[Biomod-commits] GAM error return

[Wilfried Thuiller]

Dear Hakim,
I have no clue. It seems the GAM has failed to run for whatever reason  
independent of BIOMOD.
It happens sometimes when you do not have enough data to calibrate the  
full model or when the GAM does not manage to converge.
Try to run it off BIOMOD but I am pretty sure the results will be the  
same. The easiest solution is to turn GAM as wrong for that particular  
species.

Cheers,
Wilfried


Le 23 oct. 2009 à 21:42, Hakim Abdi a écrit :

> Dear All:
>
> I'm getting the following error when loading the GAM model:
>
> > load("models/SPECIES1_GAM_full")
> > SPECIES1_GAM_full
>
> Family: binomial
> Link function: logit
>
> Formula:
> SPECIES1 ~ s(BAND5, 3) + s(BAND7, 3)
>
> Estimated degrees of freedom:
> Error in x$smooth[[i]]$first.para :
>   $ operator is invalid for atomic vectors
>
> > summary(SPECIES1_GAM_full)
> NULL
> Formula:
> NULL
>
> Error in if (x$m > 0) { : argument is of length zero
>
> While the GLM model returns the following:
>
> > load("models/SPECIES1_GLM_full")
> > SPECIES1_GLM_full
>
> Call:  glm(formula = SPECIES1 ~ I(BAND7^3) + BAND4 + I(BAND5^2),  
> family = binomial,      data = DataBIOMOD[calib.lines, ], weights =  
> Yweights[calib.lines,          i])
>
> Coefficients:
> (Intercept)   I(BAND7^3)        BAND4   I(BAND5^2)
>  -6.566e-01   -3.090e-06   -1.373e-02    1.527e-04
>
> Degrees of Freedom: 342 Total (i.e. Null);  339 Residual
> Null Deviance:      467.3
> Residual Deviance: 428.4        AIC: 436.4
>
> My model setup parameters are as follows:
>
> > Models(GLM = T, TypeGLM = "quad", Test = "AIC", GBM = F, No.trees  
> = 3000, GAM = T, CTA = F, CV.tree = 50, ANN = F, CV.ann = 2, SRE =  
> F, Perc025=T, Perc05=F, MDA = F, MARS = F, RF = F, NbRunEval = 2,  
> DataSplit = 80, Yweights=NULL, Roc=T, Optimized.Threshold.Roc=T,  
> Kappa=T, TSS=T, KeepPredIndependent = F, VarImport=5, NbRepPA=0,  
> strategy="circles", coor=CoorXY, distance=2, nb.absences=1000)
>
> Am I missing something?
>
> Thanks in advance for the assistance.
>
>
> -- 
> Mit freundlichen Grüßen/Regards
> __________________________________
> Hakim Abdi
> Erasmus Mundus Master's Programme
>
> Institut für Geoinformatik
> Universität Münster
> 32U 5755408.16mN, 404463.35mE
> ifgi.uni-muenster.de
>
> Instituto Superior de Estatística e Gestão de Informação
> Universidade Nova de Lisboa
> 29S 4287095.5mN, 486099.7mE
> www.isegi.unl.pt
>
> www.geospatialtechnologist.com
> _______________________________________________
> Biomod-commits mailing list
> Biomod-commits at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits

Dr. Wilfried Thuiller
Laboratoire d'Ecologie Alpine, UMR CNRS 5553
Université Joseph Fourier
BP53, 38041 Grenoble cedex 9, France
tel: +33 (0)4 76 63 54 53
fax: +33 (0)4 76 51 42 79

Email: wilfried.thuiller at ujf-grenoble.fr
Home page: http://www.will.chez-alice.fr
Website: http://www-leca.ujf-grenoble.fr/equipes/tde.htm

FP6 European MACIS project: http://www.macis-project.net
FP6 European EcoChange project: http://www.ecochange-project.eu



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.r-forge.r-project.org/pipermail/biomod-commits/attachments/20091025/7bac9dfc/attachment.htm