Stacked.SDM-class {SSDM}

An S4 class to represent SSDMs

Description

This is an S4 class to represent SSDMs that assembles multiple algorithms (including generalized linear model, general additive model, multivariate adaptive splines, generalized boosted regression model, classification tree analysis, random forest, maximum entropy, artificial neural network, and support vector machines) built for multiple species. It is obtained with stack_modelling or stacking.

Slots

name

character. Name of the SSDM (by default 'Species.SSDM').

diversity.map

raster. Local species richness map produced by the SSDM.

endemism.map

raster. Endemism map produced by the SSDM (see Crisp et al (2011) in references).

uncertainty

raster. Between-algorithm variance map.

evaluation

data frame. Evaluation of the SSDM (AUC, Kappa, omission rate, sensitivity, specificity, proportion of correctly predicted occurrences).

variable.importance

data frame. Relative importance of each variable in the SSDM.

algorithm.correlation

data frame. Between-algorithm correlation matrix.

esdms

list. List of ensemble SDMs used in the SSDM.

parameters

data frame. Parameters used to build the SSDM.

algorithm.evaluation

data frame. Evaluation of the algorithm averaging the metrics of all SDMs (AUC, Kappa, omission rate, sensitivity, specificity, proportion of correctly predicted occurrences).

References

M. D. Crisp, S. Laffan, H. P. Linder & A. Monro (2001) "Endemism in the Australian flora" Journal of Biogeography 28:183-198 http://biology-assets.anu.edu.au/hosted_sites/Crisp/pdfs/Crisp2001_endemism.pdf

See Also

Ensemble.SDM an S4 class to represent ensemble SDMs, and Algorithm.SDM an S4 class to represent SDMs.