[Biomod-commits] Error message using Models()

[Jonathan Heubes]

Dear colleagues,

when I try to run Models() (R version 2.10.1 with all necessary up-to-date packages and BIOMOD 1.1.3) I get an error message for some species (not for all!) which I cannot interpret.
Data.frame-structure: 51138 objects, 3 climate predictors, different species

My code:
Initial.State(Response=dat.model[,1], Explanatory=dat.model[,2:4],IndependentResponse=NULL, IndependentExplanatory=NULL, sp.name="sp1")
Models(GAM = TRUE, DataSplit = 70, Yweights=NULL, Roc=T, Optimized.Threshold.Roc=T, Kappa=T, TSS=T, KeepPredIndependent = FALSE)

Error message:

#####                    sp1                  #####
Model=GAM spline
         3  Degrees of smoothing
Error in if (a != "NA") if (a > 0) { :
  missing value where TRUE/FALSE needed
In addition: Warning message:
In (n1. * n.1) + (n2. * n.2) : NAs produced by integer overflow

I get this error messages, irrespectively of the algorithm (exception: SRE). The problem seems to arise from the inner function 'KappaRepet'. I was successful to run Models() with the same setting while using BIOMOD 1.1.0 and an older version of R as well as with gam() in the gam-package.
Any ideas, suggestions?

Thanks, cheers Jonathan