars = TRUE) {
# fix for SemiParBIVProbit package
dims <- set_dims(family, par, par2)
# set arbitrary par2 for t-copula
if (!inherits(family, "BiCop"))
par2[family == 2] <- par2[family == 2] + 4
## preprocessing of arguments
args <- preproc(c(as.list(environment()), call = match.call()),
extract_from_BiCop,
match_spec_lengths,
check_fam_par)
list2env(args, environment())
## calculate Kendall's tau
out <- vapply(1:length(par),
function(i) calcTau(family[i], par[i], par2[i]),
numeric(1))
## return result
if (length(dims) > 1)
out <- array(out, dim = dims)
out
}
frankParGrid <- c(-10^(5:2), seq(-36, 36, l = 100), 10^(2:5))
# frankTauVals <- 1 - 4/frankParGrid + 4/frankParGrid * copula::debye1(frankParGrid)
frankTauVals <- c(-0.99996000, -0.99960007, -0.99600658, -0.96065797, -0.89396585, -0.89188641, -0.88972402, -0.88747361,
-0.88512973, -0.88268644, -0.88013730, -0.87747533, -0.87469288, -0.87178163, -0.86873246, -0.86553538,
-0.86217942, -0.85865252, -0.85494133, -0.85103114, -0.84690560, -0.84254657, -0.83793380, -0.83304468,
-0.82785385, -0.82233280, -0.81644934, -0.81016705, -0.80344454, -0.79623459, -0.78848313, -0.78012801,
-0.77109744, -0.76130821, -0.75066334, -0.73904940, -0.72633308, -0.71235706, -0.69693500, -0.67984550,
-0.66082501, -0.63955977, -0.61567712, -0.58873731, -0.55822774, -0.52356346, -0.48410024, -0.43916961,
-0.38814802, -0.33057147, -0.26629772, -0.19569472, -0.11979813, -0.04035073, 0.04035073, 0.11979813,
0.19569472, 0.26629772, 0.33057147, 0.38814802, 0.43916961, 0.48410024, 0.52356346, 0.55822774,
0.58873731, 0.61567712, 0.63955977, 0.66082501, 0.67984550, 0.69693500, 0.71235706, 0.72633308,
0.73904940, 0.75066334, 0.76130821, 0.77109744, 0.78012801, 0.78848313, 0.79623459, 0.80344454,
0.81016705, 0.81644934, 0.82233280, 0.82785385, 0.83304468, 0.83793380, 0.84254657, 0.84690560,
0.85103114, 0.85494133, 0.85865252, 0.86217942, 0.86553538, 0.86873246, 0.87178163, 0.87469288,
0.87747533, 0.88013730, 0.88268644, 0.88512973, 0.88747361, 0.88972402, 0.89188641, 0.89396585,
0.96065797, 0.99600658, 0.99960007, 0.99996000)
frankTau <- function(par) {
approx(x = frankParGrid, y = frankTauVals, xout = par)$y
}
calcTau <- function(family, par, par2) {
## calculation of tau(s) depending on pair-copula family
if (family == 0) {
tau <- rep(0, times = length(par))
} else if (family == 1 | family == 2) {
tau <- 2/pi * asin(par)
} else if (family == 3 || family == 13) {
tau <- par/(par + 2)
} else if (family == 4 || family == 14) {
tau <- 1 - 1/par
} else if (family == 5) {
tau <- frankTau(par)
} else if (family == 6 || family == 16) {
# tau = 1 + 4/par^2 * integrate(function(x) log(x)*x*(1-x)^(2*(1-par)/par), 0,
# 1)$value
param1 <- 2/par + 1
tem <- digamma(2) - digamma(param1)
tau <- 1 + tem * 2/(2 - par)
tau[par == 2] <- 1 - trigamma(2)
} else if (family == 7 || family == 17) {
theta <- par
delta <- par2
tau <- 1 - 2/(delta * (theta + 2))
} else if (family == 8 || family == 18) {
theta <- par
delta <- par2
kt <- function(t, th, de) {
-log(-(1 - t)^th + 1) * (1 - t - (1 - t)^(-th) + (1 - t)^(-th) * t)/(de * th)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 9 || family == 19) {
theta <- par
delta <- par2
kt <- function(t, th, de) {
((1 - (1 - t)^th)^-de - 1)/(-th * de * (1 - t)^(th - 1) * (1 - (1 - t)^th)^(-de - 1))
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) kt(t, th = theta, de = delta), 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 10 || family == 20) {
theta <- par
delta <- par2
kt <- function(t, th, de) {
-log(((1 - t * de)^th - 1)/((1 - de)^th - 1)) * (1 - t * de - (1 - t * de)^(-th) + (1 - t * de)^(-th) * t * de)/(th * de)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 23 || family == 33) {
tau <- par/(-par + 2)
} else if (family == 24 || family == 34) {
tau <- -1 - 1/par
} else if (family == 26 || family == 36) {
theta <- -par
param1 <- 2/theta + 1
tem <- digamma(2) - digamma(param1)
tau <- 1 + tem * 2/(2 - theta)
tau[theta == 2] <- 1 - trigamma(2)
tau <- -tau
} else if (family == 27 || family == 37) {
theta <- -par
delta <- -par2
tau <- 1 - 2/(delta * (theta + 2))
tau <- -tau
} else if (family == 28 || family == 38) {
theta <- -par
delta <- -par2
kt <- function(t, th, de) {
-log(-(1 - t)^th + 1) * (1 - t - (1 - t)^(-th) + (1 - t)^(-th) * t)/(de * th)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
} else if (family == 29 || family == 39) {
theta <- -par
delta <- -par2
kt <- function(t, th, de) {
((1 - (1 - t)^th)^(-de) - 1)/(-th * de * (1 - t)^(th - 1) * (1 - (1 - t)^th)^(-de - 1))
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
} else if (family == 30 || family == 40) {
theta <- -par
delta <- -par2
kt <- function(t, th, de) {
-log(((1 - t * de)^th - 1)/((1 - de)^th - 1)) * (1 - t * de - (1 - t * de)^(-th) + (1 - t * de)^(-th) * t * de)/(th * de)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
} else if (family == 41 || family == 51) {
de <- par
ln2 <- log(2)
tem <- (2 - 2 * de) * ln2 + lgamma(2 * de) - 2 * lgamma(1 + de)
tau <- 1 - de * exp(tem)
} else if (family == 61 || family == 71) {
de <- -par
ln2 <- log(2)
tem <- (2 - 2 * de) * ln2 + lgamma(2 * de) - 2 * lgamma(1 + de)
tau <- 1 - de * exp(tem)
tau <- -tau
} else if (family == 42) {
tau <- (75 * par2 - par2^2 + par * (25 - par2))/450
} else if (family == 104 || family == 114 || family == 204 || family == 214) {
par3 <- 1
tau_int <- function(t, th, de) {
Afunc <- .C("Tawn2",
as.double(t),
as.integer(length(t)),
as.double(th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
Afunc2Deriv <- .C("d2Tawn",
as.double(t),
as.integer(length(t)),
as.double(th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
(t * (1 - t)) * Afunc2Deriv/Afunc
}
tau <- try(mapply(function(par, par2) {
integrate(function(t) {
tau_int(t, th = par, de = par2)
}, 0, 1)$value
}, par, par2), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 124 || family == 134 || family == 224 || family == 234) {
par3 <- 1
tau_int <- function(t, th, de) {
Afunc <- .C("Tawn2",
as.double(t),
as.integer(length(t)),
as.double(-th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
Afunc2Deriv <- .C("d2Tawn",
as.double(t),
as.integer(length(t)),
as.double(-th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
(t * (1 - t)) * Afunc2Deriv/Afunc
}
tau <- try(mapply(function(par, par2) {
integrate(function(t) {
tau_int(t, th = par, de = par2)
}, 0, 1)$value
}, par, par2), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
}
## return result
tau
}
Try the VineCopula package in your browser
Any scripts or data that you put into this service are public.
VineCopula documentation built on July 26, 2023, 5:23 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions calcTau frankTau BiCopPar2Tau
Documented in BiCopPar2Tau
#' Kendall's Tau Value of a Bivariate Copula
#'
#' This function computes the theoretical Kendall's tau value of a bivariate
#' copula for given parameter values.
#'
#' If the family and parameter specification is stored in a [BiCop()]
#' object `obj`, the alternative version \cr
#' \preformatted{BiCopPar2Tau(obj)} can be used.
#'
#' @param family integer; single number or vector of size `m`; defines the
#' bivariate copula family: \cr
#' `0` = independence copula \cr
#' `1` = Gaussian copula \cr
#' `2` = Student t copula (t-copula) \cr
#' `3` = Clayton copula \cr
#' `4` = Gumbel copula \cr
#' `5` = Frank copula \cr
#' `6` = Joe copula \cr
#' `7` = BB1 copula \cr
#' `8` = BB6 copula \cr
#' `9` = BB7 copula \cr
#' `10` = BB8 copula \cr
#' `13` = rotated Clayton copula (180 degrees; ``survival Clayton'') \cr
#' `14` = rotated Gumbel copula (180 degrees; ``survival Gumbel'') \cr
#' `16` = rotated Joe copula (180 degrees; ``survival Joe'') \cr
#' `17` = rotated BB1 copula (180 degrees; ``survival BB1'')\cr
#' `18` = rotated BB6 copula (180 degrees; ``survival BB6'')\cr
#' `19` = rotated BB7 copula (180 degrees; ``survival BB7'')\cr
#' `20` = rotated BB8 copula (180 degrees; ``survival BB8'')\cr
#' `23` = rotated Clayton copula (90 degrees) \cr
#' `24` = rotated Gumbel copula (90 degrees) \cr
#' `26` = rotated Joe copula (90 degrees) \cr
#' `27` = rotated BB1 copula (90 degrees) \cr
#' `28` = rotated BB6 copula (90 degrees) \cr
#' `29` = rotated BB7 copula (90 degrees) \cr
#' `30` = rotated BB8 copula (90 degrees) \cr
#' `33` = rotated Clayton copula (270 degrees) \cr
#' `34` = rotated Gumbel copula (270 degrees) \cr
#' `36` = rotated Joe copula (270 degrees) \cr
#' `37` = rotated BB1 copula (270 degrees) \cr
#' `38` = rotated BB6 copula (270 degrees) \cr
#' `39` = rotated BB7 copula (270 degrees) \cr
#' `40` = rotated BB8 copula (270 degrees) \cr
#' `104` = Tawn type 1 copula \cr
#' `114` = rotated Tawn type 1 copula (180 degrees) \cr
#' `124` = rotated Tawn type 1 copula (90 degrees) \cr
#' `134` = rotated Tawn type 1 copula (270 degrees) \cr
#' `204` = Tawn type 2 copula \cr
#' `214` = rotated Tawn type 2 copula (180 degrees) \cr
#' `224` = rotated Tawn type 2 copula (90 degrees) \cr
#' `234` = rotated Tawn type 2 copula (270 degrees) \cr
#' @param par numeric; single number or vector of size `n`; copula
#' parameter.
#' @param par2 numeric; single number or vector of size `n`; second
#' parameter for bivariate copulas with two parameters (t, BB1, BB6, BB7, BB8,
#' Tawn type 1 and type 2; default: `par2 = 0`). Note that the degrees of
#' freedom parameter of the t-copula does not need to be set, because the
#' theoretical Kendall's tau value of the t-copula is independent of this
#' choice.
#' @param obj `BiCop` object containing the family and parameter
#' specification.
#' @param check.pars logical; default is `TRUE`; if `FALSE`, checks
#' for family/parameter-consistency are omitted (should only be used with
#' care).
#' @return Theoretical value of Kendall's tau (vector) corresponding to the
#' bivariate copula `family` and parameter vector \eqn{(\theta, \delta) =}
#' `(par, par2)`.
#' \tabular{ll}{ No. (`family`) \tab Kendall's tau (`tau`) \cr
#' `1, 2` \tab \eqn{\frac{2}{\pi}\arcsin(\theta)}{2 / \pi arcsin(\theta)} \cr
#' `3, 13` \tab \eqn{\frac{\theta}{\theta+2}}{\theta / (\theta+2)} \cr
#' `4, 14` \tab \eqn{1-\frac{1}{\theta}}{1-1/\theta} \cr
#' `5` \tab \eqn{1-\frac{4}{\theta}+4\frac{D_1(\theta)}{\theta}}{1-4/\theta +
#' 4 D_1(\theta)/\theta} \cr
#' \tab with \eqn{D_1(\theta)=\int_0^\theta \frac{x/\theta}{\exp(x)-1}dx}{D_1(\theta)=
#' \int_0^\theta (x/\theta)/(exp(x)-1)dx} (Debye function) \cr
#' `6, 16` \tab \eqn{1+\frac{4}{\theta^2}\int_0^1
#' x\log(x)(1-x)^{2(1-\theta)/\theta}dx}{1+4/\theta^2\int_0^1
#' x\log(x)(1-x)^{2(1-\theta)/\theta}dx} \cr
#' `7, 17` \tab \eqn{1-\frac{2}{\delta(\theta+2)}}{1-2/(\delta(\theta+2))} \cr
#' `8, 18` \tab \eqn{1+4\int_0^1 -\log(-(1-t)^\theta+1)
#' (1-t-(1-t)^{-\theta}+(1-t)^{-\theta}t)/(\delta\theta) dt} \cr
#' `9, 19` \tab \eqn{1+4\int_0^1 ( (1-(1-t)^{\theta})^{-\delta} - 1)
#' /( -\theta\delta(1-t)^{\theta-1}(1-(1-t)^{\theta})^{-\delta-1} ) dt} \cr
#' `10, 20` \tab \eqn{1+4\int_0^1
#' -\log \left(((1-t\delta)^\theta-1)/((1-\delta)^\theta-1) \right) } \cr
#' \tab \eqn{* (1-t\delta-(1-t\delta)^{-\theta}+(1-t\delta)^{-\theta}t\delta)/(\theta\delta) dt} \cr
#' `23, 33` \tab \eqn{\frac{\theta}{2-\theta}}{\theta/(2-\theta)} \cr
#' `24, 34` \tab \eqn{-1-\frac{1}{\theta}}{-1-1/\theta} \cr
#' `26, 36` \tab \eqn{-1-\frac{4}{\theta^2}\int_0^1
#' x\log(x)(1-x)^{-2(1+\theta)/\theta}dx}{-1-4/\theta^2
#' \int_0^1 x\log(x)(1-x)^{-2(1+\theta)/\theta}dx} \cr
#' `27, 37` \tab \eqn{-1-\frac{2}{\delta(2-\theta)}}{1-2/(\delta(\theta+2))} \cr
#' `28, 38` \tab \eqn{-1-4\int_0^1 -\log(-(1-t)^{-\theta}+1)
#' (1-t-(1-t)^{\theta}+(1-t)^{\theta}t)/(\delta\theta) dt} \cr
#' `29, 39` \tab \eqn{-1-4\int_0^1 ( (1-(1-t)^{-\theta})^{\delta} - 1)
#' /( -\theta\delta(1-t)^{-\theta-1}(1-(1-t)^{-\theta})^{\delta-1} ) dt} \cr
#' `30, 40` \tab \eqn{-1-4\int_0^1 -\log
#' \left( ((1+t\delta)^{-\theta}-1)/((1+\delta)^{-\theta}-1) \right)} \cr
#' \tab \eqn{* (1+t\delta-(1+t\delta)^{\theta}-(1+t\delta)^{\theta}t\delta)/(\theta\delta) dt} \cr
#' `104,114` \tab \eqn{\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)t+[(\delta(1-t))^{\theta}+t^{\theta}]^{1/\theta}} \cr
#' `204,214` \tab \eqn{\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)(1-t)+[(1-t)^{-\theta}+(\delta t)^{-\theta}]^{-1/\theta}} \cr
#' `124,134` \tab \eqn{-\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)t+[(\delta(1-t))^{-\theta}+t^{-\theta}]^{-1/\theta}} \cr
#' `224,234` \tab \eqn{-\int_0^1 \frac{t(1-t)A^{\prime\prime}(t)}{A(t)}dt} \cr
#' \tab with \eqn{A(t) = (1-\delta)(1-t)+[(1-t)^{-\theta}+(\delta t)^{-\theta}]^{-1/\theta}} \cr
#'
#' }
#'
#' @note The number `n` can be chosen arbitrarily, but must agree across
#' arguments.
#'
#' @author Ulf Schepsmeier, Tobias Erhardt
#'
#' @seealso [BiCopTau2Par()], [BiCop()]
#'
#' @references Joe, H. (1997). Multivariate Models and Dependence Concepts.
#' Chapman and Hall, London.
#'
#' Czado, C., U. Schepsmeier, and A. Min (2012). Maximum likelihood estimation
#' of mixed C-vines with application to exchange rates. Statistical Modelling,
#' 12(3), 229-255.
#'
#' @examples
#' ## Example 1: Gaussian copula
#' tau0 <- 0.5
#' rho <- BiCopTau2Par(family = 1, tau = tau0)
#' # transform back
#' tau <- BiCopPar2Tau(family = 1, par = rho)
#' tau - 2/pi*asin(rho)
#'
#' ## Example 2:
#' vpar <- seq(from = 1.1, to = 10, length.out = 100)
#' tauC <- BiCopPar2Tau(family = 3, par = vpar)
#' tauG <- BiCopPar2Tau(family = 4, par = vpar)
#' tauF <- BiCopPar2Tau(family = 5, par = vpar)
#' tauJ <- BiCopPar2Tau(family = 6, par = vpar)
#' plot(tauC ~ vpar, type = "l", ylim = c(0,1))
#' lines(tauG ~ vpar, col = 2)
#' lines(tauF ~ vpar, col = 3)
#' lines(tauJ ~ vpar, col = 4)
#'
#' ## Example 3: different copula families
#' theta <- BiCopTau2Par(family = c(3,4,6), tau = c(0.4, 0.5, 0.6))
#' BiCopPar2Tau(family = c(3,4,6), par = theta)
#'
#' \dontshow{
#' # Test BiCopPar2Tau (one parametric families)
#' theta <- BiCopTau2Par(family = 0, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 0, par = theta)
#' theta <- BiCopTau2Par(family = 1, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 1, par = theta)
#' theta <- BiCopTau2Par(family = 3, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 3, par = theta)
#' theta <- BiCopTau2Par(family = 4, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 4, par = theta)
#' theta <- BiCopTau2Par(family = 5, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 5, par = theta)
#' theta <- BiCopTau2Par(family = 6, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 6, par = theta)
#' theta <- BiCopTau2Par(family = 13, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 13, par = theta)
#' theta <- BiCopTau2Par(family = 14, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 14, par = theta)
#' theta <- BiCopTau2Par(family = 16, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 16, par = theta)
#' theta <- BiCopTau2Par(family = 23, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 23, par = theta)
#' theta <- BiCopTau2Par(family = 24, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 24, par = theta)
#' theta <- BiCopTau2Par(family = 26, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 26, par = theta)
#' theta <- BiCopTau2Par(family = 33, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 33, par = theta)
#' theta <- BiCopTau2Par(family = 34, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 34, par = theta)
#' theta <- BiCopTau2Par(family = 36, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 36, par = theta)
#' theta <- BiCopTau2Par(family = 41, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 41, par = theta)
#' theta <- BiCopTau2Par(family = 51, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 51, par = theta)
#' theta <- BiCopTau2Par(family = 61, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 61, par = theta)
#' theta <- BiCopTau2Par(family = 71, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 71, par = theta)
#' theta <- BiCopTau2Par(family = 41, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 41, par = theta)
#' theta <- BiCopTau2Par(family = 51, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 51, par = theta)
#' theta <- BiCopTau2Par(family = 61, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 61, par = theta)
#' theta <- BiCopTau2Par(family = 71, tau = -c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 71, par = theta)
#'
#' # Test BiCopPar2Tau (two parametric families)
#' theta <- BiCopTau2Par(family = 2, tau = c(0.4,0.5,0.6))
#' BiCopPar2Tau(family = 2, par = theta)
#' theta <- 1:3
#' delta <- 1:3
#' BiCopPar2Tau(family = 7, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 17, par = theta, par2 = delta)
#' theta <- -(1:3)
#' delta <- -(1:3)
#' BiCopPar2Tau(family = 27, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 37, par = theta, par2 = delta)
#' theta <- 2:4
#' delta <- 1:3
#' BiCopPar2Tau(family = 8, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 18, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- -(1:3)
#' BiCopPar2Tau(family = 28, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 38, par = theta, par2 = delta)
#' theta <- 1:3
#' delta <- 1:3
#' BiCopPar2Tau(family = 9, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 19, par = theta, par2 = delta)
#' theta <- -(1:3)
#' delta <- -(1:3)
#' BiCopPar2Tau(family = 29, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 39, par = theta, par2 = delta)
#' theta <- 2:4
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 10, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 20, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- -c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 30, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 40, par = theta, par2 = delta)
#'
#' theta <- 2:4
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 104, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 114, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 124, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 134, par = theta, par2 = delta)
#'
#' theta <- 2:4
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 204, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 214, par = theta, par2 = delta)
#' theta <- -(2:4)
#' delta <- c(0.1, 0.5, 0.9)
#' BiCopPar2Tau(family = 224, par = theta, par2 = delta)
#' BiCopPar2Tau(family = 234, par = theta, par2 = delta)
#' }
#'
BiCopPar2Tau <- function(family, par, par2 = 0, obj = NULL, check.pars = TRUE) {
# fix for SemiParBIVProbit package
dims <- set_dims(family, par, par2)
# set arbitrary par2 for t-copula
if (!inherits(family, "BiCop"))
par2[family == 2] <- par2[family == 2] + 4
## preprocessing of arguments
args <- preproc(c(as.list(environment()), call = match.call()),
extract_from_BiCop,
match_spec_lengths,
check_fam_par)
list2env(args, environment())
## calculate Kendall's tau
out <- vapply(1:length(par),
function(i) calcTau(family[i], par[i], par2[i]),
numeric(1))
## return result
if (length(dims) > 1)
out <- array(out, dim = dims)
out
}
frankParGrid <- c(-10^(5:2), seq(-36, 36, l = 100), 10^(2:5))
# frankTauVals <- 1 - 4/frankParGrid + 4/frankParGrid * copula::debye1(frankParGrid)
frankTauVals <- c(-0.99996000, -0.99960007, -0.99600658, -0.96065797, -0.89396585, -0.89188641, -0.88972402, -0.88747361,
-0.88512973, -0.88268644, -0.88013730, -0.87747533, -0.87469288, -0.87178163, -0.86873246, -0.86553538,
-0.86217942, -0.85865252, -0.85494133, -0.85103114, -0.84690560, -0.84254657, -0.83793380, -0.83304468,
-0.82785385, -0.82233280, -0.81644934, -0.81016705, -0.80344454, -0.79623459, -0.78848313, -0.78012801,
-0.77109744, -0.76130821, -0.75066334, -0.73904940, -0.72633308, -0.71235706, -0.69693500, -0.67984550,
-0.66082501, -0.63955977, -0.61567712, -0.58873731, -0.55822774, -0.52356346, -0.48410024, -0.43916961,
-0.38814802, -0.33057147, -0.26629772, -0.19569472, -0.11979813, -0.04035073, 0.04035073, 0.11979813,
0.19569472, 0.26629772, 0.33057147, 0.38814802, 0.43916961, 0.48410024, 0.52356346, 0.55822774,
0.58873731, 0.61567712, 0.63955977, 0.66082501, 0.67984550, 0.69693500, 0.71235706, 0.72633308,
0.73904940, 0.75066334, 0.76130821, 0.77109744, 0.78012801, 0.78848313, 0.79623459, 0.80344454,
0.81016705, 0.81644934, 0.82233280, 0.82785385, 0.83304468, 0.83793380, 0.84254657, 0.84690560,
0.85103114, 0.85494133, 0.85865252, 0.86217942, 0.86553538, 0.86873246, 0.87178163, 0.87469288,
0.87747533, 0.88013730, 0.88268644, 0.88512973, 0.88747361, 0.88972402, 0.89188641, 0.89396585,
0.96065797, 0.99600658, 0.99960007, 0.99996000)
frankTau <- function(par) {
approx(x = frankParGrid, y = frankTauVals, xout = par)$y
}
calcTau <- function(family, par, par2) {
## calculation of tau(s) depending on pair-copula family
if (family == 0) {
tau <- rep(0, times = length(par))
} else if (family == 1 | family == 2) {
tau <- 2/pi * asin(par)
} else if (family == 3 || family == 13) {
tau <- par/(par + 2)
} else if (family == 4 || family == 14) {
tau <- 1 - 1/par
} else if (family == 5) {
tau <- frankTau(par)
} else if (family == 6 || family == 16) {
# tau = 1 + 4/par^2 * integrate(function(x) log(x)*x*(1-x)^(2*(1-par)/par), 0,
# 1)$value
param1 <- 2/par + 1
tem <- digamma(2) - digamma(param1)
tau <- 1 + tem * 2/(2 - par)
tau[par == 2] <- 1 - trigamma(2)
} else if (family == 7 || family == 17) {
theta <- par
delta <- par2
tau <- 1 - 2/(delta * (theta + 2))
} else if (family == 8 || family == 18) {
theta <- par
delta <- par2
kt <- function(t, th, de) {
-log(-(1 - t)^th + 1) * (1 - t - (1 - t)^(-th) + (1 - t)^(-th) * t)/(de * th)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 9 || family == 19) {
theta <- par
delta <- par2
kt <- function(t, th, de) {
((1 - (1 - t)^th)^-de - 1)/(-th * de * (1 - t)^(th - 1) * (1 - (1 - t)^th)^(-de - 1))
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) kt(t, th = theta, de = delta), 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 10 || family == 20) {
theta <- par
delta <- par2
kt <- function(t, th, de) {
-log(((1 - t * de)^th - 1)/((1 - de)^th - 1)) * (1 - t * de - (1 - t * de)^(-th) + (1 - t * de)^(-th) * t * de)/(th * de)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 23 || family == 33) {
tau <- par/(-par + 2)
} else if (family == 24 || family == 34) {
tau <- -1 - 1/par
} else if (family == 26 || family == 36) {
theta <- -par
param1 <- 2/theta + 1
tem <- digamma(2) - digamma(param1)
tau <- 1 + tem * 2/(2 - theta)
tau[theta == 2] <- 1 - trigamma(2)
tau <- -tau
} else if (family == 27 || family == 37) {
theta <- -par
delta <- -par2
tau <- 1 - 2/(delta * (theta + 2))
tau <- -tau
} else if (family == 28 || family == 38) {
theta <- -par
delta <- -par2
kt <- function(t, th, de) {
-log(-(1 - t)^th + 1) * (1 - t - (1 - t)^(-th) + (1 - t)^(-th) * t)/(de * th)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
} else if (family == 29 || family == 39) {
theta <- -par
delta <- -par2
kt <- function(t, th, de) {
((1 - (1 - t)^th)^(-de) - 1)/(-th * de * (1 - t)^(th - 1) * (1 - (1 - t)^th)^(-de - 1))
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
} else if (family == 30 || family == 40) {
theta <- -par
delta <- -par2
kt <- function(t, th, de) {
-log(((1 - t * de)^th - 1)/((1 - de)^th - 1)) * (1 - t * de - (1 - t * de)^(-th) + (1 - t * de)^(-th) * t * de)/(th * de)
}
tau <- try(1 + 4 * mapply(function(theta, delta) {
integrate(function(t) {
kt(t, th = theta, de = delta)
}, 0, 1)$value
}, theta, delta), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
} else if (family == 41 || family == 51) {
de <- par
ln2 <- log(2)
tem <- (2 - 2 * de) * ln2 + lgamma(2 * de) - 2 * lgamma(1 + de)
tau <- 1 - de * exp(tem)
} else if (family == 61 || family == 71) {
de <- -par
ln2 <- log(2)
tem <- (2 - 2 * de) * ln2 + lgamma(2 * de) - 2 * lgamma(1 + de)
tau <- 1 - de * exp(tem)
tau <- -tau
} else if (family == 42) {
tau <- (75 * par2 - par2^2 + par * (25 - par2))/450
} else if (family == 104 || family == 114 || family == 204 || family == 214) {
par3 <- 1
tau_int <- function(t, th, de) {
Afunc <- .C("Tawn2",
as.double(t),
as.integer(length(t)),
as.double(th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
Afunc2Deriv <- .C("d2Tawn",
as.double(t),
as.integer(length(t)),
as.double(th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
(t * (1 - t)) * Afunc2Deriv/Afunc
}
tau <- try(mapply(function(par, par2) {
integrate(function(t) {
tau_int(t, th = par, de = par2)
}, 0, 1)$value
}, par, par2), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
} else if (family == 124 || family == 134 || family == 224 || family == 234) {
par3 <- 1
tau_int <- function(t, th, de) {
Afunc <- .C("Tawn2",
as.double(t),
as.integer(length(t)),
as.double(-th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
Afunc2Deriv <- .C("d2Tawn",
as.double(t),
as.integer(length(t)),
as.double(-th),
as.double(de),
as.double(1),
as.double(rep(0, length(t))),
PACKAGE = "VineCopula")[[6]]
(t * (1 - t)) * Afunc2Deriv/Afunc
}
tau <- try(mapply(function(par, par2) {
integrate(function(t) {
tau_int(t, th = par, de = par2)
}, 0, 1)$value
}, par, par2), silent = TRUE)
if (inherits(tau, "try-error"))
tau <- NA
tau <- -tau
}
## return result
tau
}
Try the VineCopula package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.