Interpolating Temperature Data with Easykrig

Posted in: , on 12. May. 2009 - 19:28

Validation one leave out in easykrig

I'm working with easykrig in matlab to interpolate temperature data. Now I want validate these data and in easykrig I can use the validation one leave out but there is a problem: when the program gives to me results, predicted data, the first predicted datum (it's always the first) is coincident with the observed!Why?

A Valid Observation, Or What?

Posted on 17. May. 2009 - 02:40

Why what?

Perhaps you've got your URL's a bit mixed up. Have you tried the Matlab user group?

Why Matlab?

Posted on 17. May. 2009 - 05:52

A so-called "predicted value" is a functionally dependent value. Each predicted value has its own variance. Predicted values do not give degrees of freedom. Assume spatial dependence, krige, smooth, rig the rules of statistics, and be happy with Matlab!

JanWMerks