[R-sig-Geo] covariance matrix for kriging predictions
Edzer Pebesma
edzer.pebesma at uni-muenster.de
Thu Dec 12 22:54:02 CET 2013
On 12/12/2013 08:20 PM, Jui-Han Chang wrote:
> Hi all,
>
> I am trying to get covariance matrix for kriging predictions. I used krige0
> function in the gstat library with options fullCovariance=TRUE and
> computeVar=TRUE. I got a covariance matrix but the values seem to be too
> large. I converted the covariance matrix to correlation matrix and got
> values larger than 1. Following are the codes to derive the matrices:
>
> library(gstat)
> data(meuse)
> coordinates(meuse) = ~x+y
> data(meuse.grid)
> gridded(meuse.grid) = ~x+y
>
> m <- vgm(.59, "Sph", 874, .04)
> x <- krige0(log(zinc)~1, meuse, meuse.grid, model =
> m,fullCovariance=TRUE,computeVar=TRUE)
>
> x$var[1:5,1:5]
>
> [,1] [,2] [,3] [,4] [,5]
> 1 0.3108421 0.2794621 0.2883862 0.3019975 0.2528104
> 2 0.2794621 0.2414045 0.2546174 0.2727450 0.2074984
> 3 0.2883862 0.2546174 0.2631935 0.2769452 0.2263174
> 4 0.3019975 0.2727450 0.2769452 0.2863739 0.2499542
> 5 0.2528104 0.2074984 0.2263174 0.2499542 0.1648699
>
> cov2cor(x$var[1:5,1:5])
>
> [,1] [,2] [,3] [,4] [,5]
> 1 1.000000 1.020188 1.008247 1.012201 1.116746
> 2 1.020188 1.000000 1.010131 1.037331 1.040091
> 3 1.008247 1.010131 1.000000 1.008764 1.086450
> 4 1.012201 1.037331 1.008764 1.000000 1.150331
> 5 1.116746 1.040091 1.086450 1.150331 1.000000
>
> I am not sure what I am getting here now. Is this covariance matrix for
> kriging predictions? If not does anyone know how to get a covariance
> matrix? Any help will be appreciated.
Thanks, this is clearly wrong. My feeling is that c0 in the computation
of the kriging predictions covariance matrix needs be specified
differently, not as a constant, however I haven't sorted out how, so far.
--
Edzer Pebesma
Institute for Geoinformatics (ifgi), University of Münster
Heisenbergstraße 2, 48149 Münster, Germany. Phone: +49 251
83 33081 http://ifgi.uni-muenster.de GPG key ID 0xAC227795
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 555 bytes
Desc: OpenPGP digital signature
URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20131212/b751d607/attachment.bin>
More information about the R-sig-Geo
mailing list