rLTstochrep: Simulation of random variables from LTs used in Archimedean...

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

Simulation of random variables from LTs used in Archimedean copulas. Simulation from multivariate Archimedean copulas.

Usage

rpostable(n,alp) # 0<alp<1
rsibuya(n,alp) # 0<alp<1
rlogseries(n,cpar) # cpar as for Frank leads to better parametrization
rmitlef(n,param)
rgammaSgamma(n,param)
rpostableSgamma(n,param)
rsibuyaSpostable(n,param)
rsibuyaSgamma(n,param)
rmfrk0(n,d,cpar) # R version
rmfrk(n,d,cpar,icheck=F) # link to C
rmcop(n,d,cpar)  # choices for 'cop' for mtcj,joe,gum,bb1,bb2,bb3,bb6,bb7,bb10

Arguments

n	sample size
d	dimension for multivariate Archimedean
alp	parameter of Laplace transform (LT)
param	(vector) parameter of Laplace transform (LT)
cpar	copula parameter: could be scalar or vector depending on the copula family
icheck	flag to print out means and correlation as checks

Details

The LT families matching the Archimedean copula families are:

logseries for Frank;

gamma for MTCJ=Mardia-Takahasi-Cook-Johnson;

Sibuya for Joe;

positive stable for Gumbel;

Mittag-Leffler (or gamma stopped positive stable) for multivariate version of BB1;

gammaSgamma (gamma stopped gamma) for multivariate version of BB2;

postableSgamma (positive stable stopped gamma) for multivariate version of BB3;

sibuyaSpostable (Sibuya stopped positive stable) for multivariate version of BB6;

sibuyaSgamma (Sibuya stopped gamma) for multivariate version of BB7;

shifted negative binomial (see code) for multivariate version of BB10.

Value

vector for rpostable to rsibuyaSgamma, nxd matrix for rmfrk to rmbb10.

References

Joe H (2014). Dependence Modeling with Copulas. Chapman&Hall/CRC. See Appendix for the names of some of the LTs and the source of the algorithms.

See Also

rcop rbivcop2param

Examples

cpar=c(1,2); n=1000
r=rmitlef(n,cpar)
print(summary(r))
uu=rmbb1(n,d=3,cpar)
print(summary(uu))
print(taucor(uu[,1],uu[,2]))
print(taucor(uu[,1],uu[,3]))
print(taucor(uu[,2],uu[,3]))
tau=bb1.cpar2tau(cpar)
cat("theor.tau=",tau,"\n")

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