oblique.tree (version 1.1.1)

tree.impurity: Internal Function that Determines the Impurity of a Subtree

Description

Determines the impurity of a subtree.

Usage

tree.impurity(
	yprob, 
	number.of.observations.at.leaves, 
	leaf.classes, 
	impurity.measure = c("deviance","misclass"))

Arguments

yprob

A matrix of predicted class probabilities of leaves of the subtree considered in order of tree-growth (left to right).

number.of.observations.at.leaves

A vector denoting the number of observations at each of these leaves.

leaf.classes

A vector denoting the classes associated to these leaves.

impurity.measure

Character string denoting the measure of node heterogeneity to be used to guide cost-complexity pruning/trimming.

Value

• The impurity of the subtree.

Details

$$R_\alpha(T) = R(T) + \alpha size(T)$$ When pruning and trimming trees, $R(T)$ must be evaluated. Extracted arguments from oblique.tree objects is easiest.

References

Ripley, B. D. (1996). Pattern Recognition and Neural Networks. Cambridge University Press, Cambridge. Chapter 7.

See Also

oblique.tree.

Examples

#grow a tree on the Pima Indian dataset
data(Pima.tr, package = "MASS")
ob.tree <- oblique.tree(formula		= type~.,
			data		= Pima.tr,
			oblique.splits	= "on")

#deviance of entire tree
subtree.leaves <- ob.tree$frame$var == "<leaf>"
oblique.tree:::tree.impurity(	
	yprob 					= ob.tree$frame$yprob[subtree.leaves,,drop=FALSE],
	number.of.observations.at.leaves 	= ob.tree$frame$n[subtree.leaves],
	leaf.classes 				= ob.tree$frame$yval[subtree.leaves],
	impurity.measure 			= "deviance")