[R] Bivariate Weibull distribution -- Copula

David Duffy David.Duffy at qimr.edu.au
Fri Oct 6 03:47:38 CEST 2006


"Jenny Stadt" <jennystadt at yahoo.ca> asked:

>
> I am struggling in a bivariate Weibull distribution although I
> searched R-Site-Help and found suggestion with Copula. Seems the
> maximum likelihood estimate is beyond what I can understand.
>
> My case is: given two known marginal distribution (both are Weibull),
> and the correlation between them. How can I get the bivariate
> distribution based on these two? I wish if there is anybody has the
> experience in bivariate distribution could give me the hints. Thanks!
>
> Jen.

Not a direct answer, but you might look at Ronald Pruitt's "BIVSURV S
LIBRARY", which runs under R, snd the references in the README.  Another
way to model such data is using survreg with a frailty term -- see the
survival and kinship packages.  The most recent Kalbfleisch & Prentice
(Kalbfleisch J. D. & Prentice, R. L. (2002) The Statistical Analysis of
Failure Time Data Wiley, 2nd Edition) has a chapter on bivariate survival
analysis (with our example in it;)).

David Duffy.



More information about the R-help mailing list