Range bagging SDM predict method — predict.Rangebag • bssdm

The model prediction component of an implementation of the range bagging species distribution modelling (SDM) method (Drake, 2015).

Usage

# S3 method for class 'Rangebag'
predict(
  object,
  x,
  raw_output = NULL,
  filename = "",
  parallel_cores = NULL,
  debug = FALSE,
  ...
)

Arguments

object

A "Rangebag" model S4 object containing slots:

method: SDM method: "rangebag".
variables: List of climate (or environmental) variable names.
presence: The selected climate data corresponding to occurrences points.
coordinates: The coordinates for the selected climate data.
ch_models: A list of convex hull models (vertices).

x

Climate (or environmental) data with corresponding model variables as a terra::SpatRaster, raster::Raster*, data.frame, or matrix.

raw_output

Logical to indicate whether to return raw predicted values (TRUE) or as an object (as per x: FALSE). Default is NULL, returning either raw values or a spatial raster (as per x).

filename

Optional filename for writing spatial raster output (only). Default is "".

parallel_cores

Optional number of cores available for parallel processing, thus enable parallel processing. Default is NULL (serial).

debug

Output additional debug information (memory block info, etc.). Default is FALSE.

...

Additional parameters.

Value

Predicted values as a raw vector or a terra::SpatRaster, raster::Raster*, data.frame, or matrix (as per x).

References

Drake, J. M. (2015). Range bagging: a new method for ecological niche modelling from presence-only data. Journal of the Royal Society Interface, 12(107), 20150086. doi:10.1098/rsif.2015.0086