Learn R Programming

ggRandomForests (version 4.0.0)

.varpro_score_rules: Re-extract release rules and score them as importance() does

Description

The rule-sampling half of varPro::importance(). With local.std = TRUE each usable rule (out-of-bag and complement cases both present) gets its local importance and only those rules are kept; with FALSE the strength array's own importance is used, unfiltered. Returns the usable rules and a copy of object whose $results holds the scored rules, ready for importance.varpro.workhorse().

Usage

.varpro_score_rules(
  object,
  max.tree,
  max.rules.tree,
  seed = NULL,
  local.std = TRUE
)

Arguments

Value

A list: rules (the usable rows of the strength array) and scored (a copy of object holding the scored rules).