Extract a subset of variables from a Gaussian Copula distribution.

Extract a subset of variables from a Gaussian Copula distribution.

gcopSub(gCop, subset)

Arguments

gCop	A gaussCop object (see gcopFit).
subset	Integer or logical vector specifying which subset of variables to keep. Default is to keep all variables.

Value

A gaussCop object representing the marginal Gaussian Copula distribution on the subset of variables.

Examples

# simulate data
n <- 5e4
X <- cbind(rnorm(n, mean = 1, sd = 3),
           rnorm(n, mean=4, sd = 0.5),
           rt(n, df = 10),
           rchisq(n, df = 5),
           rnorm(n, mean=10, sd = 10))
# fit Gaussian Copula using Kernel method
gCop <- gcopFit(X, fitXD = "kernel")
#> Warning: Endpoint matching failed.  mean/sd set to empirical estimates.
# subset gCop
isub <- sample(5, 3) # subset
gCop.sub <- gcopSub(gCop, subset = isub)
# check that subsetted xDensity representations are the same
for(ii in 1:3) {
  print(identical(gCop$XDens[[isub[ii]]], gCop.sub$XDens[[ii]]))
}
#> [1] TRUE
#> [1] TRUE
#> [1] TRUE
# check that subsetted correlation matrix is the same
print(identical(gCop$Rho[isub,isub], gCop.sub$Rho))
#> [1] TRUE