Finding the best values for user-specified arguments of a prediction algorithm can be difficult, particularly if there is an interaction between argument levels. This package automates the testing of any user-defined prediction algorithm over an arbitrary number of arguments. It includes functions for testing the algorithm over the given arguments with respect to an arbitrary number of user-defined diagnostics, visualising the results of these tests, and finding the optimal argument combinations with respect to each diagnostic.
Version: | 0.0.1 |
Imports: | magrittr, ggplot2, reshape2, plyr, methods, stats, dplyr |
Suggests: | knitr, rmarkdown, markdown |
Published: | 2021-05-28 |
DOI: | 10.32614/CRAN.package.testarguments |
Author: | Matthew Sainsbury-Dale [aut, cre] |
Maintainer: | Matthew Sainsbury-Dale <msainsburydale at gmail.com> |
License: | MIT + file LICENSE |
NeedsCompilation: | no |
Materials: | README |
CRAN checks: | testarguments results |
Reference manual: | testarguments.pdf |
Vignettes: |
testarguments |
Package source: | testarguments_0.0.1.tar.gz |
Windows binaries: | r-devel: testarguments_0.0.1.zip, r-release: testarguments_0.0.1.zip, r-oldrel: testarguments_0.0.1.zip |
macOS binaries: | r-devel (arm64): testarguments_0.0.1.tgz, r-release (arm64): testarguments_0.0.1.tgz, r-oldrel (arm64): testarguments_0.0.1.tgz, r-devel (x86_64): testarguments_0.0.1.tgz, r-release (x86_64): testarguments_0.0.1.tgz, r-oldrel (x86_64): testarguments_0.0.1.tgz |
Please use the canonical form https://CRAN.R-project.org/package=testarguments to link to this page.