cor2reg: correlations to regression coefficients and vice versa for...

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

correlations to regression coefficients and vice versa for vine array (general R-vine); the matrix of regression coefficients provides a simple way for simulation for a multivariate normal distribution with correlation matrix with the structure of a truncated vine based on the order of variables in the vine array.

Usage

cor2reg(rr,A,iprint=F)
reg2cor(phm,A)

Arguments

rr	correlation matrix, dimension d
A	dxd vine array with 1:d on diagonal; only upper triangle is used.
phm	matrix of regression coefficients phi[j,k], dimension d
iprint	print flag for intermediate results

Value

phm	matrix of regression coefficients from cor2reg
rr	correlation matrix from reg2cor

See Also

cor2pcor

Examples

d=5
C=Cvinearray(d)
pp=matrix(0,d,d)
pp[1,2:d]=0.3
set.seed(123)
pp[1,2:d]=runif(d-1,0,1)
pp[2,3:d]=0.4
rr2=pcor2cor.cvine(pp)
print(rr2)
print(cor2reg(rr2,C,iprint=FALSE)) 
pp[3,4:d]=0.6
rr3=pcor2cor.cvine(pp)
print(rr3)
phm=cor2reg(rr3,C) 
print(phm)
print(reg2cor(phm,C))

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