From pep.bioalerts at gmail.com Thu May 9 12:27:35 2013
From: pep.bioalerts at gmail.com (Josep M Serra diaz)
Date: Thu, 9 May 2013 12:27:35 +0200
Subject: [Biomod-commits] you have to load the projections yourself?
Message-ID:
Dear List!
am sure it is a silly error but I am getting stack...
I did my modeling and my projection to present (hurra!)
but when I want to apply my ensemble modeling rules to the projections...it
says:
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Do
Ensemble Models Projections
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> Projecting Pinuspinaster_GLM_AllRun_EMbyTSS ...
> em.mean
!! You have to load your projections by yourself !!Error en
raster:::mean(raster:::subset(getProjection(projection.output), :
error in evaluating the argument 'x' in selecting a method for function
'mean': Error en subset.default(getProjection(projection.output),
getEMkeptModels(EM.output, :
'subset' must be logical
#################################################################
.... SO THEN I TRIED DOING FIRST
myPresenetProj <- getProjection(myBiomodProj.Present)
but...error:
!! You have to load your projections by yourself !!
##################################################################
HERE YOU HAVE THE CODE I USED
##### PROJECTION OUTPUT
myBiomodProj.Present <- BIOMOD_Projection ( modeling.output =
myBiomodModelOut,
new.env = expl.var ,
proj.name = paste(gcm,'_strategy_',strategy,sep=''),
selected.models = 'all',
binary.meth = 'TSS',
compress = 'xz',
clamping.mask = T,
keep.in.memory=F ,
do.stack=F
) #end BIOMOD_PROJECTION
# NOW THE ENSEMBLE FORECASTING FOR PROJECTIONS IN CRU FOR EACH ALGORITHM
myPresentProj <- getProjection(myBiomodProj.Present)
myBiomodProj.EF.algo.Present <- BIOMOD_EnsembleForecasting (
projection.output= myBiomodProj.Present , # name of the projection
EM.output =
myBiomodEM.algo , # modeling rules
binary.method= "TSS"
) #end
BIOMOD_PROJECTION
######################################################################
HERE YOU HAVE SESSION INFO
> sessionInfo()
R version 2.15.3 (2013-03-01)
Platform: x86_64-w64-mingw32/x64 (64-bit)
locale:
[1] LC_COLLATE=Spanish_Spain.1252 LC_CTYPE=Spanish_Spain.1252
LC_MONETARY=Spanish_Spain.1252 LC_NUMERIC=C
LC_TIME=Spanish_Spain.1252
attached base packages:
[1] splines grid stats graphics grDevices utils datasets
methods base
other attached packages:
[1] mgcv_1.7-22 foreign_0.8-52 biomod2_2.1.20 pROC_1.5.4
plyr_1.8 rpart_4.1-0 randomForest_4.6-7
mda_0.4-2
[9] class_7.3-5 gbm_2.0-8 survival_2.37-2 nnet_7.3-5
rasterVis_0.20-05 hexbin_1.26.1 latticeExtra_0.6-24
RColorBrewer_1.0-5
[17] lattice_0.20-13 abind_1.4-0 raster_2.1-16 sp_1.0-8
loaded via a namespace (and not attached):
[1] Matrix_1.0-11 nlme_3.1-108 tools_2.15.3 zoo_1.7-9
#################################################################
ANY HELP IS GREATLY APPRECIATED THANKS A LOT
Pep
From ajcanepa at gmail.com Thu May 9 13:14:06 2013
From: ajcanepa at gmail.com (Antonio Canepa)
Date: Thu, 9 May 2013 13:14:06 +0200
Subject: [Biomod-commits] Model projections failed
Message-ID:
Dear all,
I'm stack with a projection error.
I have run the models for my data (data.frame format) and as a test I have
projected the results using (BIOMOD_Projection, from myBiomodModelOut) to
the same environmental variables and xy coordenate. It actaully works but
is silly, and I did it before to select the new environmental variables and
XY corrdenates.
Now I have the new data.frame (larger scale of the same area and
environmental variables) from which I have defined the new.env()
and xy.new.env() . Actually I can see the level.plot() of my variables
plotted in space.
When I run the next projection an GLM error mesaage appears.. I have
selected different models and in al off them the same error message
appears...
Please, do you know what can be happening?
The projection code and error message are:
myBiomodProj.sp <- BIOMOD_Projection(
modeling.output = myBiomodModelOut,
new.env = myExpl.sp, # the same variables (and colnames) than those used
for models
xy.new.env = myRespXY.sp, # the same variables (and colnames) than those
used for models
proj.name = 'current',
selected.models = "all",
binary.meth = 'TSS',
compress = 'xz',
clamping.mask = F,
output.format = '.RData') # R.Data because all are data.frame format
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Do Models Projections
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> Building clamping mask
> Projecting Rhizostoma.pulmo_AllData_RUN1_GLM ...
Error en UseMethod("predict") :
no applicable method for 'predict' applied to an object of class
"c('GLM_biomod2_model', 'biomod2_model')"
Thanks!,
Antonio
--
Antonio Canepa Oneto
PhD student
Dept. Biologia Marina i Oceanografia
Institut de Ciencies Del Mar CMIMA, CSIC
Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
BARCELONA, Catalunya,Spain
Phone: +34 932 30 9500 (ext. 6067)
Mobile: +34 628 320 778
e-mail: canepa at icm.csic.es
ajcanepa at gmail.com
From jduquelazo at gmail.com Mon May 13 11:24:41 2013
From: jduquelazo at gmail.com (Joaquin DL)
Date: Mon, 13 May 2013 11:24:41 +0200
Subject: [Biomod-commits] Variable Importance Function
Message-ID:
Dear all,
I am really interested in the variable importance function and it
performaces in side Biomod.
Could you suggest any reference where i could read more about it?
Thanks
Joaqu?n
From wilfried.thuiller at ujf-grenoble.fr Mon May 13 11:30:51 2013
From: wilfried.thuiller at ujf-grenoble.fr (Wilfried Thuiller)
Date: Mon, 13 May 2013 11:30:51 +0200
Subject: [Biomod-commits] Variable Importance Function
In-Reply-To:
References:
Message-ID: <5F68D099-81DC-4B57-BD04-47C611BDCAC1@ujf-grenoble.fr>
Dear Joaquin,
This is a simple permutation scheme, which more or less reproduces what is done in Random Forest for example. The only difference is that RandomForest performs the permutation over the test data, whereas in biomod2 we do that on the calibration data (although we could also allow the possibility of doing that on the evaluation data in the future).
You may take a look at this paper:
Gromping U. 2009. Variable Importance Assessment in Regression: Linear Regression versus Random Forest. American Statistician 63: 308-319.
I do not believe permutation is necessarily the best approach for each algorithm, but this is the only one that they can be applied to all algorithms in a standard and reproducible way, hence to compare the importance across the different models.
Hope it helps
Wilfried
Le 13 mai 2013 ? 11:24, Joaquin DL a ?crit :
> Dear all,
>
> I am really interested in the variable importance function and it
> performaces in side Biomod.
> Could you suggest any reference where i could read more about it?
>
> Thanks
>
> Joaqu?n
> _______________________________________________
> 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
Universit? Joseph Fourier
BP53, 38041 Grenoble cedex 9, France
tel: +33 (0)4 76 51 44 97
fax: +33 (0)4 76 51 42 79
Email: wilfried.thuiller at ujf-grenoble.fr
Personal website: http://www.will.chez-alice.fr
Team website: http://www-leca.ujf-grenoble.fr/equipes/emabio.htm
ERC Starting Grant TEEMBIO project: http://www.will.chez-alice.fr/Research.html
From damien.georges2 at gmail.com Mon May 13 12:03:40 2013
From: damien.georges2 at gmail.com (Damien Georges)
Date: Mon, 13 May 2013 12:03:40 +0200
Subject: [Biomod-commits] you have to load the projections yourself?
In-Reply-To:
References:
Message-ID: <5190BA7C.9020402@gmail.com>
Dear Josep,
I guess the issue you have comes from a specific combination of
parameters you used (keep.in.memory=F , do.stack=F ) that are not well
supported yet within biomod2.
If the projecting step is not too much time consuming, please re-run it
with do.stack=T then try to build ensemble-models
myBiomodProj.Present <- BIOMOD_Projection ( modeling.output =
myBiomodModelOut,
new.env = expl.var ,
proj.name = paste(gcm,'_strategy_',strategy,sep=''),
selected.models = 'all',
binary.meth = 'TSS',
compress = 'xz',
clamping.mask = T,
keep.in.memory=F ,
do.stack=T
)
Is it working better?
If not, please try this code:
rasterOptions(todisk=T)
myBiomodProj.Present <- BIOMOD_Projection ( modeling.output =
myBiomodModelOut,
new.env = expl.var ,
proj.name = paste(gcm,'_strategy_',strategy,sep=''),
selected.models = 'all',
binary.meth = 'TSS',
compress = 'xz',
clamping.mask = T,
keep.in.memory=T ,
do.stack=T
)
Hope that helps,
Best,
Damien.
On 09/05/2013 12:27, Josep M Serra diaz wrote:
> Dear List!
>
> am sure it is a silly error but I am getting stack...
>
> I did my modeling and my projection to present (hurra!)
>
> but when I want to apply my ensemble modeling rules to the projections...it
> says:
>
>
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Do
> Ensemble Models Projections
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>
>
> > Projecting Pinuspinaster_GLM_AllRun_EMbyTSS ...
> > em.mean
> !! You have to load your projections by yourself !!Error en
> raster:::mean(raster:::subset(getProjection(projection.output), :
> error in evaluating the argument 'x' in selecting a method for function
> 'mean': Error en subset.default(getProjection(projection.output),
> getEMkeptModels(EM.output, :
> 'subset' must be logical
>
> #################################################################
>
> .... SO THEN I TRIED DOING FIRST
>
> myPresenetProj <- getProjection(myBiomodProj.Present)
>
> but...error:
>
> !! You have to load your projections by yourself !!
>
>
> ##################################################################
>
> HERE YOU HAVE THE CODE I USED
>
>
>
> ##### PROJECTION OUTPUT
>
> myBiomodProj.Present <- BIOMOD_Projection ( modeling.output =
> myBiomodModelOut,
> new.env = expl.var ,
> proj.name = paste(gcm,'_strategy_',strategy,sep=''),
> selected.models = 'all',
> binary.meth = 'TSS',
> compress = 'xz',
> clamping.mask = T,
> keep.in.memory=F ,
> do.stack=F
> ) #end BIOMOD_PROJECTION
> # NOW THE ENSEMBLE FORECASTING FOR PROJECTIONS IN CRU FOR EACH ALGORITHM
>
> myPresentProj <- getProjection(myBiomodProj.Present)
>
> myBiomodProj.EF.algo.Present <- BIOMOD_EnsembleForecasting (
> projection.output= myBiomodProj.Present , # name of the projection
> EM.output =
> myBiomodEM.algo , # modeling rules
>
> binary.method= "TSS"
> ) #end
> BIOMOD_PROJECTION
>
>
> ######################################################################
> HERE YOU HAVE SESSION INFO
>
>> sessionInfo()
> R version 2.15.3 (2013-03-01)
> Platform: x86_64-w64-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=Spanish_Spain.1252 LC_CTYPE=Spanish_Spain.1252
> LC_MONETARY=Spanish_Spain.1252 LC_NUMERIC=C
> LC_TIME=Spanish_Spain.1252
>
> attached base packages:
> [1] splines grid stats graphics grDevices utils datasets
> methods base
>
> other attached packages:
> [1] mgcv_1.7-22 foreign_0.8-52 biomod2_2.1.20 pROC_1.5.4
> plyr_1.8 rpart_4.1-0 randomForest_4.6-7
> mda_0.4-2
> [9] class_7.3-5 gbm_2.0-8 survival_2.37-2 nnet_7.3-5
> rasterVis_0.20-05 hexbin_1.26.1 latticeExtra_0.6-24
> RColorBrewer_1.0-5
> [17] lattice_0.20-13 abind_1.4-0 raster_2.1-16 sp_1.0-8
>
>
> loaded via a namespace (and not attached):
> [1] Matrix_1.0-11 nlme_3.1-108 tools_2.15.3 zoo_1.7-9
>
>
> #################################################################
>
> ANY HELP IS GREATLY APPRECIATED THANKS A LOT
>
>
> Pep
> _______________________________________________
> 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 jduquelazo at gmail.com Tue May 14 08:01:54 2013
From: jduquelazo at gmail.com (Joaquin DL)
Date: Tue, 14 May 2013 08:01:54 +0200
Subject: [Biomod-commits] Bug Biomod2
Message-ID:
Dear all,
I found a little bug when projectiong single models using BIOMOD projection
function.
I am trying to project single models using this comand
selected.models = BIOMOD_LoadModels(myBiomodModelOut, models=c('MAXENT',
'FDA'), run.eval= 'RUN2'),
It should project only the model selected models in the run23 but it does
all that starts by 2, so 2,20,21..... and 29...
Thanks
Joaquin Duque Lazo
From damien.georges2 at gmail.com Tue May 14 10:50:29 2013
From: damien.georges2 at gmail.com (Damien Georges)
Date: Tue, 14 May 2013 10:50:29 +0200
Subject: [Biomod-commits] Bug Biomod2
In-Reply-To:
References:
Message-ID: <5191FAD5.3080303@gmail.com>
Dear Joaquin,
Thanks for this bug report. It's fixed now (biomod2_2.1.30)
Best,
Damien
On 14/05/2013 08:01, Joaquin DL wrote:
> Dear all,
> I found a little bug when projectiong single models using BIOMOD projection
> function.
> I am trying to project single models using this comand
> selected.models = BIOMOD_LoadModels(myBiomodModelOut, models=c('MAXENT',
> 'FDA'), run.eval= 'RUN2'),
> It should project only the model selected models in the run23 but it does
> all that starts by 2, so 2,20,21..... and 29...
>
> Thanks
>
> Joaquin Duque Lazo
> _______________________________________________
> 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 pep.bioalerts at gmail.com Tue May 14 10:51:04 2013
From: pep.bioalerts at gmail.com (Josep M Serra diaz)
Date: Tue, 14 May 2013 10:51:04 +0200
Subject: [Biomod-commits] you have to load the projections yourself?
In-Reply-To: <5190BA7C.9020402@gmail.com>
References:
<5190BA7C.9020402@gmail.com>
Message-ID:
Thanks for the suggestion! it did work taking out the options you mentioned
2013/5/13 Damien Georges
> Dear Josep,
>
> I guess the issue you have comes from a specific combination of parameters
> you used (keep.in.memory=F , do.stack=F ) that are not well supported yet
> within biomod2.
>
> If the projecting step is not too much time consuming, please re-run it
> with do.stack=T then try to build ensemble-models
>
> myBiomodProj.Present <- BIOMOD_Projection ( modeling.output =
> myBiomodModelOut,
> new.env = expl.var ,proj.name = paste(gcm,'_strategy_',
> strategy,sep=''),
> selected.models = 'all',
> binary.meth = 'TSS',
> compress = 'xz',
> clamping.mask = T,
> keep.in.memory=F ,
> do.stack=T
> )
>
> Is it working better?
>
> If not, please try this code:
>
> rasterOptions(todisk=T)
>
> myBiomodProj.Present <- BIOMOD_Projection ( modeling.output =
> myBiomodModelOut,
> new.env = expl.var ,proj.name = paste(gcm,'_strategy_',
> strategy,sep=''),
> selected.models = 'all',
> binary.meth = 'TSS',
> compress = 'xz',
> clamping.mask = T,
> keep.in.memory=T ,
> do.stack=T
> )
>
>
> Hope that helps,
> Best,
> Damien.
>
>
>
> On 09/05/2013 12:27, Josep M Serra diaz wrote:
>
> Dear List!
>
> am sure it is a silly error but I am getting stack...
>
> I did my modeling and my projection to present (hurra!)
>
> but when I want to apply my ensemble modeling rules to the projections...it
> says:
>
>
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Do
> Ensemble Models Projections
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
>
>
> > Projecting Pinuspinaster_GLM_AllRun_EMbyTSS ...
> > em.mean
> !! You have to load your projections by yourself !!Error en
> raster:::mean(raster:::subset(getProjection(projection.output), :
> error in evaluating the argument 'x' in selecting a method for function
> 'mean': Error en subset.default(getProjection(projection.output),
> getEMkeptModels(EM.output, :
> 'subset' must be logical
>
> #################################################################
>
> .... SO THEN I TRIED DOING FIRST
>
> myPresenetProj <- getProjection(myBiomodProj.Present)
>
> but...error:
>
> !! You have to load your projections by yourself !!
>
>
> ##################################################################
>
> HERE YOU HAVE THE CODE I USED
>
>
>
> ##### PROJECTION OUTPUT
>
> myBiomodProj.Present <- BIOMOD_Projection ( modeling.output =
> myBiomodModelOut,
> new.env = expl.var ,proj.name = paste(gcm,'_strategy_',strategy,sep=''),
> selected.models = 'all',
> binary.meth = 'TSS',
> compress = 'xz',
> clamping.mask = T,
> keep.in.memory=F ,
> do.stack=F
> ) #end BIOMOD_PROJECTION
> # NOW THE ENSEMBLE FORECASTING FOR PROJECTIONS IN CRU FOR EACH ALGORITHM
>
> myPresentProj <- getProjection(myBiomodProj.Present)
>
> myBiomodProj.EF.algo.Present <- BIOMOD_EnsembleForecasting (
> projection.output= myBiomodProj.Present , # name of the projection
> EM.output =
> myBiomodEM.algo , # modeling rules
>
> binary.method= "TSS"
> ) #end
> BIOMOD_PROJECTION
>
>
> ######################################################################
> HERE YOU HAVE SESSION INFO
>
>
> sessionInfo()
>
> R version 2.15.3 (2013-03-01)
> Platform: x86_64-w64-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=Spanish_Spain.1252 LC_CTYPE=Spanish_Spain.1252
> LC_MONETARY=Spanish_Spain.1252 LC_NUMERIC=C
> LC_TIME=Spanish_Spain.1252
>
> attached base packages:
> [1] splines grid stats graphics grDevices utils datasets
> methods base
>
> other attached packages:
> [1] mgcv_1.7-22 foreign_0.8-52 biomod2_2.1.20 pROC_1.5.4
> plyr_1.8 rpart_4.1-0 randomForest_4.6-7
> mda_0.4-2
> [9] class_7.3-5 gbm_2.0-8 survival_2.37-2 nnet_7.3-5
> rasterVis_0.20-05 hexbin_1.26.1 latticeExtra_0.6-24
> RColorBrewer_1.0-5
> [17] lattice_0.20-13 abind_1.4-0 raster_2.1-16 sp_1.0-8
>
>
> loaded via a namespace (and not attached):
> [1] Matrix_1.0-11 nlme_3.1-108 tools_2.15.3 zoo_1.7-9
>
>
> ############################################################
> #####
>
> ANY HELP IS GREATLY APPRECIATED THANKS A LOT
>
>
> Pep
> _______________________________________________
> Biomod-commits mailing listBiomod-commits at lists.r-forge.r-project.orghttps://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
>
>
>
From pep.bioalerts at gmail.com Tue May 14 10:59:15 2013
From: pep.bioalerts at gmail.com (Josep M Serra diaz)
Date: Tue, 14 May 2013 10:59:15 +0200
Subject: [Biomod-commits] error trying to read myBiomodModelOut
Message-ID:
Dear list,
I performed some modeling two months ago. I saved the session in a .RData
file.
Now I want to project them to future conditions so I loaded the sessions,
specified correctly the working directory, but when trying to project an
error arises. It seems that it is attempting at finding a slot name
modeling id, which is not stored or something.
I checked all other results of modeling and they seemed ok
(e.g.myBiomodData)
Any hint?
Pep
##########################################################
BiomodProj.Present <- BIOMOD_Projection (modeling.output = myBiomodModelOut,
+ new.env = expl.var ,
+ proj.name = paste('CRU_strategy_',strategy,sep=''),
+ selected.models = 'all',
+ binary.meth = 'TSS',
+ compress = 'xz',
+ clamping.mask = T
+ ) #end BIOMOD_PROJECTION
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Do Models
Projections -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
-=-=-=-=
Error en cat("\nModeling id :", object at modeling.id, fill = .Options$width)
:
there is no slot with name "modeling.id" for this object of class
"BIOMOD.models.out"
#######################################################################
> myBiomodModelOut
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
BIOMOD.models.out
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Error en cat("\nModeling id :", object at modeling.id, fill = .Options$width)
:
there is no slot with name "modeling.id" for this object of class
"BIOMOD.models.out"
###################################################################
> sessionInfo()
R version 3.0.0 (2013-04-03)
Platform: x86_64-w64-mingw32/x64 (64-bit)
locale:
[1] LC_COLLATE=Spanish_Spain.1252 LC_CTYPE=Spanish_Spain.1252
LC_MONETARY=Spanish_Spain.1252 LC_NUMERIC=C
[5] LC_TIME=Spanish_Spain.1252
attached base packages:
[1] parallel splines grid stats graphics grDevices utils
datasets methods base
other attached packages:
[1] mgcv_1.7-22 MASS_7.3-26 foreign_0.8-53
biomod2_2.1.27 pROC_1.5.4 plyr_1.8 rpart_4.1-1
[8] randomForest_4.6-7 mda_0.4-2 class_7.3-7 gbm_2.1
survival_2.37-4 nnet_7.3-6 rasterVis_0.20-07
[15] hexbin_1.26.2 latticeExtra_0.6-24 RColorBrewer_1.0-5
lattice_0.20-15 raster_2.1-25 sp_1.0-9 abind_1.4-0
loaded via a namespace (and not attached):
[1] Matrix_1.0-12 nlme_3.1-109 tools_3.0.0 zoo_1.7-10
From damien.georges2 at gmail.com Tue May 14 13:09:35 2013
From: damien.georges2 at gmail.com (Damien Georges)
Date: Tue, 14 May 2013 13:09:35 +0200
Subject: [Biomod-commits] error trying to read myBiomodModelOut
In-Reply-To:
References:
Message-ID: <51921B6F.2000406@gmail.com>
Dear Josep,
This is due to an add on biomod_modelling objects between the version
you used to build your models and the one curently installed on your
computer..
If models building is quite time consuming, you could try to use the
function attached to try to convert you Modeling outputs objects. If it
fails, maybe you will have to re-run it all.
Here the code of the function :
##########################################
update_biomod2_object <- function(obj){
## This function was created to convert if possible biomod2 objects
build with an old version of the package to the current one
## It can be particulary usefull in case of package updating.
## Damien G. - 13/05/14
## some constant definitions
supported_class <-
c("BIOMOD.formated.data","BIOMOD.formated.data.PA",
"BIOMOD.Model.Options", "BIOMOD.models.out", "BIOMOD.projection.out" )
## class checking
obj_class <- class(obj)
if( ! ( obj_class %in% supported_class ) ) stop("object to update
must inherit from one of : ", toString(supported_class), "class")
## create a reference object
ref_obj <- new(obj_class) #new(obj_class,obj)
## detect the slot to update
ref_slot <- slotNames(ref_obj)
to_update_slot <- sapply(ref_slot,function(X){
test <- try(slot(obj,X),silent=T)
if(inherits(test, "try-error")){
return(X)
} else{ return(NA) }
})
to_update_slot <- na.exclude(unlist(to_update_slot))
## update object
if(length(to_update_slot)){
for(t_u_s in to_update_slot){
# set a null modeling id
if(t_u_s == "modeling.id"){
obj at modeling.id = ""
} else{
slot(obj, t_u_s) <- slot(ref_obj, t_u_s)
}
}
cat("\n object updated!")
} else { cat("\n noting to do!") }
return(obj)
}
## Example
# myBiomodModelling <- update_biomod2_object(myBiomodModelling)
##########################################
Tell me if that helps,
Best,
Damien.
On 14/05/2013 10:59, Josep M Serra diaz wrote:
> Dear list,
>
> I performed some modeling two months ago. I saved the session in a
> .RData file.
>
> Now I want to project them to future conditions so I loaded the
> sessions, specified correctly the working directory, but when trying
> to project an error arises. It seems that it is attempting at finding
> a slot name modeling id, which is not stored or something.
>
> I checked all other results of modeling and they seemed ok
> (e.g.myBiomodData)
>
>
> Any hint?
>
> Pep
> ##########################################################
>
> BiomodProj.Present <- BIOMOD_Projection (modeling.output =
> myBiomodModelOut,
> + new.env = expl.var ,
> + proj.name = paste('CRU_strategy_',strategy,sep=''),
> + selected.models = 'all',
> + binary.meth = 'TSS',
> + compress = 'xz',
> + clamping.mask = T
> + ) #end BIOMOD_PROJECTION
>
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Do
> Models Projections
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> -=-=-=-=
> Error en cat("\nModeling id :", object at modeling.id
> , fill = .Options$width) :
> there is no slot with name "modeling.id " for
> this object of class "BIOMOD.models.out"
>
>
>
>
> #######################################################################
>
>
> > myBiomodModelOut
>
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> BIOMOD.models.out
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> Error en cat("\nModeling id :", object at modeling.id
> , fill = .Options$width) :
> there is no slot with name "modeling.id " for
> this object of class "BIOMOD.models.out"
>
>
>
> ###################################################################
>
> > sessionInfo()
> R version 3.0.0 (2013-04-03)
> Platform: x86_64-w64-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=Spanish_Spain.1252 LC_CTYPE=Spanish_Spain.1252
> LC_MONETARY=Spanish_Spain.1252 LC_NUMERIC=C
> [5] LC_TIME=Spanish_Spain.1252
>
> attached base packages:
> [1] parallel splines grid stats graphics grDevices utils
> datasets methods base
>
> other attached packages:
> [1] mgcv_1.7-22 MASS_7.3-26 foreign_0.8-53
> biomod2_2.1.27 pROC_1.5.4 plyr_1.8 rpart_4.1-1
> [8] randomForest_4.6-7 mda_0.4-2 class_7.3-7 gbm_2.1
> survival_2.37-4 nnet_7.3-6 rasterVis_0.20-07
> [15] hexbin_1.26.2 latticeExtra_0.6-24 RColorBrewer_1.0-5
> lattice_0.20-15 raster_2.1-25 sp_1.0-9 abind_1.4-0
>
> loaded via a namespace (and not attached):
> [1] Matrix_1.0-12 nlme_3.1-109 tools_3.0.0 zoo_1.7-10
>
From jduquelazo at gmail.com Wed May 15 09:04:47 2013
From: jduquelazo at gmail.com (Joaquin DL)
Date: Wed, 15 May 2013 09:04:47 +0200
Subject: [Biomod-commits] Error
Message-ID:
Dear All,
I got this error for all my models... Any idea to solve it
Thanks
Joaqu?n Duque Lazo
Model=Multiple Adaptive Regression SplinesError in UseMethod("predict") :
no applicable method for 'predict' applied to an object of class
"c('MARS_biomod2_model', 'biomod2_model')"
From damien.georges2 at gmail.com Wed May 15 10:11:31 2013
From: damien.georges2 at gmail.com (Damien Georges)
Date: Wed, 15 May 2013 10:11:31 +0200
Subject: [Biomod-commits] Error
In-Reply-To:
References:
Message-ID: <51934333.1060500@gmail.com>
Dear Joaquin,
This is a really odd issue that happens sometimes with windows
computers. I guess you can solve it detaching and reloading biomod2 package.
detach("package:biomod2", unload=TRUE)
library(biomod2)
If it's still not working, try to save your session, close R, open a new
R session, load biomod2 then load you old session.
Hope that helps,
Best,
Damien
On 15/05/2013 09:04, Joaquin DL wrote:
> Dear All,
>
> I got this error for all my models... Any idea to solve it
>
> Thanks
>
> Joaqu?n Duque Lazo
>
> Model=Multiple Adaptive Regression SplinesError in UseMethod("predict") :
> no applicable method for 'predict' applied to an object of class
> "c('MARS_biomod2_model', 'biomod2_model')"
> _______________________________________________
> 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 ajcanepa at gmail.com Wed May 15 10:36:35 2013
From: ajcanepa at gmail.com (Antonio Canepa)
Date: Wed, 15 May 2013 10:36:35 +0200
Subject: [Biomod-commits] Error
In-Reply-To: <51934333.1060500@gmail.com>
References:
<51934333.1060500@gmail.com>
Message-ID:
Hi Joaquin,
I'm a windows 7 user and the same problem arise (I guess) beacuse I tried
to load the .RData from a external disk.
The problem was solved by re-run and save all the session in the HDD of the
laptop.
Good luck,
Antonio
On 15 May 2013 10:11, Damien Georges wrote:
> Dear Joaquin,
>
> This is a really odd issue that happens sometimes with windows computers.
> I guess you can solve it detaching and reloading biomod2 package.
>
> detach("package:biomod2", unload=TRUE)
> library(biomod2)
>
> If it's still not working, try to save your session, close R, open a new R
> session, load biomod2 then load you old session.
>
> Hope that helps,
>
> Best,
> Damien
>
>
>
> On 15/05/2013 09:04, Joaquin DL wrote:
>
>> Dear All,
>>
>> I got this error for all my models... Any idea to solve it
>>
>> Thanks
>>
>> Joaqu?n Duque Lazo
>>
>> Model=Multiple Adaptive Regression SplinesError in UseMethod("predict") :
>> no applicable method for 'predict' applied to an object of class
>> "c('MARS_biomod2_model', 'biomod2_model')"
>> ______________________________**_________________
>> 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
>>
>
> ______________________________**_________________
> 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
>
--
Antonio Canepa Oneto
PhD student
Dept. Biologia Marina i Oceanografia
Institut de Ciencies Del Mar CMIMA, CSIC
Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
BARCELONA, Catalunya,Spain
Phone: +34 932 30 9500 (ext. 6067)
Mobile: +34 628 320 778
e-mail: canepa at icm.csic.es
ajcanepa at gmail.com
From ajcanepa at gmail.com Thu May 16 12:09:48 2013
From: ajcanepa at gmail.com (Antonio Canepa)
Date: Thu, 16 May 2013 12:09:48 +0200
Subject: [Biomod-commits] Comparison among variable importance
Message-ID:
Dear list,
I would like to know if you can directly compare among different models the
variable importance of each variable selected?
My guess is not, because when I got the importance of each variable (
getModelsVarImport()), the sum of all of them for each model was not = 1.
I don't know if I'm doing something wrong or there is another way to
"standardise" the output from getModelsVarImport() and thus compare
throughout models ?
--
Antonio Canepa Oneto
PhD student
Dept. Biologia Marina i Oceanografia
Institut de Ciencies Del Mar CMIMA, CSIC
Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
BARCELONA, Catalunya,Spain
Phone: +34 932 30 9500 (ext. 6067)
Mobile: +34 628 320 778
e-mail: canepa at icm.csic.es
ajcanepa at gmail.com
From damien.georges2 at gmail.com Thu May 16 12:39:27 2013
From: damien.georges2 at gmail.com (Damien Georges)
Date: Thu, 16 May 2013 12:39:27 +0200
Subject: [Biomod-commits] Comparison among variable importance
In-Reply-To:
References:
Message-ID: <5194B75F.3070301@gmail.com>
Dear Antonio,
I guess you are right and you need to make a kind of standardisation of
scores to make them all comparable. Make them sum to 1 is a good way to
do it.
Cheers,
Damien.
On 16/05/2013 12:09, Antonio Canepa wrote:
> Dear list,
> I would like to know if you can directly compare among different models the
> variable importance of each variable selected?
> My guess is not, because when I got the importance of each variable (
> getModelsVarImport()), the sum of all of them for each model was not = 1.
> I don't know if I'm doing something wrong or there is another way to
> "standardise" the output from getModelsVarImport() and thus compare
> throughout models ?
>
From vioreldpopescu at gmail.com Thu May 16 16:44:25 2013
From: vioreldpopescu at gmail.com (Viorel Popescu)
Date: Thu, 16 May 2013 07:44:25 -0700
Subject: [Biomod-commits] Comparison among variable importance
In-Reply-To: <5194B75F.3070301@gmail.com>
References:
<5194B75F.3070301@gmail.com>
Message-ID:
Hi everyone, I had a similar question a few weeks ago, and from my
understanding the varImportance scores are drawn from correlation
coefficients between different models (with and without the respective
variable permuted), namely, 1 minus the correlation (r). Hence, they cannot
add up to 1, the way variable importance is performed in MaxEnt. In that
case, standardizing the scores would not a be very good way to deal
with the varImportance. Am I missing something here?
The way I dealt with them (with advice from Wilfried) was to show boxplots
of varImportance score for each variable across modeling techniques (and
taxa if you need to make such comparisons), and refer to median scores
(distribution of scores was far from normal).
Cheers,
Viorel
On Thu, May 16, 2013 at 3:39 AM, Damien Georges
wrote:
> Dear Antonio,
>
> I guess you are right and you need to make a kind of standardisation of
> scores to make them all comparable. Make them sum to 1 is a good way to do
> it.
>
> Cheers,
> Damien.
>
>
>
> On 16/05/2013 12:09, Antonio Canepa wrote:
>
>> Dear list,
>> I would like to know if you can directly compare among different models
>> the
>> variable importance of each variable selected?
>> My guess is not, because when I got the importance of each variable (
>> getModelsVarImport()), the sum of all of them for each model was not = 1.
>> I don't know if I'm doing something wrong or there is another way to
>> "standardise" the output from getModelsVarImport() and thus compare
>> throughout models ?
>>
>>
> ______________________________**_________________
> 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
>
--
Viorel D. Popescu
David H. Smith Conservation Research Fellow
University of California - Santa Cruz &
Simon Fraser University, Biological Sciences
(604) 340 4228
https://sites.google.com/site/vioreldpopescu/
From wilfried.thuiller at ujf-grenoble.fr Thu May 16 18:34:18 2013
From: wilfried.thuiller at ujf-grenoble.fr (Wilfried Thuiller)
Date: Thu, 16 May 2013 18:34:18 +0200
Subject: [Biomod-commits] Comparison among variable importance
In-Reply-To:
References:
<5194B75F.3070301@gmail.com>
Message-ID:
Hi there all,
This is, indeed, a tricky question.
The way we perform variable importance, as do RandomForest, is really without unit. As Viorel mentioned, this is 1 minus the mean of correlations (depending on the number of variable importance runs). This is done independently (as Maxent is also doing as far as I understood). I do not think there is a sound statistical way of standardizing them. You may do as Damien suggest but I fear this is not really sound and should be interpreted without caution (but I am happy to be convinced otherwise).
As Viorel suggested, box plots would do the same more or less the same but in a more sound way to me.
Best,
Wilfried
Le 16 mai 2013 ? 16:44, Viorel Popescu a ?crit :
> Hi everyone, I had a similar question a few weeks ago, and from my
> understanding the varImportance scores are drawn from correlation
> coefficients between different models (with and without the respective
> variable permuted), namely, 1 minus the correlation (r). Hence, they cannot
> add up to 1, the way variable importance is performed in MaxEnt. In that
> case, standardizing the scores would not a be very good way to deal
> with the varImportance. Am I missing something here?
>
> The way I dealt with them (with advice from Wilfried) was to show boxplots
> of varImportance score for each variable across modeling techniques (and
> taxa if you need to make such comparisons), and refer to median scores
> (distribution of scores was far from normal).
>
> Cheers,
> Viorel
>
>
> On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> wrote:
>
>> Dear Antonio,
>>
>> I guess you are right and you need to make a kind of standardisation of
>> scores to make them all comparable. Make them sum to 1 is a good way to do
>> it.
>>
>> Cheers,
>> Damien.
>>
>>
>>
>> On 16/05/2013 12:09, Antonio Canepa wrote:
>>
>>> Dear list,
>>> I would like to know if you can directly compare among different models
>>> the
>>> variable importance of each variable selected?
>>> My guess is not, because when I got the importance of each variable (
>>> getModelsVarImport()), the sum of all of them for each model was not = 1.
>>> I don't know if I'm doing something wrong or there is another way to
>>> "standardise" the output from getModelsVarImport() and thus compare
>>> throughout models ?
>>>
>>>
>> ______________________________**_________________
>> 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
>>
>
>
>
> --
> Viorel D. Popescu
> David H. Smith Conservation Research Fellow
> University of California - Santa Cruz &
> Simon Fraser University, Biological Sciences
> (604) 340 4228
> https://sites.google.com/site/vioreldpopescu/
> _______________________________________________
> 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
Universit? Joseph Fourier
BP53, 38041 Grenoble cedex 9, France
tel: +33 (0)4 76 51 44 97
fax: +33 (0)4 76 51 42 79
Email: wilfried.thuiller at ujf-grenoble.fr
Personal website: http://www.will.chez-alice.fr
Team website: http://www-leca.ujf-grenoble.fr/equipes/emabio.htm
ERC Starting Grant TEEMBIO project: http://www.will.chez-alice.fr/Research.html
From ajcanepa at gmail.com Thu May 16 22:07:36 2013
From: ajcanepa at gmail.com (Antonio Canepa)
Date: Thu, 16 May 2013 22:07:36 +0200
Subject: [Biomod-commits] Comparison among variable importance
In-Reply-To:
References:
<5194B75F.3070301@gmail.com>
Message-ID:
Hi there!,
Many thanks for your help.
It's great that thsi list works so good.
Cheers,
Antonio
On 16 May 2013 18:34, Wilfried Thuiller
wrote:
> Hi there all,
>
> This is, indeed, a tricky question.
> The way we perform variable importance, as do RandomForest, is really
> without unit. As Viorel mentioned, this is 1 minus the mean of correlations
> (depending on the number of variable importance runs). This is done
> independently (as Maxent is also doing as far as I understood). I do not
> think there is a sound statistical way of standardizing them. You may do as
> Damien suggest but I fear this is not really sound and should be
> interpreted without caution (but I am happy to be convinced otherwise).
> As Viorel suggested, box plots would do the same more or less the same but
> in a more sound way to me.
>
> Best,
> Wilfried
>
>
>
>
>
> Le 16 mai 2013 ? 16:44, Viorel Popescu a ?crit
> :
>
> > Hi everyone, I had a similar question a few weeks ago, and from my
> > understanding the varImportance scores are drawn from correlation
> > coefficients between different models (with and without the respective
> > variable permuted), namely, 1 minus the correlation (r). Hence, they
> cannot
> > add up to 1, the way variable importance is performed in MaxEnt. In that
> > case, standardizing the scores would not a be very good way to deal
> > with the varImportance. Am I missing something here?
> >
> > The way I dealt with them (with advice from Wilfried) was to show
> boxplots
> > of varImportance score for each variable across modeling techniques (and
> > taxa if you need to make such comparisons), and refer to median scores
> > (distribution of scores was far from normal).
> >
> > Cheers,
> > Viorel
> >
> >
> > On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> > wrote:
> >
> >> Dear Antonio,
> >>
> >> I guess you are right and you need to make a kind of standardisation of
> >> scores to make them all comparable. Make them sum to 1 is a good way to
> do
> >> it.
> >>
> >> Cheers,
> >> Damien.
> >>
> >>
> >>
> >> On 16/05/2013 12:09, Antonio Canepa wrote:
> >>
> >>> Dear list,
> >>> I would like to know if you can directly compare among different models
> >>> the
> >>> variable importance of each variable selected?
> >>> My guess is not, because when I got the importance of each variable (
> >>> getModelsVarImport()), the sum of all of them for each model was not =
> 1.
> >>> I don't know if I'm doing something wrong or there is another way to
> >>> "standardise" the output from getModelsVarImport() and thus compare
> >>> throughout models ?
> >>>
> >>>
> >> ______________________________**_________________
> >> Biomod-commits mailing list
> >> Biomod-commits at lists.r-forge.**r-project.org<
> Biomod-commits at lists.r-forge.r-project.org>
> >> https://lists.r-forge.r-**project.org/cgi-bin/mailman/**
> >> listinfo/biomod-commits<
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
> >
> >>
> >
> >
> >
> > --
> > Viorel D. Popescu
> > David H. Smith Conservation Research Fellow
> > University of California - Santa Cruz &
> > Simon Fraser University, Biological Sciences
> > (604) 340 4228
> > https://sites.google.com/site/vioreldpopescu/
> > _______________________________________________
> > 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
> Universit? Joseph Fourier
> BP53, 38041 Grenoble cedex 9, France
> tel: +33 (0)4 76 51 44 97
> fax: +33 (0)4 76 51 42 79
>
> Email: wilfried.thuiller at ujf-grenoble.fr
> Personal website: http://www.will.chez-alice.fr
> Team website: http://www-leca.ujf-grenoble.fr/equipes/emabio.htm
>
> ERC Starting Grant TEEMBIO project:
> http://www.will.chez-alice.fr/Research.html
>
> _______________________________________________
> 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
>
--
Antonio Canepa Oneto
PhD student
Dept. Biologia Marina i Oceanografia
Institut de Ciencies Del Mar CMIMA, CSIC
Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
BARCELONA, Catalunya,Spain
Phone: +34 932 30 9500 (ext. 6067)
Mobile: +34 628 320 778
e-mail: canepa at icm.csic.es
ajcanepa at gmail.com
From jduquelazo at gmail.com Sun May 19 10:51:42 2013
From: jduquelazo at gmail.com (Joaquin DL)
Date: Sun, 19 May 2013 10:51:42 +0200
Subject: [Biomod-commits] Biomod-commits Digest, Vol 49, Issue 6
In-Reply-To:
References:
Message-ID:
I think this is a really interesting discussion!
But how can we do those boxplots?
If we include the variable importance output within all the runs, we will
consider"bad" model inside our variable comparison and include variables
that might be high correlated with this bad prediction disturbing our
result... I am correct?
Could it be posible to compare then between models according to their rank
( the variable that is more correlated with the prediction would be
classified as the most importance, rank 1) insted of their correlation
(weight)... I wont be able to identify the weight but can say which
variable influence more in the prediction... I might be wrong
Cheers
Joaquin
2013/5/17
> Send Biomod-commits mailing list submissions to
> biomod-commits at lists.r-forge.r-project.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
>
> or, via email, send a message with subject or body 'help' to
> biomod-commits-request at lists.r-forge.r-project.org
>
> You can reach the person managing the list at
> biomod-commits-owner at lists.r-forge.r-project.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Biomod-commits digest..."
>
>
> Today's Topics:
>
> 1. Comparison among variable importance (Antonio Canepa)
> 2. Re: Comparison among variable importance (Damien Georges)
> 3. Re: Comparison among variable importance (Viorel Popescu)
> 4. Re: Comparison among variable importance (Wilfried Thuiller)
> 5. Re: Comparison among variable importance (Antonio Canepa)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 16 May 2013 12:09:48 +0200
> From: Antonio Canepa
> To: biomod-commits at lists.r-forge.r-project.org
> Subject: [Biomod-commits] Comparison among variable importance
> Message-ID:
> <
> CAN6VVbSW5nkzmeYcmOB+brDLU43+vwL2Et91t8o8RSctDzUDSg at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Dear list,
> I would like to know if you can directly compare among different models the
> variable importance of each variable selected?
> My guess is not, because when I got the importance of each variable (
> getModelsVarImport()), the sum of all of them for each model was not = 1.
> I don't know if I'm doing something wrong or there is another way to
> "standardise" the output from getModelsVarImport() and thus compare
> throughout models ?
>
> --
> Antonio Canepa Oneto
> PhD student
> Dept. Biologia Marina i Oceanografia
> Institut de Ciencies Del Mar CMIMA, CSIC
> Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
> BARCELONA, Catalunya,Spain
> Phone: +34 932 30 9500 (ext. 6067)
> Mobile: +34 628 320 778
> e-mail: canepa at icm.csic.es
> ajcanepa at gmail.com
>
>
> ------------------------------
>
> Message: 2
> Date: Thu, 16 May 2013 12:39:27 +0200
> From: Damien Georges
> To: Antonio Canepa
> Cc: biomod-commits at lists.r-forge.r-project.org
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID: <5194B75F.3070301 at gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Dear Antonio,
>
> I guess you are right and you need to make a kind of standardisation of
> scores to make them all comparable. Make them sum to 1 is a good way to
> do it.
>
> Cheers,
> Damien.
>
>
> On 16/05/2013 12:09, Antonio Canepa wrote:
> > Dear list,
> > I would like to know if you can directly compare among different models
> the
> > variable importance of each variable selected?
> > My guess is not, because when I got the importance of each variable (
> > getModelsVarImport()), the sum of all of them for each model was not = 1.
> > I don't know if I'm doing something wrong or there is another way to
> > "standardise" the output from getModelsVarImport() and thus compare
> > throughout models ?
> >
>
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 16 May 2013 07:44:25 -0700
> From: Viorel Popescu
> To: Damien Georges
> Cc: biomod-commits at lists.r-forge.r-project.org
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID:
> 0PewXgcyA at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Hi everyone, I had a similar question a few weeks ago, and from my
> understanding the varImportance scores are drawn from correlation
> coefficients between different models (with and without the respective
> variable permuted), namely, 1 minus the correlation (r). Hence, they cannot
> add up to 1, the way variable importance is performed in MaxEnt. In that
> case, standardizing the scores would not a be very good way to deal
> with the varImportance. Am I missing something here?
>
> The way I dealt with them (with advice from Wilfried) was to show boxplots
> of varImportance score for each variable across modeling techniques (and
> taxa if you need to make such comparisons), and refer to median scores
> (distribution of scores was far from normal).
>
> Cheers,
> Viorel
>
>
> On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> wrote:
>
> > Dear Antonio,
> >
> > I guess you are right and you need to make a kind of standardisation of
> > scores to make them all comparable. Make them sum to 1 is a good way to
> do
> > it.
> >
> > Cheers,
> > Damien.
> >
> >
> >
> > On 16/05/2013 12:09, Antonio Canepa wrote:
> >
> >> Dear list,
> >> I would like to know if you can directly compare among different models
> >> the
> >> variable importance of each variable selected?
> >> My guess is not, because when I got the importance of each variable (
> >> getModelsVarImport()), the sum of all of them for each model was not =
> 1.
> >> I don't know if I'm doing something wrong or there is another way to
> >> "standardise" the output from getModelsVarImport() and thus compare
> >> throughout models ?
> >>
> >>
> > ______________________________**_________________
> > Biomod-commits mailing list
> > Biomod-commits at lists.r-forge.**r-project.org<
> Biomod-commits at lists.r-forge.r-project.org>
> > https://lists.r-forge.r-**project.org/cgi-bin/mailman/**
> > listinfo/biomod-commits<
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
> >
> >
>
>
>
> --
> Viorel D. Popescu
> David H. Smith Conservation Research Fellow
> University of California - Santa Cruz &
> Simon Fraser University, Biological Sciences
> (604) 340 4228
> https://sites.google.com/site/vioreldpopescu/
>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 16 May 2013 18:34:18 +0200
> From: Wilfried Thuiller
> To: Viorel Popescu
> Cc: biomod-commits at r-forge.wu-wien.ac.at
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID:
> Content-Type: text/plain; charset=iso-8859-1
>
> Hi there all,
>
> This is, indeed, a tricky question.
> The way we perform variable importance, as do RandomForest, is really
> without unit. As Viorel mentioned, this is 1 minus the mean of correlations
> (depending on the number of variable importance runs). This is done
> independently (as Maxent is also doing as far as I understood). I do not
> think there is a sound statistical way of standardizing them. You may do as
> Damien suggest but I fear this is not really sound and should be
> interpreted without caution (but I am happy to be convinced otherwise).
> As Viorel suggested, box plots would do the same more or less the same but
> in a more sound way to me.
>
> Best,
> Wilfried
>
>
>
>
>
> Le 16 mai 2013 ? 16:44, Viorel Popescu a ?crit
> :
>
> > Hi everyone, I had a similar question a few weeks ago, and from my
> > understanding the varImportance scores are drawn from correlation
> > coefficients between different models (with and without the respective
> > variable permuted), namely, 1 minus the correlation (r). Hence, they
> cannot
> > add up to 1, the way variable importance is performed in MaxEnt. In that
> > case, standardizing the scores would not a be very good way to deal
> > with the varImportance. Am I missing something here?
> >
> > The way I dealt with them (with advice from Wilfried) was to show
> boxplots
> > of varImportance score for each variable across modeling techniques (and
> > taxa if you need to make such comparisons), and refer to median scores
> > (distribution of scores was far from normal).
> >
> > Cheers,
> > Viorel
> >
> >
> > On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> > wrote:
> >
> >> Dear Antonio,
> >>
> >> I guess you are right and you need to make a kind of standardisation of
> >> scores to make them all comparable. Make them sum to 1 is a good way to
> do
> >> it.
> >>
> >> Cheers,
> >> Damien.
> >>
> >>
> >>
> >> On 16/05/2013 12:09, Antonio Canepa wrote:
> >>
> >>> Dear list,
> >>> I would like to know if you can directly compare among different models
> >>> the
> >>> variable importance of each variable selected?
> >>> My guess is not, because when I got the importance of each variable (
> >>> getModelsVarImport()), the sum of all of them for each model was not =
> 1.
> >>> I don't know if I'm doing something wrong or there is another way to
> >>> "standardise" the output from getModelsVarImport() and thus compare
> >>> throughout models ?
> >>>
> >>>
> >> ______________________________**_________________
> >> Biomod-commits mailing list
> >> Biomod-commits at lists.r-forge.**r-project.org<
> Biomod-commits at lists.r-forge.r-project.org>
> >> https://lists.r-forge.r-**project.org/cgi-bin/mailman/**
> >> listinfo/biomod-commits<
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
> >
> >>
> >
> >
> >
> > --
> > Viorel D. Popescu
> > David H. Smith Conservation Research Fellow
> > University of California - Santa Cruz &
> > Simon Fraser University, Biological Sciences
> > (604) 340 4228
> > https://sites.google.com/site/vioreldpopescu/
> > _______________________________________________
> > 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
> Universit? Joseph Fourier
> BP53, 38041 Grenoble cedex 9, France
> tel: +33 (0)4 76 51 44 97
> fax: +33 (0)4 76 51 42 79
>
> Email: wilfried.thuiller at ujf-grenoble.fr
> Personal website: http://www.will.chez-alice.fr
> Team website: http://www-leca.ujf-grenoble.fr/equipes/emabio.htm
>
> ERC Starting Grant TEEMBIO project:
> http://www.will.chez-alice.fr/Research.html
>
>
>
> ------------------------------
>
> Message: 5
> Date: Thu, 16 May 2013 22:07:36 +0200
> From: Antonio Canepa
> To: Wilfried Thuiller
> Cc: biomod-commits at r-forge.wu-wien.ac.at
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID:
> 71LFNNgNw_WjPxDtMKWH7s0_A at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Hi there!,
> Many thanks for your help.
> It's great that thsi list works so good.
>
> Cheers,
>
> Antonio
>
>
> On 16 May 2013 18:34, Wilfried Thuiller
> wrote:
>
> > Hi there all,
> >
> > This is, indeed, a tricky question.
> > The way we perform variable importance, as do RandomForest, is really
> > without unit. As Viorel mentioned, this is 1 minus the mean of
> correlations
> > (depending on the number of variable importance runs). This is done
> > independently (as Maxent is also doing as far as I understood). I do not
> > think there is a sound statistical way of standardizing them. You may do
> as
> > Damien suggest but I fear this is not really sound and should be
> > interpreted without caution (but I am happy to be convinced otherwise).
> > As Viorel suggested, box plots would do the same more or less the same
> but
> > in a more sound way to me.
> >
> > Best,
> > Wilfried
> >
> >
> >
> >
> >
> > Le 16 mai 2013 ? 16:44, Viorel Popescu a
> ?crit
> > :
> >
> > > Hi everyone, I had a similar question a few weeks ago, and from my
> > > understanding the varImportance scores are drawn from correlation
> > > coefficients between different models (with and without the respective
> > > variable permuted), namely, 1 minus the correlation (r). Hence, they
> > cannot
> > > add up to 1, the way variable importance is performed in MaxEnt. In
> that
> > > case, standardizing the scores would not a be very good way to deal
> > > with the varImportance. Am I missing something here?
> > >
> > > The way I dealt with them (with advice from Wilfried) was to show
> > boxplots
> > > of varImportance score for each variable across modeling techniques
> (and
> > > taxa if you need to make such comparisons), and refer to median scores
> > > (distribution of scores was far from normal).
> > >
> > > Cheers,
> > > Viorel
> > >
> > >
> > > On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> > > wrote:
> > >
> > >> Dear Antonio,
> > >>
> > >> I guess you are right and you need to make a kind of standardisation
> of
> > >> scores to make them all comparable. Make them sum to 1 is a good way
> to
> > do
> > >> it.
> > >>
> > >> Cheers,
> > >> Damien.
> > >>
> > >>
> > >>
> > >> On 16/05/2013 12:09, Antonio Canepa wrote:
> > >>
> > >>> Dear list,
> > >>> I would like to know if you can directly compare among different
> models
> > >>> the
> > >>> variable importance of each variable selected?
> > >>> My guess is not, because when I got the importance of each variable (
> > >>> getModelsVarImport()), the sum of all of them for each model was not
> =
> > 1.
> > >>> I don't know if I'm doing something wrong or there is another way to
> > >>> "standardise" the output from getModelsVarImport() and thus compare
> > >>> throughout models ?
> > >>>
> > >>>
> > >> ______________________________**_________________
> > >> Biomod-commits mailing list
> > >> Biomod-commits at lists.r-forge.**r-project.org<
> > Biomod-commits at lists.r-forge.r-project.org>
> > >> https://lists.r-forge.r-**project.org/cgi-bin/mailman/**
> > >> listinfo/biomod-commits<
> >
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
> > >
> > >>
> > >
> > >
> > >
> > > --
> > > Viorel D. Popescu
> > > David H. Smith Conservation Research Fellow
> > > University of California - Santa Cruz &
> > > Simon Fraser University, Biological Sciences
> > > (604) 340 4228
> > > https://sites.google.com/site/vioreldpopescu/
> > > _______________________________________________
> > > 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
> > Universit? Joseph Fourier
> > BP53, 38041 Grenoble cedex 9, France
> > tel: +33 (0)4 76 51 44 97
> > fax: +33 (0)4 76 51 42 79
> >
> > Email: wilfried.thuiller at ujf-grenoble.fr
> > Personal website: http://www.will.chez-alice.fr
> > Team website: http://www-leca.ujf-grenoble.fr/equipes/emabio.htm
> >
> > ERC Starting Grant TEEMBIO project:
> > http://www.will.chez-alice.fr/Research.html
> >
> > _______________________________________________
> > 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
> >
>
>
>
> --
> Antonio Canepa Oneto
> PhD student
> Dept. Biologia Marina i Oceanografia
> Institut de Ciencies Del Mar CMIMA, CSIC
> Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
> BARCELONA, Catalunya,Spain
> Phone: +34 932 30 9500 (ext. 6067)
> Mobile: +34 628 320 778
> e-mail: canepa at icm.csic.es
> ajcanepa 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
>
> End of Biomod-commits Digest, Vol 49, Issue 6
> *********************************************
>
From jduquelazo at gmail.com Mon May 20 18:24:46 2013
From: jduquelazo at gmail.com (Joaquin DL)
Date: Mon, 20 May 2013 18:24:46 +0200
Subject: [Biomod-commits] Comparison among variable importance
Message-ID:
I think this is a really interesting discussion!
But how can we do those boxplots?
If we include the variable importance output within all the runs, we will
consider"bad" model inside our variable comparison and include variables
that might be high correlated with this bad prediction disturbing our
result... I am correct?
Could it be posible to compare variables between models according to their
rank ( the variable that is more correlated with the prediction would be
classified as the most importance, rank 1) insted of their correlation
(weight)... I wont be able to identify the weight but can say which
variable influence more in the prediction... I might be wrong
I found a paper where they did it like that. have a look to: Syphard, A.D.
& Franklin, J. (2009) Differences in spatial predictions among species
distribution modeling methods vary with species traits and environmental
predictors. Ecography, 32, 907-918.
Cheers
Joaquin
2013/5/17
> Send Biomod-commits mailing list submissions to
> biomod-commits at lists.r-forge.r-project.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
>
> or, via email, send a message with subject or body 'help' to
> biomod-commits-request at lists.r-forge.r-project.org
>
> You can reach the person managing the list at
> biomod-commits-owner at lists.r-forge.r-project.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Biomod-commits digest..."
>
>
> Today's Topics:
>
> 1. Comparison among variable importance (Antonio Canepa)
> 2. Re: Comparison among variable importance (Damien Georges)
> 3. Re: Comparison among variable importance (Viorel Popescu)
> 4. Re: Comparison among variable importance (Wilfried Thuiller)
> 5. Re: Comparison among variable importance (Antonio Canepa)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 16 May 2013 12:09:48 +0200
> From: Antonio Canepa
> To: biomod-commits at lists.r-forge.r-project.org
> Subject: [Biomod-commits] Comparison among variable importance
> Message-ID:
> <
> CAN6VVbSW5nkzmeYcmOB+brDLU43+vwL2Et91t8o8RSctDzUDSg at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Dear list,
> I would like to know if you can directly compare among different models the
> variable importance of each variable selected?
> My guess is not, because when I got the importance of each variable (
> getModelsVarImport()), the sum of all of them for each model was not = 1.
> I don't know if I'm doing something wrong or there is another way to
> "standardise" the output from getModelsVarImport() and thus compare
> throughout models ?
>
> --
> Antonio Canepa Oneto
> PhD student
> Dept. Biologia Marina i Oceanografia
> Institut de Ciencies Del Mar CMIMA, CSIC
> Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
> BARCELONA, Catalunya,Spain
> Phone: +34 932 30 9500 (ext. 6067)
> Mobile: +34 628 320 778
> e-mail: canepa at icm.csic.es
> ajcanepa at gmail.com
>
>
> ------------------------------
>
> Message: 2
> Date: Thu, 16 May 2013 12:39:27 +0200
> From: Damien Georges
> To: Antonio Canepa
> Cc: biomod-commits at lists.r-forge.r-project.org
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID: <5194B75F.3070301 at gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Dear Antonio,
>
> I guess you are right and you need to make a kind of standardisation of
> scores to make them all comparable. Make them sum to 1 is a good way to
> do it.
>
> Cheers,
> Damien.
>
>
> On 16/05/2013 12:09, Antonio Canepa wrote:
> > Dear list,
> > I would like to know if you can directly compare among different models
> the
> > variable importance of each variable selected?
> > My guess is not, because when I got the importance of each variable (
> > getModelsVarImport()), the sum of all of them for each model was not = 1.
> > I don't know if I'm doing something wrong or there is another way to
> > "standardise" the output from getModelsVarImport() and thus compare
> > throughout models ?
> >
>
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 16 May 2013 07:44:25 -0700
> From: Viorel Popescu
> To: Damien Georges
> Cc: biomod-commits at lists.r-forge.r-project.org
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID:
> 0PewXgcyA at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Hi everyone, I had a similar question a few weeks ago, and from my
> understanding the varImportance scores are drawn from correlation
> coefficients between different models (with and without the respective
> variable permuted), namely, 1 minus the correlation (r). Hence, they cannot
> add up to 1, the way variable importance is performed in MaxEnt. In that
> case, standardizing the scores would not a be very good way to deal
> with the varImportance. Am I missing something here?
>
> The way I dealt with them (with advice from Wilfried) was to show boxplots
> of varImportance score for each variable across modeling techniques (and
> taxa if you need to make such comparisons), and refer to median scores
> (distribution of scores was far from normal).
>
> Cheers,
> Viorel
>
>
> On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> wrote:
>
> > Dear Antonio,
> >
> > I guess you are right and you need to make a kind of standardisation of
> > scores to make them all comparable. Make them sum to 1 is a good way to
> do
> > it.
> >
> > Cheers,
> > Damien.
> >
> >
> >
> > On 16/05/2013 12:09, Antonio Canepa wrote:
> >
> >> Dear list,
> >> I would like to know if you can directly compare among different models
> >> the
> >> variable importance of each variable selected?
> >> My guess is not, because when I got the importance of each variable (
> >> getModelsVarImport()), the sum of all of them for each model was not =
> 1.
> >> I don't know if I'm doing something wrong or there is another way to
> >> "standardise" the output from getModelsVarImport() and thus compare
> >> throughout models ?
> >>
> >>
> > ______________________________**_________________
> > Biomod-commits mailing list
> > Biomod-commits at lists.r-forge.**r-project.org<
> Biomod-commits at lists.r-forge.r-project.org>
> > https://lists.r-forge.r-**project.org/cgi-bin/mailman/**
> > listinfo/biomod-commits<
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
> >
> >
>
>
>
> --
> Viorel D. Popescu
> David H. Smith Conservation Research Fellow
> University of California - Santa Cruz &
> Simon Fraser University, Biological Sciences
> (604) 340 4228
> https://sites.google.com/site/vioreldpopescu/
>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 16 May 2013 18:34:18 +0200
> From: Wilfried Thuiller
> To: Viorel Popescu
> Cc: biomod-commits at r-forge.wu-wien.ac.at
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID:
> Content-Type: text/plain; charset=iso-8859-1
>
> Hi there all,
>
> This is, indeed, a tricky question.
> The way we perform variable importance, as do RandomForest, is really
> without unit. As Viorel mentioned, this is 1 minus the mean of correlations
> (depending on the number of variable importance runs). This is done
> independently (as Maxent is also doing as far as I understood). I do not
> think there is a sound statistical way of standardizing them. You may do as
> Damien suggest but I fear this is not really sound and should be
> interpreted without caution (but I am happy to be convinced otherwise).
> As Viorel suggested, box plots would do the same more or less the same but
> in a more sound way to me.
>
> Best,
> Wilfried
>
>
>
>
>
> Le 16 mai 2013 ? 16:44, Viorel Popescu a ?crit
> :
>
> > Hi everyone, I had a similar question a few weeks ago, and from my
> > understanding the varImportance scores are drawn from correlation
> > coefficients between different models (with and without the respective
> > variable permuted), namely, 1 minus the correlation (r). Hence, they
> cannot
> > add up to 1, the way variable importance is performed in MaxEnt. In that
> > case, standardizing the scores would not a be very good way to deal
> > with the varImportance. Am I missing something here?
> >
> > The way I dealt with them (with advice from Wilfried) was to show
> boxplots
> > of varImportance score for each variable across modeling techniques (and
> > taxa if you need to make such comparisons), and refer to median scores
> > (distribution of scores was far from normal).
> >
> > Cheers,
> > Viorel
> >
> >
> > On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> > wrote:
> >
> >> Dear Antonio,
> >>
> >> I guess you are right and you need to make a kind of standardisation of
> >> scores to make them all comparable. Make them sum to 1 is a good way to
> do
> >> it.
> >>
> >> Cheers,
> >> Damien.
> >>
> >>
> >>
> >> On 16/05/2013 12:09, Antonio Canepa wrote:
> >>
> >>> Dear list,
> >>> I would like to know if you can directly compare among different models
> >>> the
> >>> variable importance of each variable selected?
> >>> My guess is not, because when I got the importance of each variable (
> >>> getModelsVarImport()), the sum of all of them for each model was not =
> 1.
> >>> I don't know if I'm doing something wrong or there is another way to
> >>> "standardise" the output from getModelsVarImport() and thus compare
> >>> throughout models ?
> >>>
> >>>
> >> ______________________________**_________________
> >> Biomod-commits mailing list
> >> Biomod-commits at lists.r-forge.**r-project.org<
> Biomod-commits at lists.r-forge.r-project.org>
> >> https://lists.r-forge.r-**project.org/cgi-bin/mailman/**
> >> listinfo/biomod-commits<
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
> >
> >>
> >
> >
> >
> > --
> > Viorel D. Popescu
> > David H. Smith Conservation Research Fellow
> > University of California - Santa Cruz &
> > Simon Fraser University, Biological Sciences
> > (604) 340 4228
> > https://sites.google.com/site/vioreldpopescu/
> > _______________________________________________
> > 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
> Universit? Joseph Fourier
> BP53, 38041 Grenoble cedex 9, France
> tel: +33 (0)4 76 51 44 97
> fax: +33 (0)4 76 51 42 79
>
> Email: wilfried.thuiller at ujf-grenoble.fr
> Personal website: http://www.will.chez-alice.fr
> Team website: http://www-leca.ujf-grenoble.fr/equipes/emabio.htm
>
> ERC Starting Grant TEEMBIO project:
> http://www.will.chez-alice.fr/Research.html
>
>
>
> ------------------------------
>
> Message: 5
> Date: Thu, 16 May 2013 22:07:36 +0200
> From: Antonio Canepa
> To: Wilfried Thuiller
> Cc: biomod-commits at r-forge.wu-wien.ac.at
> Subject: Re: [Biomod-commits] Comparison among variable importance
> Message-ID:
> 71LFNNgNw_WjPxDtMKWH7s0_A at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Hi there!,
> Many thanks for your help.
> It's great that thsi list works so good.
>
> Cheers,
>
> Antonio
>
>
> On 16 May 2013 18:34, Wilfried Thuiller
> wrote:
>
> > Hi there all,
> >
> > This is, indeed, a tricky question.
> > The way we perform variable importance, as do RandomForest, is really
> > without unit. As Viorel mentioned, this is 1 minus the mean of
> correlations
> > (depending on the number of variable importance runs). This is done
> > independently (as Maxent is also doing as far as I understood). I do not
> > think there is a sound statistical way of standardizing them. You may do
> as
> > Damien suggest but I fear this is not really sound and should be
> > interpreted without caution (but I am happy to be convinced otherwise).
> > As Viorel suggested, box plots would do the same more or less the same
> but
> > in a more sound way to me.
> >
> > Best,
> > Wilfried
> >
> >
> >
> >
> >
> > Le 16 mai 2013 ? 16:44, Viorel Popescu a
> ?crit
> > :
> >
> > > Hi everyone, I had a similar question a few weeks ago, and from my
> > > understanding the varImportance scores are drawn from correlation
> > > coefficients between different models (with and without the respective
> > > variable permuted), namely, 1 minus the correlation (r). Hence, they
> > cannot
> > > add up to 1, the way variable importance is performed in MaxEnt. In
> that
> > > case, standardizing the scores would not a be very good way to deal
> > > with the varImportance. Am I missing something here?
> > >
> > > The way I dealt with them (with advice from Wilfried) was to show
> > boxplots
> > > of varImportance score for each variable across modeling techniques
> (and
> > > taxa if you need to make such comparisons), and refer to median scores
> > > (distribution of scores was far from normal).
> > >
> > > Cheers,
> > > Viorel
> > >
> > >
> > > On Thu, May 16, 2013 at 3:39 AM, Damien Georges
> > > wrote:
> > >
> > >> Dear Antonio,
> > >>
> > >> I guess you are right and you need to make a kind of standardisation
> of
> > >> scores to make them all comparable. Make them sum to 1 is a good way
> to
> > do
> > >> it.
> > >>
> > >> Cheers,
> > >> Damien.
> > >>
> > >>
> > >>
> > >> On 16/05/2013 12:09, Antonio Canepa wrote:
> > >>
> > >>> Dear list,
> > >>> I would like to know if you can directly compare among different
> models
> > >>> the
> > >>> variable importance of each variable selected?
> > >>> My guess is not, because when I got the importance of each variable (
> > >>> getModelsVarImport()), the sum of all of them for each model was not
> =
> > 1.
> > >>> I don't know if I'm doing something wrong or there is another way to
> > >>> "standardise" the output from getModelsVarImport() and thus compare
> > >>> throughout models ?
> > >>>
> > >>>
> > >> ______________________________**_________________
> > >> Biomod-commits mailing list
> > >> Biomod-commits at lists.r-forge.**r-project.org<
> > Biomod-commits at lists.r-forge.r-project.org>
> > >> https://lists.r-forge.r-**project.org/cgi-bin/mailman/**
> > >> listinfo/biomod-commits<
> >
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/biomod-commits
> > >
> > >>
> > >
> > >
> > >
> > > --
> > > Viorel D. Popescu
> > > David H. Smith Conservation Research Fellow
> > > University of California - Santa Cruz &
> > > Simon Fraser University, Biological Sciences
> > > (604) 340 4228
> > > https://sites.google.com/site/vioreldpopescu/
> > > _______________________________________________
> > > 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
> > Universit? Joseph Fourier
> > BP53, 38041 Grenoble cedex 9, France
> > tel: +33 (0)4 76 51 44 97
> > fax: +33 (0)4 76 51 42 79
> >
> > Email: wilfried.thuiller at ujf-grenoble.fr
> > Personal website: http://www.will.chez-alice.fr
> > Team website: http://www-leca.ujf-grenoble.fr/equipes/emabio.htm
> >
> > ERC Starting Grant TEEMBIO project:
> > http://www.will.chez-alice.fr/Research.html
> >
> > _______________________________________________
> > 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
> >
>
>
>
> --
> Antonio Canepa Oneto
> PhD student
> Dept. Biologia Marina i Oceanografia
> Institut de Ciencies Del Mar CMIMA, CSIC
> Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
> BARCELONA, Catalunya,Spain
> Phone: +34 932 30 9500 (ext. 6067)
> Mobile: +34 628 320 778
> e-mail: canepa at icm.csic.es
> ajcanepa 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
>
> End of Biomod-commits Digest, Vol 49, Issue 6
> *********************************************
>
From ajcanepa at gmail.com Tue May 21 23:20:04 2013
From: ajcanepa at gmail.com (Antonio Canepa)
Date: Tue, 21 May 2013 23:20:04 +0200
Subject: [Biomod-commits] How to plot individual projections?
Message-ID:
Dear all,
I have a silly question but I can't find the solution.
I have my myBiomodProj, using the BIOMOD_Projection() function.
Now I have like 36 projected models (4 models, 3PA, 3RUN), but I would like
to plot them, individually). In the help file the most closer solution is:
plot(myBiomodProj, str.grep = 'GAM'), but all the 3 Runs and 3 PA (9 plots)
appears.
Also I would like to change the color of the probabilities but I can't
modify it through the color() command.
I appreciate so much if you can give a hand on it!,
Thank you, so much.
Cheers,
Antonio
--
Antonio Canepa Oneto
PhD student
Dept. Biologia Marina i Oceanografia
Institut de Ciencies Del Mar CMIMA, CSIC
Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
BARCELONA, Catalunya,Spain
Phone: +34 932 30 9500 (ext. 6067)
Mobile: +34 628 320 778
e-mail: canepa at icm.csic.es
ajcanepa at gmail.com
From postmaster at r-forge.wu-wien.ac.at Wed May 22 07:07:09 2013
From: postmaster at r-forge.wu-wien.ac.at (Returned mail)
Date: Wed, 22 May 2013 12:07:09 +0700
Subject: [Biomod-commits] (no subject)
Message-ID:
Dear user biomod-commits at r-forge.wu-wien.ac.at,
We have received reports that your account was used to send a huge amount of junk e-mail messages during the last week.
Most likely your computer had been infected by a recent virus and now runs a hidden proxy server.
Please follow our instructions in order to keep your computer safe.
Sincerely yours,
r-forge.wu-wien.ac.at technical support team.
From damien.georges2 at gmail.com Wed May 22 10:22:25 2013
From: damien.georges2 at gmail.com (Damien Georges)
Date: Wed, 22 May 2013 10:22:25 +0200
Subject: [Biomod-commits] How to plot individual projections?
In-Reply-To:
References:
Message-ID: <519C8041.8020508@gmail.com>
Dear Antonio,
The simplest way to process if you want to produce non -default plots is
to extract the 'formal' projection information and to plot it with
non-biomod plotting function.
You can access "formal" projection values with getProjection(...)
function (e.g. proj_val <- getProjection(myBiomodProj) ). Then,
depending on if you made your projections with environmental raster or
data.frame, the plotting function you will use will change.
for RasterStack environment :
plot(subset(proj_val,1)) will plot the first layer (i.e your first model
projection)
for data.frame, you have to get the XY data.fame that correspond to your
projection points. Then you can make your plots with level.plot function
(see ?level.plot for examples)
the last thing you can do is to use the biomod2 plotting function you
mentioned and increase the str.grep arg to precise the selection(e.g
plot(myBiomodProj, str.grep = 'PA1_RUN1_GAM')
Hope that helps,
Damien.
On 21/05/2013 23:20, Antonio Canepa wrote:
> Dear all,
> I have a silly question but I can't find the solution.
> I have my myBiomodProj, using the BIOMOD_Projection() function.
> Now I have like 36 projected models (4 models, 3PA, 3RUN), but I would like
> to plot them, individually). In the help file the most closer solution is:
> plot(myBiomodProj, str.grep = 'GAM'), but all the 3 Runs and 3 PA (9 plots)
> appears.
> Also I would like to change the color of the probabilities but I can't
> modify it through the color() command.
> I appreciate so much if you can give a hand on it!,
> Thank you, so much.
>
> Cheers,
>
> Antonio
>
From ajcanepa at gmail.com Wed May 22 11:16:42 2013
From: ajcanepa at gmail.com (Antonio Canepa)
Date: Wed, 22 May 2013 11:16:42 +0200
Subject: [Biomod-commits] How to plot individual projections?
In-Reply-To: <519C8041.8020508@gmail.com>
References:
<519C8041.8020508@gmail.com>
Message-ID:
Hi Damien,
Thanks for the advice. the command plot(subset(proj_val,1)) works fine.
Cheers,
Antonio
On 22 May 2013 10:22, Damien Georges wrote:
> Dear Antonio,
>
> The simplest way to process if you want to produce non -default plots is
> to extract the 'formal' projection information and to plot it with
> non-biomod plotting function.
>
> You can access "formal" projection values with getProjection(...) function
> (e.g. proj_val <- getProjection(myBiomodProj) ). Then, depending on if you
> made your projections with environmental raster or data.frame, the plotting
> function you will use will change.
>
> for RasterStack environment :
> plot(subset(proj_val,1)) will plot the first layer (i.e your first model
> projection)
>
> for data.frame, you have to get the XY data.fame that correspond to your
> projection points. Then you can make your plots with level.plot function
> (see ?level.plot for examples)
>
> the last thing you can do is to use the biomod2 plotting function you
> mentioned and increase the str.grep arg to precise the selection(e.g
> plot(myBiomodProj, str.grep = 'PA1_RUN1_GAM')
>
> Hope that helps,
> Damien.
>
>
>
> On 21/05/2013 23:20, Antonio Canepa wrote:
>
> Dear all,
> I have a silly question but I can't find the solution.
> I have my myBiomodProj, using the BIOMOD_Projection() function.
> Now I have like 36 projected models (4 models, 3PA, 3RUN), but I would like
> to plot them, individually). In the help file the most closer solution is:
> plot(myBiomodProj, str.grep = 'GAM'), but all the 3 Runs and 3 PA (9 plots)
> appears.
> Also I would like to change the color of the probabilities but I can't
> modify it through the color() command.
> I appreciate so much if you can give a hand on it!,
> Thank you, so much.
>
> Cheers,
>
> Antonio
>
>
>
>
--
Antonio Canepa Oneto
PhD student
Dept. Biologia Marina i Oceanografia
Institut de Ciencies Del Mar CMIMA, CSIC
Ps. Mar?tim de la Barceloneta,37-49, c.p. 08003
BARCELONA, Catalunya,Spain
Phone: +34 932 30 9500 (ext. 6067)
Mobile: +34 628 320 778
e-mail: canepa at icm.csic.es
ajcanepa at gmail.com
From kpitts at siu.edu Fri May 24 21:32:31 2013
From: kpitts at siu.edu (Kristen Bouska)
Date: Fri, 24 May 2013 14:32:31 -0500
Subject: [Biomod-commits] Question regarding training/testing/evaluating
data sets
Message-ID:
I wanted to make sure I am understanding my model commands correctly. If I
have the following commands (below), I am calling in a training data set
(expl.var) and an evaluation data set (eval.expl.var), then I am
cross-validating the training data set 10 times, splitting it 75/25 for
each cross-validation. If this is correct, then my evaluation data set
(eval.expl.var) is not used in the BIOMOD_Modeling evaluation step,
correct? Is the evaluation data set only used to evaluate the ensembled
models?
Sorry if this is confusing, I just want to make sure I am doing what I
think I am doing when I am running my models.
myBiomodData <- BIOMOD_FormatingData(resp.var=myResp, resp.xy=myRespXY,
expl.var=myExpl, resp.name=myRespName, eval.resp.var=myEvalpa,
eval.expl.var=myEvalenv, eval.resp.xy=myRespEvalXY)
myBiomodModelOut <- BIOMOD_Modeling(myBiomodData, models = c('GLM', 'GBM',
'CTA', 'RF', 'MARS'), models.options=NULL, NbRunEval=10, DataSplit=75,
Yweights=NULL, Prevalence=NULL, VarImport=3, models.eval.meth =
c('KAPPA','TSS','ROC'),SaveObj = TRUE, rescal.all.models = TRUE,
do.full.models = TRUE, modeling.id = ?test_date? )
Thanks!
Kristen
--
Kristen Bouska
PhD Candidate
Environmental Resources and Policy Program
Life Sciences II, Room 375 (Mail Code 4325)
Southern Illinois University - Carbondale
Carbondale, IL 62901
618-924-2592
From jeaggu at gmail.com Wed May 29 09:38:27 2013
From: jeaggu at gmail.com (Jesus)
Date: Wed, 29 May 2013 09:38:27 +0200
Subject: [Biomod-commits] MaxLike
Message-ID:
Hi Damien, I was wondering if you have thought about implementing MaxLike
(Royle et al. 2012, MEE) in the BIOMOD package? have you considered it?
Cheers,
Jes?s
From damien.georges2 at gmail.com Wed May 29 14:17:42 2013
From: damien.georges2 at gmail.com (Damien Georges)
Date: Wed, 29 May 2013 14:17:42 +0200
Subject: [Biomod-commits] Question regarding training/testing/evaluating
data sets
In-Reply-To:
References:
Message-ID: <51A5F1E6.3080001@gmail.com>
Dear Kristen,
In fact if you give a specific independent dataset for model evaluation
this dataset will be use both for all models evaluations. In case you
decide to do do cross validation (i.e. 75/25) , models will be calibrate
with 75% of expl.var data, then the optimal threshold to convert
continuous prediction to binaries (needed to evaluate models) will
determined with 25% of expl.var dataset. Then the model will evaluation
score will be calculated using eval.expl.var dataset (using threshold
calculate at previous step).
Ensemble models will be evaluated with same thresholds (25% of expl.var)
and same reference data (eval.expl.var).
Working with this "3 separated dataset" is the fairest way to construct
your models. If you have data enough, I recommend you to follow this way
to work.
Hope that helps,
Damien
On 24/05/2013 21:32, Kristen Bouska wrote:
> I wanted to make sure I am understanding my model commands correctly. If I
> have the following commands (below), I am calling in a training data set
> (expl.var) and an evaluation data set (eval.expl.var), then I am
> cross-validating the training data set 10 times, splitting it 75/25 for
> each cross-validation. If this is correct, then my evaluation data set
> (eval.expl.var) is not used in the BIOMOD_Modeling evaluation step,
> correct? Is the evaluation data set only used to evaluate the ensembled
> models?
>
> Sorry if this is confusing, I just want to make sure I am doing what I
> think I am doing when I am running my models.
>
> myBiomodData <- BIOMOD_FormatingData(resp.var=myResp, resp.xy=myRespXY,
> expl.var=myExpl, resp.name=myRespName, eval.resp.var=myEvalpa,
> eval.expl.var=myEvalenv, eval.resp.xy=myRespEvalXY)
>
> myBiomodModelOut <- BIOMOD_Modeling(myBiomodData, models = c('GLM', 'GBM',
> 'CTA', 'RF', 'MARS'), models.options=NULL, NbRunEval=10, DataSplit=75,
> Yweights=NULL, Prevalence=NULL, VarImport=3, models.eval.meth =
> c('KAPPA','TSS','ROC'),SaveObj = TRUE, rescal.all.models = TRUE,
> do.full.models = TRUE, modeling.id = ?test_date? )
> Thanks!
> Kristen
From postmaster at r-forge.wu-wien.ac.at Thu May 30 10:42:03 2013
From: postmaster at r-forge.wu-wien.ac.at (Bounced mail)
Date: Thu, 30 May 2013 15:42:03 +0700
Subject: [Biomod-commits] ERROR
Message-ID:
The original message was received at Thu, 30 May 2013 15:42:03 +0700 from 102.144.44.52
----- The following addresses had permanent fatal errors -----
biomod-commits at r-forge.wu-wien.ac.at
From pep.bioalerts at gmail.com Thu May 30 12:06:31 2013
From: pep.bioalerts at gmail.com (Josep M Serra diaz)
Date: Thu, 30 May 2013 12:06:31 +0200
Subject: [Biomod-commits] MaxLike
In-Reply-To:
References:
Message-ID:
Hey Jesus! I think it is a great idea...but great concern should be used
when applying Royle 2012 approach... I am not for or against...but the
debate is open :)
See
http://onlinelibrary.wiley.com/doi/10.1111/j.1600-0587.2013.00321.x/abstract
Pep
2013/5/29 Jesus
> Hi Damien, I was wondering if you have thought about implementing MaxLike
> (Royle et al. 2012, MEE) in the BIOMOD package? have you considered it?
>
> Cheers,
>
> Jes?s
> _______________________________________________
> 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 Fri May 31 14:00:15 2013
From: postmaster at r-forge.wu-wien.ac.at (Bounced mail)
Date: Fri, 31 May 2013 19:00:15 +0700
Subject: [Biomod-commits] Mail System Error - Returned Mail
Message-ID:
The original message was received at Fri, 31 May 2013 19:00:15 +0700
from r-forge.wu-wien.ac.at [140.22.201.109]
----- The following addresses had permanent fatal errors -----