weka.classifiers.trees

Class RandomTree

java.lang.Object

weka.classifiers.Classifier

weka.classifiers.trees.RandomTree

All Implemented Interfaces:

Serializable, Cloneable, CapabilitiesHandler, Drawable, OptionHandler, Randomizable, RevisionHandler, WeightedInstancesHandler

public class RandomTree
extends Classifier
implements OptionHandler, WeightedInstancesHandler, Randomizable, Drawable

Class for constructing a tree that considers K randomly chosen attributes at each node. Performs no pruning. Also has an option to allow estimation of class probabilities based on a hold-out set (backfitting).

Valid options are:

 -K <number of attributes>
  Number of attributes to randomly investigate
  (<0 = int(log_2(#attributes)+1)).

 -M <minimum number of instances>
  Set minimum number of instances per leaf.

 -S <num>
  Seed for random number generator.
  (default 1)

 -depth <num>
  The maximum depth of the tree, 0 for unlimited.
  (default 0)

 -N <num>
  Number of folds for backfitting (default 0, no backfitting).

 -U
  Allow unclassified instances.

 -D
  If set, classifier is run in debug mode and
  may output additional info to the console

Version:

$Revision: 5535 $

Author:

Eibe Frank (eibe@cs.waikato.ac.nz), Richard Kirkby (rkirkby@cs.waikato.ac.nz)

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	m_AllowUnclassifiedInstances Whether unclassified instances are allowed
protected int	m_Attribute The attribute to split on.
protected double[]	m_ClassDistribution Class probabilities from the training data.
protected Instances	m_Info The header information.
protected int	m_KValue The number of attributes considered for a split.
protected int	m_MaxDepth The maximum depth of the tree (0 = unlimited)
protected double	m_MinNum Minimum number of instances for leaf.
protected int	m_NumFolds Determines how much data is used for backfitting
protected double[]	m_Prop The proportions of training instances going down each branch.
protected int	m_randomSeed The random seed to use.
protected double	m_SplitPoint The split point.
protected RandomTree[]	m_Successors The subtrees appended to this tree.
protected Classifier	m_ZeroR a ZeroR model in case no model can be built from the data

Fields inherited from class weka.classifiers.Classifier

m_Debug

Fields inherited from interface weka.core.Drawable

BayesNet, Newick, NOT_DRAWABLE, TREE

Constructor Summary

Constructors

Constructor and Description
RandomTree()

Method Summary

Modifier and Type	Method and Description
String	allowUnclassifiedInstancesTipText() Returns the tip text for this property
void	backfitData(Instances data) Backfits the given data into the tree.
protected void	backfitData(Instances data, double[] classProbs) Recursively backfits data into the tree.
void	buildClassifier(Instances data) Builds classifier.
protected void	buildTree(Instances data, double[] classProbs, Instances header, double minNum, boolean debug, int[] attIndicesWindow, Random random, int depth, boolean allow) Recursively generates a tree.
protected double	distribution(double[][] props, double[][][] dists, int att, Instances data) Computes class distribution for an attribute.
double[]	distributionForInstance(Instance instance) Computes class distribution of an instance using the decision tree.
protected double	gain(double[][] dist, double priorVal) Computes value of splitting criterion after split.
boolean	getAllowUnclassifiedInstances() Get the value of NumFolds.
Capabilities	getCapabilities() Returns default capabilities of the classifier.
int	getKValue() Get the value of K.
int	getMaxDepth() Get the maximum depth of trh tree, 0 for unlimited.
double	getMinNum() Get the value of MinNum.
int	getNumFolds() Get the value of NumFolds.
String[]	getOptions() Gets options from this classifier.
String	getRevision() Returns the revision string.
int	getSeed() Gets the seed for the random number generations
String	globalInfo() Returns a string describing classifier
String	graph() Returns graph describing the tree.
int	graphType() Returns the type of graph this classifier represents.
String	KValueTipText() Returns the tip text for this property
protected String	leafString() Outputs a leaf.
Enumeration	listOptions() Lists the command-line options for this classifier.
static void	main(String[] argv) Main method for this class.
String	maxDepthTipText() Returns the tip text for this property
String	minNumTipText() Returns the tip text for this property
String	numFoldsTipText() Returns the tip text for this property
int	numNodes() Computes size of the tree.
protected double	priorVal(double[][] dist) Computes value of splitting criterion before split.
String	seedTipText() Returns the tip text for this property
void	setAllowUnclassifiedInstances(boolean newAllowUnclassifiedInstances) Set the value of AllowUnclassifiedInstances.
void	setKValue(int k) Set the value of K.
void	setMaxDepth(int value) Set the maximum depth of the tree, 0 for unlimited.
void	setMinNum(double newMinNum) Set the value of MinNum.
void	setNumFolds(int newNumFolds) Set the value of NumFolds.
void	setOptions(String[] options) Parses a given list of options.
void	setSeed(int seed) Set the seed for random number generation.
protected Instances[]	splitData(Instances data) Splits instances into subsets based on the given split.
String	toGraph() Outputs the decision tree as a graph
int	toGraph(StringBuffer text, int num) Outputs one node for graph.
protected int	toGraph(StringBuffer text, int num, RandomTree parent) Outputs one node for graph.
String	toString() Outputs the decision tree.
protected String	toString(int level) Recursively outputs the tree.

Methods inherited from class weka.classifiers.Classifier

classifyInstance, debugTipText, forName, getDebug, makeCopies, makeCopy, runClassifier, setDebug

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

m_Successors

protected RandomTree[] m_Successors

The subtrees appended to this tree.

m_Attribute

protected int m_Attribute

The attribute to split on.

m_SplitPoint

protected double m_SplitPoint

The split point.

m_Info

protected Instances m_Info

The header information.

m_Prop

protected double[] m_Prop

The proportions of training instances going down each branch.

m_ClassDistribution

protected double[] m_ClassDistribution

Class probabilities from the training data.

m_MinNum

protected double m_MinNum

Minimum number of instances for leaf.

m_KValue

protected int m_KValue

The number of attributes considered for a split.

m_randomSeed

protected int m_randomSeed

The random seed to use.

m_MaxDepth

protected int m_MaxDepth

The maximum depth of the tree (0 = unlimited)

m_NumFolds

protected int m_NumFolds

Determines how much data is used for backfitting

m_AllowUnclassifiedInstances

protected boolean m_AllowUnclassifiedInstances

Whether unclassified instances are allowed

m_ZeroR

protected Classifier m_ZeroR

a ZeroR model in case no model can be built from the data

Constructor Detail

RandomTree

public RandomTree()

Method Detail

globalInfo

public String globalInfo()

Returns a string describing classifier

Returns:

a description suitable for displaying in the explorer/experimenter gui

minNumTipText

public String minNumTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getMinNum

public double getMinNum()

Get the value of MinNum.

Returns:

Value of MinNum.

setMinNum

public void setMinNum(double newMinNum)

Set the value of MinNum.

Parameters:

newMinNum - Value to assign to MinNum.

KValueTipText

public String KValueTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getKValue

public int getKValue()

Get the value of K.

Returns:

Value of K.

setKValue

public void setKValue(int k)

Set the value of K.

Parameters:

k - Value to assign to K.

seedTipText

public String seedTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

setSeed

public void setSeed(int seed)

Set the seed for random number generation.

Specified by:

setSeed in interface Randomizable

Parameters:

seed - the seed

getSeed

public int getSeed()

Gets the seed for the random number generations

Specified by:

getSeed in interface Randomizable

Returns:

the seed for the random number generation

maxDepthTipText

public String maxDepthTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getMaxDepth

public int getMaxDepth()

Get the maximum depth of trh tree, 0 for unlimited.

Returns:

the maximum depth.

numFoldsTipText

public String numFoldsTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getNumFolds

public int getNumFolds()

Get the value of NumFolds.

Returns:

Value of NumFolds.

setNumFolds

public void setNumFolds(int newNumFolds)

Set the value of NumFolds.

Parameters:

newNumFolds - Value to assign to NumFolds.

allowUnclassifiedInstancesTipText

public String allowUnclassifiedInstancesTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getAllowUnclassifiedInstances

public boolean getAllowUnclassifiedInstances()

Get the value of NumFolds.

Returns:

Value of NumFolds.

setAllowUnclassifiedInstances

public void setAllowUnclassifiedInstances(boolean newAllowUnclassifiedInstances)

Set the value of AllowUnclassifiedInstances.

Parameters:

newAllowUnclassifiedInstances - Value to assign to AllowUnclassifiedInstances.

setMaxDepth

public void setMaxDepth(int value)

Set the maximum depth of the tree, 0 for unlimited.

Parameters:

value - the maximum depth.

listOptions

public Enumeration listOptions()

Lists the command-line options for this classifier.

Specified by:

listOptions in interface OptionHandler

Overrides:

listOptions in class Classifier

Returns:

an enumeration over all possible options

getOptions

public String[] getOptions()

Gets options from this classifier.

Specified by:

getOptions in interface OptionHandler

Overrides:

getOptions in class Classifier

Returns:

the options for the current setup

setOptions

public void setOptions(String[] options)
                throws Exception

Parses a given list of options.

Valid options are:

 -K <number of attributes>
  Number of attributes to randomly investigate
  (<0 = int(log_2(#attributes)+1)).

 -M <minimum number of instances>
  Set minimum number of instances per leaf.

 -S <num>
  Seed for random number generator.
  (default 1)

 -depth <num>
  The maximum depth of the tree, 0 for unlimited.
  (default 0)

 -N <num>
  Number of folds for backfitting (default 0, no backfitting).

 -U
  Allow unclassified instances.

 -D
  If set, classifier is run in debug mode and
  may output additional info to the console

Specified by:

setOptions in interface OptionHandler

Overrides:

setOptions in class Classifier

Parameters:

options - the list of options as an array of strings

Throws:

Exception - if an option is not supported

getCapabilities

public Capabilities getCapabilities()

Returns default capabilities of the classifier.

Specified by:

getCapabilities in interface CapabilitiesHandler

Overrides:

getCapabilities in class Classifier

Returns:

the capabilities of this classifier

See Also:

Capabilities

buildClassifier

public void buildClassifier(Instances data)
                     throws Exception

Builds classifier.

Specified by:

buildClassifier in class Classifier

Parameters:

data - the data to train with

Throws:

Exception - if something goes wrong or the data doesn't fit

backfitData

public void backfitData(Instances data)
                 throws Exception

Backfits the given data into the tree.

Throws:

Exception

distributionForInstance

public double[] distributionForInstance(Instance instance)
                                 throws Exception

Computes class distribution of an instance using the decision tree.

Overrides:

distributionForInstance in class Classifier

Parameters:

instance - the instance to compute the distribution for

Returns:

the computed class distribution

Throws:

Exception - if computation fails

toGraph

public String toGraph()

Outputs the decision tree as a graph

Returns:

the tree as a graph

toGraph

public int toGraph(StringBuffer text,
                   int num)
            throws Exception

Outputs one node for graph.

Parameters:

text - the buffer to append the output to

num - unique node id

Returns:

the next node id

Throws:

Exception - if generation fails

toString

public String toString()

Outputs the decision tree.

Overrides:

toString in class Object

Returns:

a string representation of the classifier

leafString

protected String leafString()
                     throws Exception

Outputs a leaf.

Returns:

the leaf as string

Throws:

Exception - if generation fails

toString

protected String toString(int level)

Recursively outputs the tree.

Parameters:

level - the current level of the tree

Returns:

the generated subtree

backfitData

protected void backfitData(Instances data,
                           double[] classProbs)
                    throws Exception

Recursively backfits data into the tree.

Parameters:

data - the data to work with

classProbs - the class distribution

Throws:

Exception - if generation fails

buildTree

protected void buildTree(Instances data,
                         double[] classProbs,
                         Instances header,
                         double minNum,
                         boolean debug,
                         int[] attIndicesWindow,
                         Random random,
                         int depth,
                         boolean allow)
                  throws Exception

Recursively generates a tree.

Parameters:

data - the data to work with

classProbs - the class distribution

header - the header of the data

minNum - the minimum number of instances per leaf

debug - whether debugging is on

attIndicesWindow - the attribute window to choose attributes from

random - random number generator for choosing random attributes

depth - the current depth

determineStructure - whether to determine structure

Throws:

Exception - if generation fails

numNodes

public int numNodes()

Computes size of the tree.

Returns:

the number of nodes

splitData

protected Instances[] splitData(Instances data)
                         throws Exception

Splits instances into subsets based on the given split.

Parameters:

data - the data to work with

Returns:

the subsets of instances

Throws:

Exception - if something goes wrong

distribution

protected double distribution(double[][] props,
                              double[][][] dists,
                              int att,
                              Instances data)
                       throws Exception

Computes class distribution for an attribute.

Parameters:

props -

dists -

att - the attribute index

data - the data to work with

Throws:

Exception - if something goes wrong

priorVal

protected double priorVal(double[][] dist)

Computes value of splitting criterion before split.

Parameters:

dist - the distributions

Returns:

the splitting criterion

gain

protected double gain(double[][] dist,
                      double priorVal)

Computes value of splitting criterion after split.

Parameters:

dist - the distributions

priorVal - the splitting criterion

Returns:

the gain after the split

getRevision

public String getRevision()

Returns the revision string.

Specified by:

getRevision in interface RevisionHandler

Overrides:

getRevision in class Classifier

Returns:

the revision

main

public static void main(String[] argv)

Main method for this class.

Parameters:

argv - the commandline parameters

graph

public String graph()
             throws Exception

Returns graph describing the tree.

Specified by:

graph in interface Drawable

Returns:

the graph describing the tree

Throws:

Exception - if graph can't be computed

graphType

public int graphType()

Returns the type of graph this classifier represents.

Specified by:

graphType in interface Drawable

Returns:

Drawable.TREE

toGraph

protected int toGraph(StringBuffer text,
                      int num,
                      RandomTree parent)
               throws Exception

Outputs one node for graph.

Parameters:

text - the buffer to append the output to

num - the current node id

parent - the parent of the nodes

Returns:

the next node id

Throws:

Exception - if something goes wrong