Calculates first, and second order, average minimal depth of a maximal subtree for a forest object produced by the R 'ranger' package. This variable importance metric is implemented as described in Ishwaran et. al. ("High-Dimensional Variable Selection for Survival Data", March 2010, <doi:10.1198/jasa.2009.tm08622>).
|Author||John Karlen [aut, cre]|
|Date of publication||2016-11-19 10:13:21|
|Maintainer||John Karlen <email@example.com>|
|License||MIT + file LICENSE|
baseRPlotting: Plot Second Order vs First Order minimal depth of a maximal...
binForestByDepth: Bin Forest by Depth
calculateAMDMS: Forest Averaged Minimal Depth of a Maximal Subtree (AMDMS)
countSplitsPerVar: Count Splits Per Variable
getAndPlotMetric: Calculate the metric and make Second Order vs First Order...
lookForVarsAbsentInForest: Look for Variable ID's that didn't occur in the Forest.
plotAMDMSvsNumSplits: Plot metric vs number of splits per variable.
plotFirstAndSecondOrderMetric: Plot Second Order vs First Order, minimal depth of a maximal...
recursiveDepthBinning: Recursive Depth Binning
startRecursiveDepthBinning: Start Recursive Depth Binning