I haven't seen any replies to your post, and I know nothing
about "copula". In case you are still interested in help with this, I
will offer a few suggestions / questions. First, have you reviewed the
posting guide! "www.R-project.org/posting-guide.html"? I believe that
the process described there has helped people answer their own
questions, and failing that, has increased on average the speed and
quality of replies received.
Also, have you tried a direct email to the maintainer [whose
email address can be obtained via 'help(package="copula")']?
Beyond that, have you made some attempt to use 'copula' to do
what you want? For example, have you tried working the "examples" in
the help file for 'mvdc'? If you have and you still have questions,
please submit another post including a self-contained example, e.g.,
modifying the "exampe" in the help file for 'mvdc'.
hope this helps.
spencer graves
Horace Tso wrote:
> Dear r list,
>
> I posted this on the S list last week since i'm using some of the
> FinMetrics functions on copula. Knowing there is a copula package in R,
> I figure this would be an appropriate forum to ask this question.
>
> I want to model inverse relationship between two (non-normal,
> non-symmetric) marginals with the gumbel copula, or with any copula.
> Say, x is lognormal and y is norm. Since gumbel's delta must be greater
> than one, how do I specify the equivalence of a negative correlation?
>
> If both are symmetric, I think I could get away by using a positive
> delta, simulate the bivariate realizations and then flipping the sign on
> one of them. Or am I completely off.
>
> I did search through the archive but found no related posting. Thanks
> in advance.
>
> Horace W. Tso
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html