Statistics Toolbox™

cutcategories - Cut categories

Class

@classregtree

Syntax

C = cutcategories(t) C = cutcategories(t,nodes)

Description

C = cutcategories(t) returns an n-by-2 cell array C of the categories used at branches in the decision tree t, where n is the number of nodes. For each branch node i based on a categorical predictor variable x, the left child is chosen if x is among the categories listed in C{i,1}, and the right child is chosen if x is among those listed in C{i,2}. Both columns of C are empty for branch nodes based on continuous predictors and for leaf nodes.

C = cutcategories(t,nodes) takes a vector nodes of node numbers and returns the categories for the specified nodes.

Example

Create a classification tree for car data:

load carsmall

t = classregtree([MPG Cylinders],Origin,...
                 'names',{'MPG' 'Cyl'},'cat',2)
t = 
Decision tree for classification
 1  if Cyl=4 then node 2 else node 3
 2  if MPG<31.5 then node 4 else node 5
 3  if Cyl=6 then node 6 else node 7
 4  if MPG<21.5 then node 8 else node 9
 5  if MPG<41 then node 10 else node 11
 6  if MPG<17 then node 12 else node 13
 7  class = USA
 8  class = France
 9  class = USA
10  class = Japan
11  class = Germany
12  class = Germany
13  class = USA

view(t)

C = cutcategories(t)
C = 
    [4]    [1x2 double]
     []              []
    [6]    [         8]
     []              []
     []              []
     []              []
     []              []
     []              []
     []              []
     []              []
     []              []
     []              []
     []              []
C{1,2}
ans =
     6     8

Reference