[R-sig-Geo] autoKrige: singular matrix in function LDLfactor

[ElenaC]

Dear List,

I have a dataset (with an irregular grid) which I am trying to interpolate
to a regular grid with much higher spatial resolution using the autoKrige
function from the package 'automap'. But I get the error:

> SST_data <- autoKrige(SST~1, HG1990, center_points2, nmax=50)
[using ordinary kriging]

"chfactor.c", line 131: singular matrix in function LDLfactor()
Error in predict.gstat(g, newdata = newdata, block = block, nsim = nsim,  : 
  LDLfactor


I understand this error could be due to duplicated points either within the
dataset itself or between the dataset and the points I want to interpolate
to. But I fail to find any duplicates with the function zerodist (with
default minimum distance):

> zerodist(HG1990)
     [,1] [,2]
> zerodist2(HG1990,center_points2)
     [,1] [,2]


(In the example above both 'HG1990' and 'center_points2' are
SpatialPointsDataFrame objects).

Any other ideas of what could be causing this error or how to deal with it
would be greatly appreciated! (I will be happy to send the data to anybody
who wants to take a closer look).

Thanks in advance,
Elena




--
View this message in context: http://r-sig-geo.2731867.n2.nabble.com/autoKrige-singular-matrix-in-function-LDLfactor-tp7584343.html
Sent from the R-sig-geo mailing list archive at Nabble.com.