[SOLVED] What causes this singular matrix error when kriging

[markm0705]

I am kriging some simulation samples into blocks and have run into a problem with a singular matrix.
In the code below I can krige the first sample(dt.tmp1.csv) there is no problem but when I attempt to krige the second (from a different distribution) kriging reports a singular matrix error.("Error in matrix inversion (rank=1324) : Pivot #2 is null Singular matrix")
I'm confused as to why this should happen to given the structure of the two datasets is the same.
My R script and data files are in the

Kriging-Question.zip

attachment

(46.85 KiB) Downloaded 194 times

[Fabien Ors]

Dear
Singular matrix error are usually due to the presence of duplicated samples (two or more samples are located to the same position or are too close).
The second dataset has 1698 samples (instead of 1681 for the first dataset) and 17 samples are duplicated.
Please use the duplicate function before launching kriging.
Hope this helps.

Best regards

[markm0705]

Fabien thank you for your prompt reply Regards