On Sat, 30 Sep 2006, Sara Mouro wrote:
> Dear all,
>
> Could you pleas advise me on the following?
>
> I need to use general(ized) linear models (binomial distribution + logit
> link function) , to describe the preferred environment of each species
(each
> sample is an individual in which I have measured several variables and also
> recorded the species it belongs to)
>
>
>
> However, must account for the spatial autrefoocorrelation between
> individuals.
>
>
>
> -> So I think I need something like the so called
"autologistic" models.
> isn't it?
>
> -> What package would you advise me to use for that?
>
RSiteSearch("autologistic")
tells you what is known about this, which I'm afraid seems to be that no
such function is available. Your question is very similar to the first hit
in the site search, and that received no answer.
If you are willing to try a different framework, the ME() Moran
eigenvector function in the spdep package is a possibility, albeit not yet
well-proven. It adds selected eigenvectors from a centred spatial weights
matrix to the RHS of the glm to "whiten out" spatial dependence, but
it
will also "whiten out" or paint over other spatially patterned
mis-specification problems.
>
>
> Thank you in advance.
>
>
>
> Best regards,
>
> Sara Mouro
>
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> 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
> and provide commented, minimal, self-contained, reproducible code.
>
--
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no