splitTree: splitTree

In IntTreeR/IntTreeR: Build classification trees in an interactive way

Description

Performs a new binary split at a selected splitpoint.

Usage

splitTree(tree.obj, split.point, split.feature, max.surrogates, custom.split)

Arguments

tree.obj	The tree object, on which the split will be performed with the given split point.
split.point	(integer) non-negative numeric; Node ID of terminal node where you want to perform your split.
split.featurecharacter	or string; Column name of the attribute which should be used for the split.
max.surrogates(Optional)	(Integer) non-negative; the maximum number of surrogates, which will be calculated for the split. (Default is 3)
custom.split(Optional)	(Integer) custom point to split (only for numeric attributes)

Details

Missing values are handled via Surrogate Splits. In most details, it follows „Leo Breiman et. al. (1984). “quite closely.

Examples

tree <- splitTree(tree.obj = tree, split.point = 1, split.feature = “income”)

IntTreeR/IntTreeR documentation built on May 7, 2019, 6:38 a.m.