the problem is that gstat/R now used name spaces, a feature of R 1.7.0, but I forgot to modify the version requirement (yes, a bug): therefore your first problem is solved when upgrading R to >= 1.7.0
I will have a further look into your second problem.
Best regards, -- Edzer
Nick Hamm wrote:
Hi
I've come accross a curious problem when using the latest gstat package for R (gstat_0.9-8.tar.gz), which I downloaded from www.gstat.org
I have installed this on two systems.
The first is my desktop system which is a Athlon based PC running Redhat
8.0 (2.4.18-14). I'm running R version 1.6.1 (2002-11-01). I'm
experiencing problems running basic gstat functions eg:
ve <- variogram(rb1~1, loc=~i+j, data=asphaltMmrPcC)Error in .Call("gstat_init", as.integer(debug.level), PACKAGE = "gstat") :
.Call function name not in load table
vm <- fit.variogram(ve, model=vgm(0.05, "Sph", 4, 0.05))
Error in .Call("gstat_get_n_variogram_models", 0, PACKAGE = "gstat") : .Call function name not in load table
The second is one of our shared compute facilities. This is an is an
eight processor IBM x440. Its running RedHat 7.3 (2.4.21-). Its
running R version R 1.7.1 (2003-06-16). The same gstat functions seem
to work fine!
Does anybody have any thoughts about this? I think some changes to the way .Call works were made in R version 1.7.0? Perhaps this is part of the problem? Can anybody suggest a solution? I may try upgrading my desktop PC to run the latest version of R. However, this doesn't strike me as being an ideal solution. I've had no problems running earlier versions of the gstat (gstat_0.9-7 and gstat_0.9-6) package on my desktop (under R version 1.6.1).
cheers
Nick