Soil as predictor in SDM

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Soil as predictor in SDM

maurizio
Dear R users,
I would like to include soil as predictor in my model and in addition to
climatic variables. Anyway I have some doubts concerning the usd of
categorical values such as the format of my raster file. Actually, my soil
type data are categorical codes, should these codes be described as
categories prior to be used? Or shoul I reclassify my raster into numeric
values? My main concern is if I would create directional influence and bias
in the results.
Thanks
--
Maurizio Marchi
Skype ID: maurizioxyz
linux user 552742

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Soil as predictor in SDM

Patrick Schratz
Hi Maurizio,

I do not really understand what exactly your problem is.
Getting the soil information stored in a raster file into a data.frame?
General use of factor variables in a model?
I also do not see the link between a factor variable and model bias.

Happy to help after clarification.

Regards, Patrick

On 09/27/2017 04:59 PM, Maurizio Marchi wrote:
> Dear R users,
> I would like to include soil as predictor in my model and in addition to
> climatic variables. Anyway I have some doubts concerning the usd of
> categorical values such as the format of my raster file. Actually, my soil
> type data are categorical codes, should these codes be described as
> categories prior to be used? Or shoul I reclassify my raster into numeric
> values? My main concern is if I would create directional influence and bias
> in the results.
> Thanks

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Soil as predictor in SDM

Narayani Barve-2
In reply to this post by maurizio
You can use categorical variable as categorical even with number. If you
are using Maxent, then you can specify that this variable as categorical.
And maxent will take care of it. On the safer side instead of specifying
the categories as numbers, change categories to characters.

Regards,
Narayani

On Wed, Sep 27, 2017 at 10:59 AM, Maurizio Marchi <
[hidden email]> wrote:

> Dear R users,
> I would like to include soil as predictor in my model and in addition to
> climatic variables. Anyway I have some doubts concerning the usd of
> categorical values such as the format of my raster file. Actually, my soil
> type data are categorical codes, should these codes be described as
> categories prior to be used? Or shoul I reclassify my raster into numeric
> values? My main concern is if I would create directional influence and bias
> in the results.
> Thanks
> --
> Maurizio Marchi
> Skype ID: maurizioxyz
> linux user 552742
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> [hidden email]
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Soil as predictor in SDM

Ben Tupper
In reply to this post by maurizio
Hi,

This is addressed briefly in section 4.2 of the dismo package vignette.

https://cran.r-project.org/web/packages/dismo/vignettes/sdm.pdf <https://cran.r-project.org/web/packages/dismo/vignettes/sdm.pdf>

Does that answer your question?

Ben

> On Sep 27, 2017, at 10:59 AM, Maurizio Marchi <[hidden email]> wrote:
>
> Dear R users,
> I would like to include soil as predictor in my model and in addition to
> climatic variables. Anyway I have some doubts concerning the usd of
> categorical values such as the format of my raster file. Actually, my soil
> type data are categorical codes, should these codes be described as
> categories prior to be used? Or shoul I reclassify my raster into numeric
> values? My main concern is if I would create directional influence and bias
> in the results.
> Thanks
> --
> Maurizio Marchi
> Skype ID: maurizioxyz
> linux user 552742
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> [hidden email]
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo

Ben Tupper
Bigelow Laboratory for Ocean Sciences
60 Bigelow Drive, P.O. Box 380
East Boothbay, Maine 04544
http://www.bigelow.org

Ecocast Reports: http://seascapemodeling.org/ecocast.html
Tick Reports: https://report.bigelow.org/tick/
Jellyfish Reports: https://jellyfish.bigelow.org/jellyfish/




        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Soil as predictor in SDM

maurizio
In reply to this post by maurizio
Dear Narayani and Ben,
thank you very much for your replies. Actually I'm working with biomod2,
not dismo neither other models outside biomod2 (i.e. I use maxent but
inside the biomod2 package) but I did it, I saw that biomod2 is able to
handle categorical (factorial) raster too :)
Thanks

--
Maurizio Marchi
Skype ID: maurizioxyz
linux user 552742

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo