BIOMOD Error message: argument of length 0

Question

I’m using the package BIOMOD in R to modelling habitat in rodent species. The data formatting works but I'm getting this error message when I run Modelling:

## Conputing the models ##
myBiomodModelOut <- BIOMOD_Modeling(biomodDataL,
                            models = c('GLM','MAXENT','GAM'),
                            models.options = myBiomodOption,
                            NbRunEval = 2,
                            DataSplit = 80,
                            Yweights = NULL,
                            VarImport = 2,
                            models.eval.meth = c('TSS','ROC'),
                            SaveObj = TRUE,
                            rescal.all.models = TRUE)

################################################################################ Creating suitable Workdir...

                    ! Weights where defined to rise a 0.5 prevalence !  

-=-=-=- Lemming Modeling Summary -=-=-=-=-=-=-=-=-=-

1 environmental variables ( X2 )

Number of evaluation repetitions : 3

Models selected : GLM MAXENT GAM Total number of model runs : 12

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

-=-=-=- Run : Lemming_PA1

-=-=-=--=-=-=- Lemming_PA1_RUN1 Model=GLM quadratic with no interaction

    Stepwise procedure using AIC criteriaError in 1:ncol(explVar) : argument of length 0

#

The explanatory variable is an index of productivity (i.e. NDVI). This raster, with explanatory variable (2'x 2' resolution), and vector with the response variable are in the same projection. Do you know why I have this kind of message?

Thanks in advance for your help, Mary

Answer 1

You are getting an error because the GLM is trying to use a stepAIC and you are only including a single variable.