[Biomod-commits] Yweights and Projections

[Damien Georges]

Hi Martin,

Unfortunately, biomod is not able to deal with "user's weights" and  
pseudo-absences sampling yet. In modeling  with pseudo absences 
sampling, some weights are automatically created to keep a 0.5 prevalence.
But, as consequence of your remark, I plane to add this functionality 
soon in the new version of biomod (i.e. biomod2).

Concerning the second point. Are you sure that it's not just an 
automatic raster color ramp issue?
Extrapolating model prediction on a big area must create some kinds of 
outliers that may biased visualizations (larger values scale).

Best,

Damien


On 17/07/2012 17:33, Martin Videla wrote:
>
> Hi Biomodellers!
>
> I have a few questions:
>
> 1) I'm trying to use the Yweight function to weight presence points 
> according their georreferencing accuracy. I have these data for 
> presence points but obviously not for pseudoabsences so I guess this 
> is causing a problem when I run the models. This is the error message
>
> Error in `[.data.frame`(tr, , 4) : undefined columns selected
>
> So my question is how can I incorporate the georreferencing accuracy 
> when using pseudoabsences?
>
> 2) I run the models successfully and then I projected them using the 
> Projection.raster function to a large region (Latin America). The 
> resulting maps predict far more presences than expected. When I 
> perform the projection of the exact same model to a smaller area 
> (Central America) within the range of the species the maps look very 
> different (better)! Does anyone know what could be the reason for such 
> variation?
>
> Thank you!
>
> Martin
>
>
>
> -- 
> Martín Videla
> Centro de Investigaciones Entomológicas de Córdoba (CIEC)
> Faculdad de Ciencias Exactas,Fisicas y Naturales (FCEFyN)
> Universidad Nacional de Córdoba (UNC)
> Velez Sarfield 299, CP: 5000, Córdoba, Argentina
> TE: 0054351-4332090 int 43
> e-mail: videla.martin at gmail.com <mailto:videla.martin at gmail.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


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/biomod-commits/attachments/20120719/61a5339e/attachment.html>