<HTML><HEAD>
<META content="text/html; charset=utf-8" http-equiv=Content-Type>
<META name=GENERATOR content="MSHTML 8.00.6001.19088"></HEAD>
<BODY style="MARGIN: 4px 4px 1px; FONT: 10pt Tahoma">
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>Hi,<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>I am currently working with R 2.13.0 and Biomod 1.1-6.9.<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>I discovered that predictors are automatically dropped from GAM/GLM models, when their importance (via VarImportance) is zero. Is BIOMOD intended to work like that? Are there any possibilities to suppress the dropping? I am asking, because the evaluation of variable importance can be extremely different, pending on the models which are used. Example:<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>> VarImportance<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>$SP1<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri><SPAN style="mso-spacerun: yes"> </SPAN>Var1 Var2 Var3 Var4<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>GAM 0.000 0.973 0.000 0.000<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>GBM 0.192 0.344 0.42 0.015<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>The GAM model was calibrated with one predictor only (Var2)! All other variables were automatically dropped from the GAM, including Var3, which is suggested to be the most important one by GBM. When I used Models() and a priori excluded Var2, then Var3 became most important for GAMs. Actually this indicates collinearity, which is 0.77 for Var2 and Var3 (pearson correlation coefficient). This doesn’t occur with GBMs, because they are more robust against multi-collinearity? Any suggestions? Many thanks, cheers Jonathan <o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>PS: The settings in Models():<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri>Models(GAM = T, GBM = T, NbRepPA=1, strategy="random", nb.absences=1000,<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri><SPAN style="mso-spacerun: yes"> </SPAN>NbRunEval = 5, DataSplit = 70, Yweights=NULL, Roc=TRUE, Optimized.Threshold.Roc=T,<o:p></o:p></FONT></SPAN></P>
<P style="MARGIN: 0cm 0cm 10pt" class=MsoNormal><SPAN style="mso-ansi-language: EN-GB" lang=EN-GB><FONT face=Calibri><SPAN style="mso-spacerun: yes"> </SPAN>Kappa=F, TSS=F, KeepPredIndependent = FALSE, VarImport=5)<o:p></o:p></FONT></SPAN></P></BODY></HTML>