Skip to content

Projecting model to landscape

7 messages · AdrianR, Tim Meehan, Nicole K.S. Barker +2 more

#
Hi, Li. 

I am far from certain, but I believe this might be a good place to look:
http://cran.r-project.org/web/packages/spatstat/index.html
http://www.spatstat.org/

Regards, 
Adrian Rasmussen


Li Wen-2 wrote

            

        
        
--
View this message in context: http://r-sig-ecology.471788.n2.nabble.com/Projecting-model-to-landscape-tp7577935p7577936.html
Sent from the r-sig-ecology mailing list archive at Nabble.com.
#
Thanks, Nicole

I want to compare a few methods as in Elith et al 2006 Ecography 29: 129-51. So, what I am after is a generic method to do "spatial projection" with fitted statistical models, could be GAM, GLM, or MARS, etc. Maybe using prediction.model is the only way to this at the moment. 

Cheers,
Li

-----Original Message-----
From: r-sig-ecology-bounces at r-project.org [mailto:r-sig-ecology-bounces at r-project.org] On Behalf Of Nicole K.S. Barker
Sent: Wednesday, 6 March 2013 5:00 AM
To: Li Wen
Cc: r-sig-ecology at r-project.org
Subject: Re: [R-sig-eco] Projecting model to landscape

If you extract the information from your rasters into a dataframe (one
column per raster), can you use the predict.randomForest function in the
randomForest package?

If yes, you can then input your predictions back into a raster of the same
dimensions as the environmental variables.

I don't have any experience with random forest model, but that's what I do
with my Boosted Regression Tree predictions.

Nicole K.S. Barker
Ph.D Student / ?tudiante au Doctorat
Laval University / Universit? Laval
Ducks Unlimited Canada / Canards Illimit?s Canada
LinkedIn: www.linkedin.com/in/nksbarker
Phone: 647-470-3207
On Tue, Mar 5, 2013 at 5:21 AM, Li Wen <li01.wen at gmail.com> wrote:

            
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
This email is intended for the addressee(s) named and may contain confidential and/or privileged information. 
If you are not the intended recipient, please notify the sender and then delete it immediately.
Any views expressed in this email are those of the individual sender except where the sender expressly and with authority states them to be the views of the Office of Environment and Heritage, NSW Department of Premier and Cabinet.

PLEASE CONSIDER THE ENVIRONMENT BEFORE PRINTING THIS EMAIL
#
You may want to consider the  Biomod
<http://www.will.chez-alice.fr/Software.html>   package ( Thuiller 2009
<http://www.will.chez-alice.fr/pdf/ThuillerEcography2009.PDF>  ).

This sounds like what you are looking for, assuming you are working with
presence/absence data.   Biomod does not accommodate abundance data.  It can
generate pseudo-absences if you have presence-only data.

Biomod is an ensemble forecasting package that can output multiple kinds of
models including random forests, ANN, GLM, GAM, classification trees, etc.
It will make predictions for all grid cells in your raster layers based on
any single model or using an ensemble consensus approach.    It provides
parameters that can be used to compare models and assess variable
importance.

Hope this helps. Good luck!



--
View this message in context: http://r-sig-ecology.471788.n2.nabble.com/Projecting-model-to-landscape-tp7577935p7577941.html
Sent from the r-sig-ecology mailing list archive at Nabble.com.
#
Thanks, Doug

It seems that Biomod is the one I am after. Thanks again.

Li

-----Original Message-----
From: r-sig-ecology-bounces at r-project.org [mailto:r-sig-ecology-bounces at r-project.org] On Behalf Of doug.leasure
Sent: Wednesday, 6 March 2013 9:30 AM
To: r-sig-ecology at r-project.org
Subject: Re: [R-sig-eco] Projecting model to landscape

You may want to consider the  Biomod
<http://www.will.chez-alice.fr/Software.html>   package ( Thuiller 2009
<http://www.will.chez-alice.fr/pdf/ThuillerEcography2009.PDF>  ).

This sounds like what you are looking for, assuming you are working with
presence/absence data.   Biomod does not accommodate abundance data.  It can
generate pseudo-absences if you have presence-only data.

Biomod is an ensemble forecasting package that can output multiple kinds of
models including random forests, ANN, GLM, GAM, classification trees, etc.
It will make predictions for all grid cells in your raster layers based on
any single model or using an ensemble consensus approach.    It provides
parameters that can be used to compare models and assess variable
importance.

Hope this helps. Good luck!



--
View this message in context: http://r-sig-ecology.471788.n2.nabble.com/Projecting-model-to-landscape-tp7577935p7577941.html
Sent from the r-sig-ecology mailing list archive at Nabble.com.

_______________________________________________
R-sig-ecology mailing list
R-sig-ecology at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
----------------------------------------------------------------------------------------------------------------------------------------------------------------------
This email is intended for the addressee(s) named and may contain confidential and/or privileged information. 
If you are not the intended recipient, please notify the sender and then delete it immediately.
Any views expressed in this email are those of the individual sender except where the sender expressly and with authority states them to be the views of the Office of Environment and Heritage, NSW Department of Premier and Cabinet.

PLEASE CONSIDER THE ENVIRONMENT BEFORE PRINTING THIS EMAIL