chkcop: Checks for pcop pcond dcop qcond functions

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

Check pcop pcond dcop qcond functions for copula cdf, conditional cdf, pdf, conditional quantile

Usage

chkcopderiv(u,vvec,cpar,bcdf,pcond,bpdf,str=" ",eps=1.e-4)
chkcopcond(uvec,vvec,cpar,pcond,qcond,str=" ",tol=1.e-5)
chkincrease(mat)

Arguments

u	scalar in (0,1)
vvec	vector of values in (0,1)
uvec	vector of values in (0,1), same length as vvec
cpar	copula parameter
bcdf	function for bivariate copula cdf = pcop
bpdf	function for bivariate copula pdf = dcop
pcond	function for conditional copula cdf
qcond	function for conditional copula quantile
str	string for copula name
eps	step size for numerical derivative in chkcopderiv
tol	tolerance for pcond/qcond composition versus identity in chkcopcond
mat	matrix or dataframe

Value

For chkcopderiv and chkcopcond, output is printed, nothing is returned. chkcopcond checks pcond vs numerical derivative of pcop; chkcopderiv checks also dcop vs numerical derivative of pcond.

For chkincrease, a value of 1 means the values are increasing in each row, otherwise 0 is returned.

Examples

# checks cdf, conditional cdf, pdf for Plackett copula
be=.8
par.pla=pla.b2cpar(be)
u=.3
vvec=seq(.4,.9,.1)
chkcopderiv(u,vvec,par.pla,bcdf=ppla,pcond=pcondpla,bpdf=dpla,str="pla")
uvec=seq(.1,.6,.1)
vvec=seq(.4,.9,.1)
chkcopcond(uvec,vvec,par.pla,pcondpla,qcondpla,"pla")
# check monotonicity for Plackett copula
u=seq(.1,.9,.2)
umat=matrix(rep(u,5),5,5,byrow=TRUE)
vmat=matrix(rep(u,5),5,5)
cdf=ppla(umat,vmat,par.pla)
chkincrease(cdf)
# check monotonicity of dependence measure for Galambos and Plackett copulas
data(deptabder)
chkincrease(gal.deptab[,1:6]) # cols cpar beta tau rhoS rhoN lambda
apply(gal.deptab,2,diff)
chkincrease(pla.deptab[,1:5]) # cols cpar beta tau rhoS rhoN l

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