|
No, but I can see there is a need for this. Writing this is not a straightforward extension to krige.cv; I'd rather extend predict.gstat -- you'd like to extend it to cokriging with arbitratrily many variables. One catch is that it is not really straightforward whether secondary variables should be omitted at prediction locations or not; also, whether they should be cross validated as well. I you don't see it appear (or announce), keep reminding me this is missing; right now I have too little time to accomplish this. -- Edzer Jorge wrote:
|