SDMSelect: A R-package for cross-validation model selection and species distribution mapping

A R-package for cross-validation model selection and species distribution mapping

• Added pkgdown page on github
• Allow for factors in presence-absence models (transformed as 0/1)
• Correct options with version 0.1.0 of {GlobalOptions}
• fixed NaN could be returned when comparing distribution with itself in best_distri.
• Reduce number of outputs saved at each iteration to only the best ones. With a lot of covariates tested, saving all model tested was too huge. Too keep more models, modelselect_opt$lim_pvalue can be lowered.
• Save outputs of model selection as rds, not RData.
• Added gplot_data, a modified version of rasterVis::gplot to retrieve data of raster for visualisation instead of directly plotting it in ggplot.