copderiv: Derivatives of log copula density and copula conditional cdf
In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description Usage Arguments Details Value See Also Examples

Derivatives of log copula density and copula conditional cdf

logdfrk.deriv(u,v,cpar)  # derivatives of c_{12}(u,v,cpar)
pcondfrk.deriv(v,u,cpar) # deriv of C_{2|1}(v+u;cpar)
# also frk can be replaced with gum or bb1 or ..
logdbvtcop.deriv(u,v,param,df=dfdefault)
pcondbvtcop.deriv(v,u,param,df=dfdefault)

`u`	value in interval 0,1; could be a vector
`v`	value in interval 0,1; could be a vector
`cpar`	copula parameter: could be scalar or vector depending on the copula family
`param`	for t copula, this is either rho in (-1,1) or (rho,df) where df>0; in the former case, set dfdefault before using
`df`	global default shape parameter for t copula

These are templates. A user can add other functions like these as needed. These could be useful for R-vine negative log-likelihoods with analytic derivatives for input into nlm. Result is faster with t copulas with fixed shape parameters.

log pdf or C_{2|1} value(s) with derivatives with respect to u, v and cpar. Output is a vector of length(3+length(cpar)) with u,v are scalars; otherwise output is a matrix of nx [length(3+length(cpar))] if at least one of u or v is a vector of length n.

pcop pcond

u=seq(.1,.9,.2)
v=c(.2,.2,.4,.6,.8)
dfdefault=5
logdfrk.deriv(u,v,2)
logdbvtcop.deriv(u,v,.6)
logdbvtcop.deriv(u,v,c(.6,5))
logdbb1.deriv(u,v,c(.5,2))
pcondfrk.deriv(u,v,2)
pcondbvtcop.deriv(u,v,.6)
pcondbvtcop.deriv(u,v,c(.6,5))
pcondbb1.deriv(u,v,c(.5,2))