keel.Algorithms.Preprocess.NoiseFilters.IterativePartitioningFilter

Class C45

java.lang.Object

keel.Algorithms.Decision_Trees.C45.Algorithm

keel.Algorithms.Preprocess.NoiseFilters.IterativePartitioningFilter.C45

public class C45
extends Algorithm

Class to implement the C4.5 algorithm

Since:

JDK1.6

Version:

1.0

Author:

Written by Jose A. Saez Munoz, research group SCI2S (Soft Computing and Intelligent Information Systems). DECSAI (DEpartment of Computer Science and Artificial Intelligence), University of Granada - Spain. Date: 06/01/10

Field Summary

Fields inherited from class keel.Algorithms.Decision_Trees.C45.Algorithm

correct, log, modelDataset, modelFileName, resultFileName, startTime, testCorrect, testDataset, testFileName, testOutputFileName, trainDataset, trainFileName, trainOutputFileName

Constructor Summary

Constructors

Constructor and Description
C45(java.lang.String trainfn, java.lang.String testfn) Constructor.

Method Summary

Modifier and Type	Method and Description
double[]	classificationForItemset(Itemset itemset) Returns class probabilities for an itemset.
double	evaluateItemset(Itemset itemset) Function to evaluate the class which the itemset must have according to the classification of the tree.
void	generateTree(Dataset itemsets) Generates the tree.
int[]	getPredictions() Returns the predicted classes for each test instance.
static int	maxIndex(double[] doubles) Returns index of maximum element in a given array of doubles.
void	printResult() Writes the tree and the results of the training and the test in the file.
void	printTest() Evaluates the test dataset and writes the results in the file.
void	printTrain() Evaluates the training dataset and writes the results in the file.
void	priorsProbabilities() Sets the class prior probabilities.
protected void	setOptions(java.io.StreamTokenizer option) Function to read the options from the execution file and assign the values to the parameters.
protected void	setOptions(java.lang.String trainfn, java.lang.String testfn) Function to read the options from the execution file and assign the values to the parameters.
java.lang.String	toString() Function to print the tree.

Methods inherited from class keel.Algorithms.Decision_Trees.C45.Algorithm

getHeader, getNextToken, initTokenizer

Methods inherited from class java.lang.Object

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

Constructor Detail

C45

public C45(java.lang.String trainfn,
           java.lang.String testfn)
    throws java.lang.Exception

Constructor.

Parameters:

trainfn - Training dataset filename.

testfn - Test dataset filename.

Throws:

java.lang.Exception - If the algorithm cannot be executed.

Method Detail

getPredictions

public int[] getPredictions()

Returns the predicted classes for each test instance.

Returns:

the predicted classes for each test instance.

setOptions

protected void setOptions(java.io.StreamTokenizer option)
                   throws java.lang.Exception

Description copied from class: Algorithm

Function to read the options from the execution file and assign the values to the parameters.

Specified by:

setOptions in class Algorithm

Parameters:

option - The StreamTokenizer that reads the parameters file.

Throws:

java.lang.Exception - If the format of the file is not correct.

setOptions

protected void setOptions(java.lang.String trainfn,
                          java.lang.String testfn)
                   throws java.lang.Exception

Function to read the options from the execution file and assign the values to the parameters.

Parameters:

trainfn - Training dataset filename.

testfn - Test dataset filename.

Throws:

java.lang.Exception - If the format of the file is not correct.

generateTree

public void generateTree(Dataset itemsets)
                  throws java.lang.Exception

Generates the tree.

Parameters:

itemsets - The dataset used to build the tree.

Throws:

java.lang.Exception - If the tree cannot be built.

evaluateItemset

public double evaluateItemset(Itemset itemset)
                       throws java.lang.Exception

Function to evaluate the class which the itemset must have according to the classification of the tree.

Parameters:

itemset - The itemset to evaluate.

Returns:

The index of the class index predicted.

Throws:

java.lang.Exception - if the instance can not be evaluated.

classificationForItemset

public final double[] classificationForItemset(Itemset itemset)
                                        throws java.lang.Exception

Returns class probabilities for an itemset.

Parameters:

itemset - The itemset.

Returns:

class probabilities for an itemset.

Throws:

java.lang.Exception - If cannot compute the classification.

maxIndex

public static int maxIndex(double[] doubles)

Returns index of maximum element in a given array of doubles. First maximum is returned.

Parameters:

doubles - The array of elements.

Returns:

index of maximum element in a given array of doubles.

priorsProbabilities

public void priorsProbabilities()
                         throws java.lang.Exception

Sets the class prior probabilities.

Throws:

java.lang.Exception - If cannot compute the probabilities.

printResult

public void printResult()
                 throws java.io.IOException

Writes the tree and the results of the training and the test in the file.

Specified by:

printResult in class Algorithm

Throws:

java.io.IOException - If the file cannot be written.

printTrain

public void printTrain()

Evaluates the training dataset and writes the results in the file.

Specified by:

printTrain in class Algorithm

printTest

public void printTest()

Evaluates the test dataset and writes the results in the file.

Specified by:

printTest in class Algorithm

toString

public java.lang.String toString()

Function to print the tree.

Overrides:

toString in class java.lang.Object

Returns:

a String representation of the tree.