[R-sig-Geo] applying a regression kriging model outside gstats

[Tomislav Hengl]

Ordinary kriging and hence also RK are matrix algebra 
(http://spatial-analyst.net/book/system/files/RK5points.zip <-- look for 
the excel file). If you have model parameters (betas, vgm parameters), 
then it should be relatively easy to implement to generate predictions 
(you do not need gstat or similar). Problems might happen if the number 
of points is very high >>1000 so solving matrices can take time (in 
which case using a search radius etc is a good idea) or if you have to 
generate predictions for large areas... or you do not work with a linear 
model...

HTH,

T. Hengl

On 03/20/2018 12:12 PM, Naura, Marc wrote:
> Hi,
> I have produced a regression kriging model using gstats that I want to deliver to a wide range of users to predict values for new sites using the web or as part of a simple computer software (e.g. Visual Basic software).  Is there a simple way of doing that or do I need to use R and gstats to produce predictions?
> Many thanks
> Marc
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>