[Biomod-commits] BIOMOD_EnsembleModeling (Error: undefined columns selected)

[semra yalcin]

Hi Damien,

I get error message while running the BIOMOD_EnsembleModeling argument.
Briefly, ensembling of PA1 models and PA2 models worked fine, but it
stopped after selecting of proper models of PA3.

It said "models kept :.... Error in
`[.data.frame`(getModelsPrediction(modeling.output, as.data.frame =
TRUE): undefined
column selection" for PA3 models.

I've tried to rerun of models with changing the number of PA and the number
of evaluation run etc., but none of these changes worked.

What could be the reason of this?

Thanks so much in advance..

Semra



Here my codes:

###########################

library(biomod2)

myResp.ras <- raster("speciesdata/kn500.img")
myRespName <- 'Kayin'
myRespXY <- xyFromCell(object=myResp.ras, cell=which(myResp.ras[]>0))
myResp <- extract(x=myResp.ras, y=myRespXY)

datacurr = stack("Current/tr/aspect","Current/tr/bio04","Current/tr/bio05",
"Current/tr/bio06","Current/tr/bio09",
"Current/tr/bio12", "Current/tr/bio14", "Current/tr/wetness")

myBiomodDataSet1KN <- BIOMOD_FormatingData (resp.var = myResp,
expl.var = datacurr,
resp.xy = myRespXY,
resp.name = myRespName,
PA.nb.rep = 5,
PA.nb.absences = 10000,
PA.strategy = 'random')

myBiomodOption <- BIOMOD_ModelingOptions()

myBiomodModelOutSet1KN <- BIOMOD_Modeling(
myBiomodDataSet1KN,
models = c('GLM','GAM', 'GBM', 'CTA','ANN', 'FDA', 'RF','MAXENT'),
models.options = myBiomodOption,
NbRunEval=5,
DataSplit=70,
Yweights=NULL,
VarImport=3,
models.eval.meth = c('TSS','ROC'),
SaveObj = TRUE,
rescal.all.models = TRUE)

myBiomodEMSET1KN<- BIOMOD_EnsembleModeling( myBiomodModelOutSet1KN,
                         chosen.models = 'all',
                         eval.metric = c('ROC'),
                         eval.metric.quality.threshold = c(0.85),
                         prob.mean = TRUE,
                         prob.cv = TRUE,
                         prob.ci = TRUE,
                         prob.ci.alpha = 0.05,
                         prob.median = FALSE,
                         committee.averaging = TRUE,
                         prob.mean.weight = TRUE,
                         prob.mean.weight.decay = 'proportional')

###########################
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/biomod-commits/attachments/20120830/edaf5149/attachment.html>