stabilityScore: Stability score

In karlkumbier/iRF2.0: Iterative Random Forests

Stability score

Description

Run outer layer bootstrap stability analysis

Usage

stabilityScore(
  x,
  y,
  ntree = 500,
  mtry = floor(sqrt(ncol(x))),
  mtry.select.prob = rep(1, ncol(x)),
  ints.idx.eval = NULL,
  ints.eval = NULL,
  rit.param = list(depth = 5, ntree = 500, nchild = 2, class.id = 1, min.nd = 1,
    class.cut = NULL),
  varnames.grp = NULL,
  n.bootstrap = 1,
  bs.sample = NULL,
  weights = rep(1, nrow(x)),
  signed = TRUE,
  oob.importance = TRUE,
  type = "randomForest",
  n.core = 1,
  ...
)

Arguments

x	numeric feature matrix
y	response vector. If factor, classification is assumed.
ntree	number of random forest trees.
mtry.select.prob	feature weights for first iteration. Defaults to equal weights
ints.idx.eval	like ints.eval, but specifies the indice of the interactions instead of their names. Intended for internal use only.
ints.eval	interactions to evaluate. If specified, importance metrics will be evaluated for these interactions instead of those recovered by RIT.
rit.param	named list specifying RIT parameters. Entries include depth: depths of RITs, ntree: number of RITs, nchild: number of child nodes for each RIT, class.id: 0-1 indicating which leaf nodes RIT should be run over, min.nd: minimum node size to run RIT over, class.cut: threshold for converting leaf nodes in regression to binary classes.
n.bootstrap	number of bootstrap samples to calculate stability scores.
bs.sample	list of observation indices to use for bootstrap samples. If NULL, iRF will take standard bootstrap samples of observations.
weights	numeric weight for each observation. Leaf nodes will be sampled for RIT with probability proprtional to the total weight of observations they contain.
signed	if TRUE, signed interactions will be returned
oob.importance	if TRUE, importance measures are evaluated on OOB samples.
n.core	number of cores to use. If -1, all available cores are used.
...	additional arguments passed to iRF::randomForest.

Value

a data table containing the recovered interactions and importance scores.

karlkumbier/iRF2.0 documentation built on April 12, 2025, 1:44 p.m.