load2pcor: Operations of loading matrix of Gaussian factor analysis

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

Operations of loading matrix of Gaussian factor analysis

Usage

load2pcor(amat)   
pcor2load(rhmat)
grotate2(amat,row,iprint=F) # apply one Givens rotation
grotate3(amat,row1=1,row2=2,iprint=F) # apply three Givens rotations

Arguments

amat	dxp matrix of loadings for load2pcor, dx2 for grotate2, dx3 matrix for grotate3
rhmat	dxp matrix for pcor2load, correlations with factor 1 in column 1, partial correlations with factor k given previous factors in column k
row	index of row to set to 0 in second column
row1	index of row to set to 0 in second and third columns
row2	index of row to set to 0 in third column
iprint	print flag for intermediate results

Value

For load2pcor, dxp matrix with correlations with factor 1 in the first column and partial correlations with factor k given previous factors in the kth column.

For grotate2 and grotate3, a rotated loading matrix with same dimension as inputted matrix.

For pcor2load, a matrix of loadings.

Examples

d=7
be1=c(.7,.6,.7,.6,.7,.6,.7)
be2=c(.4,.4,.4,.4,.3,.3,.3)
cpar1.gum=gum.b2cpar(be1)
cpar2.gum=gum.b2cpar(be2)
n=300
set.seed(123)
gumdat=sim2fact(n,cpar1.gum,cpar2.gum,qcondgum,qcondgum,"gum","gum")
zdat=nscore(gumdat,iopt=TRUE)
rmat=cor(zdat)
out2=factanal(covmat=rmat,factors=2)
amat2=matrix(c(out2$loadings),d,2)
amat2=grotate2(amat2,row=1)
print(amat2) # 0 in (1,2) position
rhmat=load2pcor(amat2)
amat=pcor2load(rhmat)
out3=factanal(covmat=rmat,factors=3)
amat3=matrix(c(out3$loadings),d,3)
amat3=grotate3(amat3,row1=1,row2=2)
print(amat3) # 0 in (1,2), (1,3) and (2,3) positions
load2pcor(amat3)

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