[R-sig-Geo] Question niche based model with R

[Maurizio Marchi]

Hi Abou,
it is a quite broad question...a simple and easy answer would be "you
can just model yous data.frame with ?lm function or ?glm but you would
lose many of the iteresting stuff!! Then, in my opinion, first of all
you must decide between "presence-only" and "presence-absence" models
(my suggestion is for the second and in your case you must learn how
to generate pseudo-absences or background points. So, in my
experience, there are two main packages which are the most interesting
and suited for this purpose and which may help you: dismo and biomod2
(https://cran.r-project.org/web/packages/dismo/dismo.pdf -
https://cran.r-project.org/web/packages/biomod2/biomod2.pdf). There
are also many tutorials which can help you, for instance
https://cran.r-project.org/web/packages/dismo/vignettes/sdm.pdf
Then once you will handle the subject you will also be able to create
your own SDM using the algorithm you prefer (e.g. RandomForest,
MaxEnt, GLM, MARS, GAM and so on...) and without any additional
package. Anyway, my advice is to use one of the two above cited
packages because they are object-oriented and optimized for Ecological
Modelling.

Just to cover possible next question (which model is the best
solution) you will see that it is impossible to define the BEST method
which can perform correctly in all the circumstances. Many Researchers
are used to work with a single-algorithm (for example Generalized
Linear Models, Maximum Entropy, Generalized Additive Models, Neural
network, Random Forest etc.) while others are more willing to work
with ensemble projections and consensus models (i.e. mean or weighted
mean of more than one algorithm). Some examples of single-model and
ensemble model are here:

http://dx.doi.org/10.5424/fs/2016253-09476
http://dx.doi.org/10.1111/gcb.12604
http://dx.doi.org/10.1007/s13595-014-0439-4
http://dx.doi.org/10.1111/gcb.12476

Cheers,

-- 
Maurizio Marchi
Skype ID: maurizioxyz
linux user 552742