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: cartsplitregr

Quantlet: cartsplitclass
Description: builds the classification tree.

Reference(s):

Link:
Usage: Tree = cartsplitclass(VarMatrix,ClassVector,SplitRule,MinSize)
Input:
VarMatrix n x m matrix, variables
ClassVector n x 1 vector, classes
SplitRule scalar, splitting rule either 0 (Ginni) or 1 (Twoing)
MinSize scalar, stop condition, minimum number of observations in terminal node
Output:
Tree list, composed of following elements: Variable - number of variable of the split questions SplitValue - critical value of the question Class - dominating class of the current node (class with maximum number of observations) Impurity - missclassification error for each node (between 0 and 1) NumberOfPoints - total number of points (all classes) in current node ParentNode - records of parent node Index - used for transformation from XploRe to dll and back DrawIndex - used for drawing trees in XploRe

Example:
library("xclust");
x = #(1,2,3,4,40,50,60,80,90, 100, 110, 300, 500)~#(10,0,40,60,100,1,-91,20,20,34, 1, 3, -5);
y = #(1,1,1,1,2,1,1,2,3,2, 3, 3, 3)
tr = cartsplitclass(x,y,0,1)
cartdisptree(tr)

Result:
Generates data structure tr representing a binary tree.



Author: R. Timofeev, A. Andriyashin 20041215 license MD*Tech
(C) MD*TECH Method and Data Technologies, 05.02.2006