Prepare_covarStack: Merge covariate rasters from different extents and...

In statnmap/SDMSelect: Cross-validation model selection and species distribution mapping

Description

Merge covariate rasters from different extents and projections into a multi-layer Raster

Usage

Prepare_covarStack(
  cov.paths,
  r.ref,
  ext,
  outWD,
  names.cov,
  is.factor,
  verbose = 1,
  cl = NULL
)

Arguments

cov.paths	vector of paths of all raster files to be added in the multi-layer raster
r.ref	Raster used as a reference for extent, resolution, crs. Default is the first one of the list.
ext	Optional object of class extent to crop r.ref before stack.
outWD	Directory where to save new reprojected, cropped rasters. Default to "tempdir()/extent".
names.cov	Optional. vector of names of multi-layer raster. Same length as x.
is.factor	position of covariates that are factors or categories. Resampling (if required) is bilinear by default but "ngb" (nearest neighbour) for factors.
verbose	level of messages shown
cl	a cluster as made with makeCluster. If cl is empty, nbclust in modelselect_opt will be used to create a cluster.

Details

Rasters are merged in a unique RasterStack. If the projection of a raster is different than the reference (r.ref), it is reprojected. If the extent is different, it is resampled. If the resolution is smaller than 0.9*res, it is first aggregated, then resampled. Method depends if projection is really different, otherwise this is only a change in name of proj

statnmap/SDMSelect documentation built on April 1, 2021, 2:01 p.m.