[R-sig-Geo] Regression kriging

[Edzer Pebesma]

Jose, is your model linear, or are you using a generalized linear model?

The questions is not so much: model parameters before or after kriging
residuals, but rather: model parameters under the assumption of
independent observations (the usual regression approach), or model
parameters under the assumption of spatially dependent observations (the
geostatistical, or generalized linear regression approach). In the case
you are using a GLM, the report mentioned earlier on regression kriging
(another name for universal kriging) may not be very helpful, as it
deals with linear models.
--
Edzer

Jose Funes wrote:
> Hi,
>
> I have used regression kriging to model abundance of an invasive
> species. After performing a stepwise regression of the model, I fitted
> a theoretical variogram to a empirical variogram of the residuals. My
> question is how to obtain the parameter estimate of the model after
> kriging the residuals. Do the parameters of the regression model
> differ after kriging the residuals? if so, how can I get them from the
> R output?
>
> Looking forward to hearing from you, thanks
>
> Jose Funes
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>