Boruta: getImpRf – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

getImpRf

ranger Random Forest importance adapters

Description

Those function is intended to be given to a getImp argument of Boruta function to be called by the Boruta algorithm as an importance source. getImpRfZ generates default, normalized permutation importance, getImpRfRaw raw permutation importance, finally getImpRfGini generates Gini index importance.

Usage

getImpRfZ(x, y, ntree = 500, num.trees = ntree, ...)

getImpRfGini(x, y, ntree = 500, num.trees = ntree, ...)

getImpRfRaw(x, y, ntree = 500, num.trees = ntree, ...)

Arguments

`x`	data frame of predictors including shadows.
`y`	response vector.
`ntree`	Number of trees in the forest; copied into `ranger`'s native num.trees, put to retain transparent compatibility with randomForest.
`num.trees`	Number of trees in the forest, as according to `ranger`'s nomenclature. If not given, set to `ntree` value. If both are given, `num.trees` takes precedence.
`...`	parameters passed to the underlying `ranger` call; they are relayed from `...` of `Boruta`.

Note

Prior to Boruta 5.0, getImpLegacyRfZ function was a default importance source in Boruta; see getImpLegacyRf for more details.

Boruta

Wrapper Algorithm for All Relevant Feature Selection

v7.0.0

GPL (>= 2)

Authors

Miron Bartosz Kursa [aut, cre] (<https://orcid.org/0000-0001-7672-648X>), Witold Remigiusz Rudnicki [aut]

Initial release