[R-sig-Geo] Variable search radius for interpolation

[Tomislav Hengl]

I've implemented recently a function that helps speeding up kriging with 
spatially clustered data (for it fits kernel density function first and 
then sub-samples regular points proportionally to the density of 
observations). It comes together with the GSIF package:

http://gsif.r-forge.r-project.org/spline.krige.html

This should be especially useful if you need to interpolate values over 
a large grid e.g. millions of prediction locations. For example, I've 
used it to interpolate values over whole Africa at 250 m (over 30,000 by 
30,000 pixels) and it took only 3-4 minutes to run.

Note that this function needs SAGA GIS (v2.1 or higher).

HTH,

T. Hengl

On 24-11-2014 18:34, Moshood Agba Bakare wrote:
> Hi All,
> Could anyone guide me on how to define variable search radius in GSTAT
> while interpolating into a grid cell? The GSTAT documentation used MAXDIST
> to define fixed search radius. (2) Is there a function in GSTAT to
> determine the point density while interpolating into different grids size?
>
> Thank you
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>