[R-SIG-Finance] question of fitMvdc and fitCopula functions

Xiaochen Sun Xiaochen.Sun at brunel.ac.uk
Tue Aug 21 01:26:16 CEST 2007

Dear all,
I wonder if anyone has experience on copula package regarding the copula parameter estimation.
I was reading "Enjoy the Joy of Copulas" and the manual of the package "copula", 
The example code:
fit <- fitMvdc(dat, myMvd, start = start, optim.control = list(trace=TRUE,maxit=200)

the item myMvd reflects the copula functions, but it is already defined by
myMvd <- mvdc(copula=ellipCopula(family= "nomal", param=0.5),margins=c("gamma","gamma"), paramMargins=list(list(shape=2, scale=1),list(shape=3,scale=2)))
My question is suppose I have a series data X1....Xn;Y1....Yn, however without knowing the margins, and I intend to estimate the parameter of the chosen copula, how can I run this function????
Same question to fitCopula function......fitCopula(eu, myMvd at copula <mailto:myMvd at copula> , start = a.0)
If you go to "QRMlib" package, you will find similar function 
fit.gausscopula and fit.Archcopula2d, in both case, you can simply code as fit.Archcopula2d(Udata, "clayton")
[Ref: Enjoy the Joy of Coulas, Jun Yan, 2006]
Appreciate for any reply.

More information about the R-SIG-Finance mailing list