[R-sig-Geo] [R-sig-geo] quantify prediction variance by point with ordinary kriging

Edzer Pebesma edzer.pebesma at uni-muenster.de
Mon Jan 9 23:46:46 CET 2012


Jessica, there are many things in your email that I don't understand
("direction of variance", "translation needed to quantify variance by
point"), but it might be that cross validation (krige.cv) is what you're
looking for.

Hth,

On 01/09/2012 11:02 PM, Jessica Ruth Hullman wrote:
> 
> Hi,
> 
> I'm still fairly new to kriging, and have been using ordinary kriging with
> the gstat package on radon data from location in the north east US. I'm
> basically using the method here to do this - 
> http://casoilresource.lawr.ucdavis.edu/drupal/node/442
> (create coordinates, then expand to grid, convert to gstat object, fit
> variogram, update gstat object, then use predict() command, visualizing
> with spplot()  
> 
> I'm creating kriging plots from a set of data sets, and then having people
> estimate the radon value for each plot at a given location. It appears that
> the kriging prediction variance sometimes causes the value shown on a plot
> to fall in a different range (according to the legend) than the actual
> value for that location in the input data set. I need to be able to
> quantify exactly what the variance is (including direction) at my target
> location by comparing the value from the output of the prediction (which I
> understand is a SpatialPixelsDataFrame) to my value from the input data
> set. 
> 
> I've been reading gstat documentation, but haven't been able to determine
> the easiest way to do the translation I need to quantify variance by point.
> Could anyone point me in the right direction? 
> 
> Thank you very much,
> 
> Jessica Hullman
> Doctoral Candidate, Information Visualization
> University of Michigan School of Information
> 
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo

-- 
Edzer Pebesma
Institute for Geoinformatics (ifgi), University of Münster
Weseler Straße 253, 48151 Münster, Germany. Phone: +49 251
8333081, Fax: +49 251 8339763  http://ifgi.uni-muenster.de
http://www.52north.org/geostatistics      e.pebesma at wwu.de



More information about the R-sig-Geo mailing list