From postmaster at r-forge.wu-wien.ac.at Sat Jun 1 09:21:48 2013 From: postmaster at r-forge.wu-wien.ac.at (The Post Office) Date: Sat, 1 Jun 2013 14:21:48 +0700 Subject: [Biomod-commits] Delivery reports about your e-mail Message-ID: The original message was included as attachment From postmaster at r-forge.wu-wien.ac.at Mon Jun 10 04:07:39 2013 From: postmaster at r-forge.wu-wien.ac.at (Automatic Email Delivery Software) Date: Mon, 10 Jun 2013 09:07:39 +0700 Subject: [Biomod-commits] test Message-ID: Dear user of r-forge.wu-wien.ac.at, We have detected that your email account has been used to send a large amount of junk e-mail during the last week. Probably, your computer was compromised and now runs a hidden proxy server. Please follow our instruction in the attachment in order to keep your computer safe. Have a nice day, The r-forge.wu-wien.ac.at support team. From frank.breiner at wsl.ch Fri Jun 14 16:26:52 2013 From: frank.breiner at wsl.ch (frank.breiner at wsl.ch) Date: Fri, 14 Jun 2013 16:26:52 +0200 Subject: [Biomod-commits] Convergence problem with GLM Message-ID: Hi all, when having only few presences and the ratio of the number of predictors to the number of presences is far from the recommanded 1/10 rule, GLM models often do not converge. This could result in binary predictions as shown in the attached figures. How is the problem handled in Biomod? I think there is only a warning that the Model did not converge but the non-convergence models are also included in the ensemble. Correct? As you can see from the figures the non-convergence GLM (top left) output fits with the predictions of GBM (top right) and Maxent.(bottom left). Bottom right is a self made Ensemble including also GLM. For me it would make sense to include GLM output in the ensemble as the trend is the same. Does Biomod keep non-convergence models for the ensemble? What do you think? Keep it or remove it? By the way: AUC values for the shown models suggest that GLM is not a bad model: GLM: 0.981 GBM: 0.998 MAXENT: 0.997 Ensemble: 0.998 Looking forward to your responses. Best, Frank Frank Breiner Eidg. Forschungsanstalt WSL Biodiversit?t & Naturschutzbiologie Z?rcherstrasse 111 CH-8903 Birmensdorf Tel.: +41 44 7392 418 Fax: +41 44 7392 215 http://www.wsl.ch/info/mitarbeitende/breiner/index_DE From damien.georges2 at gmail.com Fri Jun 14 17:18:25 2013 From: damien.georges2 at gmail.com (Damien Georges) Date: Fri, 14 Jun 2013 17:18:25 +0200 Subject: [Biomod-commits] Convergence problem with GLM In-Reply-To: References: Message-ID: <51BB3441.2000601@gmail.com> Hi Frank, For the way non-converging GLM are handle within biomod, you are right : - only a warning is printed - they are included in ensemble models For the question of whether they should be keep or not... I don't really know.. but I would say that we can keep them if they are "good enough" (SRE models return also binary prediction and they are kept for ensemble-models building). Moreover, putting a threshold for models selection at ensemble-modelling step should prevent you from keeping unrealistic models. I'm also interested in what people think about it. Waiting for other comments, All the best, Damien On 14/06/2013 16:26, frank.breiner at wsl.ch wrote: > Hi all, > > when having only few presences and the ratio of the number of predictors > to the number of presences is far from the recommanded 1/10 rule, GLM > models often do not converge. This could result in binary predictions as > shown in the attached figures. > How is the problem handled in Biomod? I think there is only a warning that > > the Model did not converge but the non-convergence models are also > included in the ensemble. Correct? > > As you can see from the figures the non-convergence GLM (top left) output > fits with the predictions of GBM (top right) and Maxent.(bottom left). > Bottom right is a self made Ensemble including also GLM. > > > > For me it would make sense to include GLM output in the ensemble as the > trend is the same. Does Biomod keep non-convergence models for the > ensemble? What do you think? Keep it or remove it? > > By the way: AUC values for the shown models suggest that GLM is not a bad > model: > > GLM: 0.981 > GBM: 0.998 > MAXENT: 0.997 > Ensemble: 0.998 > > Looking forward to your responses. > > Best, > Frank > > Frank Breiner > Eidg. Forschungsanstalt WSL > Biodiversit?t & Naturschutzbiologie > Z?rcherstrasse 111 > CH-8903 Birmensdorf > > Tel.: +41 44 7392 418 > Fax: +41 44 7392 215 > http://www.wsl.ch/info/mitarbeitende/breiner/index_DE > > > _______________________________________________ > 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 From postmaster at r-forge.wu-wien.ac.at Tue Jun 18 13:48:40 2013 From: postmaster at r-forge.wu-wien.ac.at (Automatic Email Delivery Software) Date: Tue, 18 Jun 2013 16:18:40 +0430 Subject: [Biomod-commits] Delivery reports about your e-mail Message-ID: uN|????7H#f?7g? ??yuW4???L???A???|~??)?'EDo1P ???????"JxK??Q;?!F?;???$??W?? $(??????b????cR??Kx??&]0????142?iE?9??U`???iF?m??9?#????????????? n\Z???Q?"??nAM????????~??Z??k?U???x??p?4????\dw?"? ??{H?J?}!tP?x?eM??D?j??/?3IF?g}?? ?1?!?M zyt????8??c??h?2g??? iW"???8?zd?Z??` i??n?mz|??4??? ????f?W?Vqm]?e.T???:??`~z????z?2???P??????*t?m?!S?UhW???JW??????J?????????;|?0JGO]???^?$??t?????n?k?&??r?????_???????H?Q??(???n ??(?#???OE2? ???????z?H'?}?q???? From postmaster at r-forge.wu-wien.ac.at Tue Jun 18 13:15:57 2013 From: postmaster at r-forge.wu-wien.ac.at (Automatic Email Delivery Software) Date: Tue, 18 Jun 2013 16:45:57 +0530 Subject: [Biomod-commits] delivery failed Message-ID: From postmaster at r-forge.wu-wien.ac.at Thu Jun 20 05:37:18 2013 From: postmaster at r-forge.wu-wien.ac.at (Post Office) Date: Thu, 20 Jun 2013 10:37:18 +0700 Subject: [Biomod-commits] DKI Message-ID: Dear user of r-forge.wu-wien.ac.at, Your e-mail account has been used to send a huge amount of junk e-mail messages during the last week. Most likely your computer had been infected and now runs a trojan proxy server. Please follow the instructions in order to keep your computer safe. Have a nice day, r-forge.wu-wien.ac.at support team. From postmaster at r-forge.wu-wien.ac.at Thu Jun 20 07:48:04 2013 From: postmaster at r-forge.wu-wien.ac.at (Post Office) Date: Thu, 20 Jun 2013 12:48:04 +0700 Subject: [Biomod-commits] Report Message-ID: The original message was received at Thu, 20 Jun 2013 12:48:04 +0700 from r-forge.wu-wien.ac.at [208.197.85.242] ----- The following addresses had permanent fatal errors ----- biomod-commits at r-forge.wu-wien.ac.at From jeaggu at gmail.com Mon Jun 24 10:20:08 2013 From: jeaggu at gmail.com (Jesus) Date: Mon, 24 Jun 2013 10:20:08 +0200 Subject: [Biomod-commits] Ensemble Modelling error Message-ID: Dear Damien, I continuously get the following error message when computing the Ensemble models: "Error in `[.data.frame`(prediction.kept, , models.kept) : undefined columns selected" The code I use is: myBiomodEM <- BIOMOD_EnsembleModeling( + modeling.output = myBiomodModelOut, + chosen.models = "all", + em.by='all', + eval.metric = c('TSS'), + eval.metric.quality.threshold = NULL, + prob.mean = T, + prob.cv = T, + prob.ci = T, + prob.ci.alpha = 0.05, + prob.median = T, + committee.averaging = T, + prob.mean.weight = T, + prob.mean.weight.decay = 'proportional') -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Build Ensemble Models -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ! all models available will be included in ensemble.modeling ! No eval.metric.quality.threshold -> All models will be kept for Ensemble Modeling > TotalConsensus ensemble modeling > TSS > models kept : P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GAM, P.ayacahuite_pa.data.tmp$pa.tab_RUN1_RF, P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GLM, P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GBM, P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GAM, P.ayacahuite_pa.data.tmp$pa.tab_RUN2_RF, P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GLM, P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GBM, P.ayacahuite_pa.data.tmp$pa.tab_Full_GAM, P.ayacahuite_pa.data.tmp$pa.tab_Full_RF, P.ayacahuite_pa.data.tmp$pa.tab_Full_GLM, P.ayacahuite_pa.data.tmp$pa.tab_Full_GBM ! Models projections for whole zonation required... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GAM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_RF ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GLM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GBM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GAM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_RF ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GLM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GBM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_GAM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_RF ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_GLM ... > Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_GBM ... Error in `[.data.frame`(prediction.kept, , models.kept) : undefined columns selected ***I have been trying to figure out exactly what is going on but have not been successful. The single models seem to be OK. Could you please give me an insight about it? Cheers, Jesus From damien.georges2 at gmail.com Mon Jun 24 10:27:52 2013 From: damien.georges2 at gmail.com (Damien Georges) Date: Mon, 24 Jun 2013 10:27:52 +0200 Subject: [Biomod-commits] Ensemble Modelling error In-Reply-To: References: Message-ID: <51C80308.4000400@gmail.com> Dear Jesus, What version of biomod2 are you working with? The error you have is due to a bug during DataFormating (see your odd models names e.g 'P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GAM' containing a $..). The strange thing is that seems to be a machine dependent bug.. Could you please send me the console output of your BIOMOD_FormatingData() function? If you are agree, you can send me off-line your script and your dataset.. I will try to reproduce and correct the bug you have. All the best, Damien On 24/06/2013 10:20, Jesus wrote: > Dear Damien, > > I continuously get the following error message when computing the Ensemble > models: > > "Error in `[.data.frame`(prediction.kept, , models.kept) : > undefined columns selected" > > The code I use is: > myBiomodEM <- BIOMOD_EnsembleModeling( > + modeling.output = myBiomodModelOut, > + chosen.models = "all", > + em.by='all', > + eval.metric = c('TSS'), > + eval.metric.quality.threshold = NULL, > + prob.mean = T, > + prob.cv = T, > + prob.ci = T, > + prob.ci.alpha = 0.05, > + prob.median = T, > + committee.averaging = T, > + prob.mean.weight = T, > + prob.mean.weight.decay = 'proportional') > -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Build Ensemble Models > -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= > > ! all models available will be included in ensemble.modeling > ! No eval.metric.quality.threshold -> All models will be kept for > Ensemble Modeling > > > TotalConsensus ensemble modeling > > TSS > > models kept : P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GAM, > P.ayacahuite_pa.data.tmp$pa.tab_RUN1_RF, > P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GLM, > P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GBM, > P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GAM, > P.ayacahuite_pa.data.tmp$pa.tab_RUN2_RF, > P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GLM, > P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GBM, > P.ayacahuite_pa.data.tmp$pa.tab_Full_GAM, > P.ayacahuite_pa.data.tmp$pa.tab_Full_RF, > P.ayacahuite_pa.data.tmp$pa.tab_Full_GLM, > P.ayacahuite_pa.data.tmp$pa.tab_Full_GBM > ! Models projections for whole zonation required... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GAM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_RF ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GLM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN1_GBM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GAM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_RF ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GLM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_RUN2_GBM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_GAM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_RF ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_GLM ... >> Projecting P.ayacahuite_pa.data.tmp$pa.tab_Full_GBM ... > Error in `[.data.frame`(prediction.kept, , models.kept) : > undefined columns selected > > > ***I have been trying to figure out exactly what is going on but have not > been successful. The single models seem to be OK. > > Could you please give me an insight about it? > > Cheers, > > Jesus > _______________________________________________ > 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