hypervolumesToFile: Identify hypervolumes over multiple files

In eliotmiller/ebirdr: Manipulate eBird data

Description

Given a set of eBird files and specified traits, identify hypervolumes and save to disk.

Usage

hypervolumesToFile(files, hv.type, bandwidth.arg, r.points, species.col,
  trait.cols, read.wd, write.wd, cores, ...)

Arguments

files	Character vector specifying which files from the read.wd to process. If missing, will assume all files in read.wd should be included. Full file names (without path) are required.
hv.type	Options are "box", "gaussian", and "svm".
bandwidth.arg	Options are either "silverman", "cross-validation", or a fixed number, per standard hypervolume calculations. See Blonder et al. 2014 for more details, including his FAQ page.
r.points	The repsperpoint argument from hypervolume. If missing, will be set to NULL and will default to hypervolume's default, ceiling((10^(3 + sqrt(ncol(data))))/nrow(data)).
species.col	The name of the column identifying the species to which the traits belong. Not needed per se, but the function currently takes whatever column the user provides here and binds the traits.col to it, ignoring this first column for the hypervolume calculations. In other words, it is expected that you provide the name of some column here.
trait.cols	A character vector detailing which columns from the eBird files you would like to calculate a hypervolume for.
read.wd	The path to the directory where the eBird records are housed.
write.wd	The path to the directory where the hypervolumes will be saved.
cores	The number of cores to use for parallel processing.
...	Can be used to pass things to the hypervolume function, e.g., svm.nu for the svm method, quantile.requested for gaussian, etc.

Details

Interesting and seemingly very useful method from Blonder et al. 2014, although not entirely clear how best to set the bandwidth.arg parameters.

Value

Prints progress to the workspace, but saves nothing there. Results written as RDS files to the write.wd.

References

Blonder, B., C. Lamanna, C. Violle, and B. J. Enquist. 2014. The n-dimensional hypervolume. Global Ecology and Biogeography 23:595-609.

eliotmiller/ebirdr documentation built on May 14, 2019, 10:33 a.m.