Functions provided by this pedagogic package allow to compute models with two popular machine learning approaches, BRT (Boosted Regression Trees) and MaxEnt (Maximum Entropy) applied on sets of marine biological and environmental data. They include the possibility of managing the main parameters for the construction of the models. Classic tools to evaluate model performance are provided (Area Under the Curve, omission rate and confusion matrix, map standard deviation) and are completed with tools to perform null models. The biological dataset includes original occurrences of two species of the class Echinoidea (sea urchins) present on the Kerguelen Plateau and that show contrasted ecological niches. The environmental dataset includes the corresponding statistics for 15 abiotic and biotic descriptors summarized for the Kerguelen Plateau and for different periods in a raster format. The package can be used for practicals to teach and learn the basics of species distribution modelling. Maps of potential distribution can be produced based on the example data included in the package, which brings prior observations of the influence of spatial and temporal heterogeneities on modelling performances. The user can also provide his own datasets to use the modelling functions.
|Author||Guillaumot Charlene [aut, cre], Martin Alexis [aut], Eleaume Marc [aut], Saucede Thomas [aut]|
|Date of publication||2016-08-08 19:09:02|
|Maintainer||Guillaumot Charlene <email@example.com>|
brisaster.antarcticus: Records of _Brisaster antarcticus_ echinoid presences on the...
compute.brt: Compute BRT (Boosted Regression Trees) model
compute.maxent: Compute MaxEnt model
ctenocidaris.nutrix: Records of _Ctenocidaris nutrix_ echinoid presences on the...
delim.area: RasterStack preparation for modelling
null.model: Compute null model
predictors1965_1974: Environmental descriptors for 1965-1974 on the Kerguelen...
predictors2005_2012: Environmental descriptors for 2005-2012 on the Kerguelen...
predictors2200AIB: IPCC environmental descriptors predicted for 2200 (AIB...
SDMdata.quality: Evaluate dataset quality
SDMeval: Evaluate species distribution models
SDMtab: Compile species distribution dataset for modelling