[R] Zero truncated Poisson distribution & R2WinBUGS

[Achim Zeileis]

On Thu, 3 Mar 2011, Timothée wrote:

> Hi,
> I have a very similar problem...
> In some sites, counts data>0
> In the other sites, counts = 0
> I already applied zero inflated models with the zero-trick (Martin et al
> 2005 in Ecology letters), but I would like to use truncated distributions
> (Poisson and negative binomial) to model my counts in order to estimate the
> number of sites where no counts but presence.
> The problem is that I can't manage to write the likelihood of such truncated
> distributions.

If you're looking for a maximum likelihood regression model (without 
random effects), then the "countreg" package on R-Forge may be helpful for 
you.

The package is not yet on CRAN because we are planning to add more 
functionality, but the code is well tested. It's the same code-base 
underlying zeroinfl() and hurdle() in "pscl".

hth,
Z

> Did you get any valuable answers to your post ?
> Would you mind to share with me how you succeeded (if so) ?
> Thank you very much
> Timothée
>
> --
> View this message in context: http://r.789695.n4.nabble.com/Zero-truncated-Poisson-distribution-R2WinBUGS-tp3043121p3333406.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.