• Metadata Catalogue
  •   Search
  •   Map

SDMselect R package - Cross-validation model selection and species distribution mapping

A set of functions for covariate selection and model specifications with cross-validation and forward model selection. It selects GAM, GLM models using a multiple k-fold cross-validation and map species distribution while accounting for model uncertainties. Multiple families are possible and compared all together using RMSE or AUC as a result of cross-validation procedure. Covariates correlation may be tested. The model selection procedure will test different combinations of covariates with LM, GLM, GLM natural splines and GAM models, with different distributions (Gaussian, Gamma, Log-Normal, Tweedie; Binomial) and with different maximum degrees of freedom for GLM with polynoms or natural splines. Calculations are parallelized when possible. Outputs are numerous, allowing for summary of the model selection and the comparison of the different models all together. The final model selected is used to map species distribution along with maps of uncertainty.

Default

Date ( Publication)
2019-06-25
Status
Completed
Creator
  - Sébastien Rochette

Keywords

modelling

Keywords

biogeography

Keywords

GAM models

Keywords

GLM models

Keywords

cross-validation

Keywords

model selection

Keywords

species distribution

Keywords

R

Access constraints
License
Use limitation

https://cran.r-project.org/web/licenses/GPL-3

OnLine resource
Development site (

WWW:LINK-1.0-http--link

)
Required Services

R (>= 3.3.0)

Topic Category
Species distribution
Service Category

modelling

Service Category

model preparation

Service Language
eng
Service TRL
TRL 9 – Actual system proven in operational environment

Metadata

File identifier
daebb3b9-44d5-4f65-b375-2cc607406f0c XML
Metadata language
en
Hierarchy level
Service
Metadata Schema Version

0.1.6

 
 

Overviews

overview
sdmpredictors.jpg

Spatial extent

Keywords



Provided by

logo
Access to the portal
Read here the full details and access to the data.