Significance Level for Random Forest Impurity Importance Scores

Sets a significance level for Random Forest MDI (Mean Decrease in Impurity, Gini or sum of squares) variable importance scores, using an empirical Bayes approach. See Dunne et al. (2022) .

RFlocalfdr

Provides a method for setting the significance level of the MDI (mean decrease in impurity) importances from a random forest model. Based on an empirical Bayes model. See https://www.biorxiv.org/content/10.1101/2022.04.06.487300v2 Thresholding Gini Variable Importance with a single trained Random Forest: An Empirical Bayes Approach (Robert Dunne, Roc Reguant, Priya Ramarao-Milne, Piotr Szul, Letitia Sng, Mischa Lundberg, Natalie A. Twine, Denis C. Bauer) for full details.

Until I figure out how to manage the cran repository:

• the data sets are not available in the cran version
• many of the examples are enclosed in "dontrun" environments

Install devtools from CRAN

install.packages("RFlocalfdr")

Or from GitHub:

devtools::install_github("parsifal9/RFlocalfdr", build_vignettes = TRUE)

Usage

library(RFlocalfdr)
vignette("simulated",package="RFlocalfdr")
vignette("Smoking",package="RFlocalfdr")

License

GNU General Public License