mdiscretenllk: negative log-likelihood of multivariate copula with discrete...
In vincenzocoia/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description Usage Arguments Value See Also Examples

negative log-likelihood of multivariate copula with discrete margins

1 2	mdiscretenllk(cpar,uudat,pmcop,LB,UB) emvndiscretenllk(cpar,zzdat) # exchangeable multivariate normal

`cpar`	copula parameter for pmcop
`uudat`	n x (2d) matrix with uudat[,1:d] lower corner and uudat[(d+1):(2d)] upper corner of rectangle, after fitting univariate models and converting margin via cdf
`pmcop`	function for the cdf of the d-variate copula
`zzdat`	n x (2d) matrix with zzdat[,1:d] lower corner and zzdat[(d+1):(2d)] upper corner of rectangle, after fitting univariate models and converting margin to standard normal (for emvndiscretenllk which is positive exchangeable multivariate normal)
`LB`	lower bound vector for cpar
`UB`	upper bound vector for cpar

negative log-likelihood

discreteresponse

# d=4 dimensional data with transformed univariate 
uu=matrix(c(.1,.1,.1,.1, .5,.6,.7,.8, 0,0,0,0, .4,.5,.4,.3),2,8,byrow=TRUE)
zz=qnorm(uu)
zz[zz< -6]=-6
mdiscretenllk(2,uu,pmfrk,0,30)
emvndiscretenllk(0.3,zz)