[R-sig-Geo] !SPAM: prediction error variance

[Roger Bivand]

On Thu, 28 Oct 2010, Pinar Aslantas Bostan wrote:

> Dear all,
>
> I performed gwr() and obtained predictions for 31223 data points. Now I 
> want to calculate prediction error variance at each estimation point. Is 
> there a function to calculate them or do you have any idea about 
> calculating these values?

You need to pass in a fittedGWRobject, and set se.fit, to provide the 
estimate of sigma needed to achieve this. The fitted object must have a 
hatmatrix, which is not possible with this N, so you will have to 
subsample to be able to get the sigma based on effective degrees of 
freedom. I don't think that this has been tried, and I'm not sure what it 
provides of value, given the uncertainty present about the method.

Hope this helps,

Roger

>
> Thanks!
>
> Pinar
>
> ********************************************************************************
> Pinar Aslantas Bostan
> Research Assistant
> Department of Geodetic and
> Geographic Information Technologies (GGIT)
> Middle East Technical University
> 06531 Ankara/TURKEY
> aslantas at metu.edu.tr
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo

-- 
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