cor2pcor: correlations to partial correlations and vice versa for vine...

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

correlations to partial correlations and vice versa for vine array (C vine, D vine or general R vine)

Usage

cor2pcor.cvine(rr)
pcor2cor.cvine(pc)
cor2pcor.dvine(rr)
pcor2cor.dvine(pc)
cor2pcor.rvine(rr,A)
pcor2cor.rvine(pc,A,byrow=T)
pcor2cor.1tr(pp,A)  # 1-truncated vine
pcor2cor.2tr(pp,A)  # 2-truncated vine
pcor2cor.3tr(pp,A,iprint=F)  # 3-truncated vine
pcor2cor.truncvine(pp,A,ntrunc,iprint=F)
vineResidVar(rmatobj,A,ntrunc)

Arguments

rr	correlation matrix, dimension d
pc	partial correlation array, dimension d, stored in C-vine or D-vine format
pp	partial correlation array, dimension d, stored in format of vine array A
byrow	TRUE if partial correlations by tree are stored by rows in pc, FALSE if partial correlation in location (j,k) of pc is ρ_{j,k;S} for a conditioning set S.
A	dxd vine array with 1:d on diagonal; only upper triangle is used
ntrunc	truncation level of vine; only first ntrunc rows of A are used
rmatobj	object from pcor2cor.truncvine() with $rmat and $phmat
iprint	print flag for intermediate results

Details

pcor2cor.1tr, pcor2cor.2tr, pcor2cor.3tr are specific versions of pcor2cor.truncvine; pcor2cor.1tr is called by pcor2cor.truncvine and the other two are absorbed into pcor2cor.truncvine.

Value

pc	from cor2pcor.cvine and cor2pcor.dvine, partial correlation array
pcobj	from cor2pcor.rvine, list with $pctree and $pcmat, the former with partial correlations by tree in rows, the latter with ρ_{j,k;S} in location (j,k).
rr	correlation matrix from pcor2cor.cvine, pcor2cor.dvine, pcor2cor.rvine
rmatobj	component $rmat for correlation matrix and $phmat for matrix of regression coefficients from pcor2cor.truncvine
psi2	vector of residual variances from vineResidVar

See Also

cor2reg

Examples

d=5
D=Dvinearray(d)
C=Cvinearray(d)
A=vnum2array(d,bnum=3)  
# bnum in 0 to 2^dpow-1 where dpow=2^dcase, dcase=(d-2)*(d-3)/2
rr=toeplitz(c(1,.5,.25,.125,.05))
cor2pcor.dvine(rr)
cor2pcor.rvine(rr,D)
cor2pcor.cvine(rr)
cor2pcor.rvine(rr,C)
pcobj=cor2pcor.rvine(rr,A)
pcor2cor.rvine(pcobj$pctree,A)
pcor2cor.rvine(pcobj$pcmat,A,byrow=FALSE)
rmatobj=pcor2cor.truncvine(pcobj$pctree,A,ntrunc=3)
print(rmatobj$phmat)
print(rmatobj$rmat)
vineResidVar(rmatobj,A,3)

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