Regression Tree Class¶

This is the TPZ.Rtree class in some detail, refer to the source code for mode information and methods.

Module author: Matias Carrasco Kind

class TPZ.Rtree(X, Y, minleaf=4, forest='yes', mstar=2, dict_dim='')¶

Creates a regression tree class instance

Parameters:	X (float or int array, 1 row per object) – Preprocessed attributes array (all columns are considered) Y (float) – Attribute to be predicted minleaf (int, def = 4) – Minimum number of objects on terminal leaf forest (str, ‘yes’/’no’) – Random forest key mstar (int) – Number of random subsample of attributes if forest is used dict_dim (dict) – dictionary with attributes names

get_branch(line)¶

Get the branch in string format given a line search, where the line is a vector of attributes per individual object

Parameters:	line (float) – input data line to look in the tree, same dimensions as input X
Returns:	str – branch array in string format, ex., [‘L’,’L’,’R’]

get_vals(line)¶

Get the predictions given a line search, where the line is a vector of attributes per individual object

Parameters:	line (float) – input data line to look in the tree, same dimensions as input X
Returns:	float – array with the leaf content

leaves()¶

Return an array with all branches in string format ex: [‘L’,’R’,’L’] is a branch of depth 3 where L and R are the left or right branches

Returns:	str – Array of all branches in the tree

leaves_dim()¶

Returns an array of the used dimensions for all the the nodes on all the branches

Returns:	int – Array of all the dimensions for each node on each branch

plot_tree(itn=-1, fileout='TPZ', path='', save_png='no')¶

Plot a tree using dot (Graphviz) Saves it into a png file by default

Parameters:	itn (int) – Number of tree to be included on path, use -1 to ignore this number fileout (str) – Name of file for the png files path (str) – path for the output files save_png (str) – save png created by Graphviz (‘yes’/’no’)

print_branch(branch)¶

Returns the content of a leaf on a branch (given in string format)

save_tree(itn=-1, fileout='TPZ', path='')¶

Saves the tree

Parameters:	itn (int) – Number of tree to be included on path, use -1 to ignore this number fileout (str) – Name of output file path (str) – path for the output file