Skip to content
Prev 18025 / 29559 Next

Variogram Error

Saman,

Ben Graeler managed to reproduce the same effect by:

library(gstat)
data(meuse)
coordinates(meuse) <- ~x+y

obj = gstat(NULL, "D1", zinc~1, meuse[1,], set = list(zero_dist = 3))
obj = gstat(obj, "D2", zinc~1, meuse[2,])
variogram(obj, cross = "ONLY", pseudo = T, boundaries=c(0,100))

Essentially, for ST variograms, for each time slice the mean was removed
before computing the (pseudo) cross variogram. This will usually have
little effect, unless you have very few, or as in this (and your) case,
only one observation per time slice.

We changed the code now such that the mean is not removed. Updates are
committed r-forge; the next version on CRAN will include this update.
On 04/14/2013 06:10 PM, Saman Monfared wrote: