ecograph/clusterapply

Apply a function in parallel on data sets that are split by some factor variable

clusterapply

Clusterapply is meant to break datasets into smaller frames on which functions (regressions, for example) can be applied. Much of this can already be accomplished with the family of *apply functions, but here we are building towards parallelization with an eye on memory use. The package does not yet contain all intended features and should be considered provisional.

This package is used primarily in the EPIDEMIA forecasting system’s epidemiar package: https://github.com/EcoGRAPH/epidemiar/releases/latest with an associated demonstration project available at: https://github.com/EcoGRAPH/epidemiar-demo/releases/latest

Installation

The development version, whenever public, is available from GitHub with:

# install.packages("devtools")
devtools::install_github("EcoGRAPH/clusterapply")

Example

Please see the associated ca-vignette.Rmd file for an example of clusterapply’s use.