XploRe Help : cartdisptree

Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Library:	xclust
See also:	cartsplitclass cartsplitregr cartpredict cartdrawpdfclass cartdrawpdfregr

Quantlet:		cartdisptree
Description:		plots the regression/classification tree of CART in a display.

Reference(s):

Breiman, L., Friedman, J., Olshen, R., and Stone, C. J.(1984): Classification and Regression Trees. Chapman & Hall, New York

Link:

http://www.xplore-stat.de/tutorials/cartstart.html

Usage:	`cartdisptree(tr,string)`
Input:
	`tr`	list of vectors, data structure representing a binary tree, output from cartsplitregr or cartsplitclass. See cartsplitclass or cartsplitregr for detailed description.
	`string`	string, optional, either "NumberOfPoints", "Impurity" or "Class", default: see notes below

Note:

Each node on the tree displays some information. The kind of information is determined by the optional parameter string: - if string = "NumberOfPoints", each node shows its respective number of observations - if string = "Impurity", each node shows the impurity of its corresponding tree. For classification tree - sum of missclassification error of termninal nodes. For regression tree - sum of variances ratios. - if string = "Class", each node shows its respective mean of observations - as default, nonterminal nodes show the splitting rule (for example x1<=1.34) and the leaf nodes show the values of the regression function (the mean of the observations in that node)

Example:

; loads the library xclust library("xclust") ; generates an artifcal data set x1=#(0,0,0,0,1,1,1,1,1,2) x2=#(0,0,0,0,0,0,0,1,1,1) x=x1~x2 y=#(0,0,0,0,100,100,100,120,120,120) ; runs CART on our data tr=cartsplitclass(x,y,0,1) ; plots CART tree cartdisptree(tr)

Result:

shows the tree with the splitting rules

Author: R. Timofeev, A. Andriyashin 20050115 license MD*Tech