semicor: Semi-correlations for bivariate or multivariate data set

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

Semi-correlations for bivariate or multivariate data set

Usage

semicor(bivdat,inscore=T)
semicortable(mdat,inscore=F)

Arguments

bivdat	data matrix with 2 columns
mdat	data matrix with d columns
inscore	T if bivdat or mdat has already been converted to normal scores

Value

For semicor, $ncorr: correlation of normal scores; $lcorr: lower semi-correlation of normal scores; $ucorr: upper semi-correlation of normal scores.

For semicortable, d*(d-1)/2 by 6 table with columns j1,j2,ncorr,lcorr,ucorr,bvnsemic

See Also

bvnsemic rhoNsemic tailweightedDepmeas

Examples

set.seed(1234)
n=300
x=matrix(rnorm(2*n),n,2)
rho=0.6
x[,2]=rho*x[,1]+sqrt(1-rho^2)*x[,2]
semicor(x,inscore=TRUE)
semicor(x,inscore=FALSE)
bvnsemic(rho)  # theoretical value
# log return data
data(euro0306)
semicortable(euro0306[,2:6],inscore=FALSE)

YafeiXu/CopulaModel documentation built on May 9, 2019, 11:07 p.m.