bouldr README
The bouldr package is a toolkit for running Receiver Operator Characteristic (ROC) Curve analyses using a simple, formula interface. It also allows for intuitive visualization and statistical comparison of the curves. The statistical core of the package is pROC, a package developed by Xavier Robin and colleagues (see citation below).
Many ROC packages focus on machine learning and classification use-cases. However, ROC is useful in psychology as well -- particularly in developing and evaluating assessment instruments. A continuous score on a test can be ROC-ed against a target diagnosis to test the diagnostic efficiency of the test. This is similar to diagnostic accuracy, but ROC is the preferred approach because it assesses sensativity and specificity across the range of the scale, and can be used to identify optimal cut-scores. This package relies on pROC to perform the computation.
Requirements
Rversion 3.5.0 or greatertidyrdplyrmagrittrggplot2pROCRcppAlgosbroomstats
Installation
bouldr is not currently on CRAN, so you'll need devtools to install:
devtools::install_github('jlangaa/bouldr')
Examples
Use ?bouldr and ?generate_data to see examples.
References
Xavier Robin, Natacha Turck, Alexandre Hainard, Natalia Tiberti, Frédérique Lisacek, Jean-Charles Sanchez and Markus Müller (2011). pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics, 12, p. 77. DOI:10.1186/1471-2105-12-77 http://www.biomedcentral.com/1471-2105/12/77/