[Biomod-commits] sample size problem?

[Wilfried Thuiller]

Dear Andreia,
Yes, it seems the neural networks do not work with very few points. It 
is not a very good practice of fitting models with so little information.
These models are correlative. They cannot be properly adjusted with very 
few points.

First, remove the neural networks from your runs. They are pretty 
sensitive to sample size. Then run the other ones. With a bit of chance, 
the other would fit.

Hope it helps,
Wilfried



Andreia Mendonça a écrit :
> Hello all,
>
> I am using BIOMOD to model some data, but I'm facing some errors that 
> i can't seem to overcome.
>
> As i tried to run the models on some of the populations, the following 
> error showed up in R:
> *Error in nnet.default (x, y, w,...): **initial value in 'vmmin' is not finite 
>
>
>
> *The error only happens when I try to model populations with 12, 9 or 7 individuals. When i model populations with 40, 26, 19 or even 14 individuals it works just fine.
>
>
> So this error has to do with the sample size? Is there any way to solve this?
>
> I really appreciate all the help you can give me, since i am much of a newbie with modelling methods. 
>
>
> *Andreia*
> ------------------------------------------------------------------------
>
> _______________________________________________
> 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
>   

-- 
--
Wilfried Thuiller
Laboratoire d'Ecologie Alpine, UMR-CNRS 5553
Université J. Fourier
BP 53, 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/20090319/73e0327d/attachment.htm