[R-sig-Geo] spautolm variable selection

[Roger Bivand]

On Sun, 25 Jul 2010, Peter Larson wrote:

> Hello,
>
> Is there any way to iterate spautolm to run through all possible
> models using a set of covariates and choose the one with the lowest
> AIC?

No, there are no step or update facilities for spautolm.

Since the choice of RHS variables (and their functional forms) will affect 
the spatial parameter, this would be rather tricky anyway, as dropping 
variables might well increase autocorrelation in the residual. You could 
look at ways of using Spatial Filtering for this purpose, but the 
functions used for this do not support variable selection either - you'd 
need to write your own.

Roger

>
> Thanks,
>
> pete
>
> _______________________________________________
> 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