measure_distances: Measure pair-wise distances between trees of a random forest

In imbs-hl/timbR: Tree interpretation methods based on ranger

Measure pair-wise distances between trees of a random forest

Description

measure_distances uses tree-based distance metrics to calculate standardized values for pair-wise dissimilarity of trees in a random forest trained with ranger.

Usage

measure_distances(rf, metric = "splitting variables", test_data = NULL)

Arguments

rf	Object of class ranger used with write.forest = TRUE.
metric	Specification of the tree metric. Available are "splitting variables", "weighted splitting variables", "terminal nodes" and "prediction".
test_data	Additional data set comparable to the data set rf was build on.

Value

distances

matrix of size num.treesxnum.trees

Author(s)

Lea Louisa Kronziel, M.Sc.

Examples

require(ranger)
require(timbR)

set.seed(12345)
# Train random forest with ranger
rg.iris <- ranger(Species ~ ., data = iris, write.forest=TRUE, num.trees = 10)

#
measure_distances(rf = rg.iris, metric = "splitting variables")
measure_distances(rf = rg.iris, metric = "weighted splitting variables")
measure_distances(rf = rg.iris, metric = "terminal nodes", test_data = iris)
measure_distances(rf = rg.iris, metric = "prediction", test_data = iris)

imbs-hl/timbR documentation built on April 17, 2025, 2:08 p.m.