bham.leakiest

Class ARFFFile

java.lang.Object

bham.leakiest.ARFFFile

public class ARFFFile
extends java.lang.Object

This file contains the methods needed to process ARFF files.

Version:

1.4.7

Author:

Tom Chothia, Yusuke Kawamoto, Chris Novakovic

Field Summary

Fields

Modifier and Type	Field and Description
java.lang.String[]	attributes Array of attributes
int	noOfTests The number of samples

Constructor Summary

Constructors

Constructor and Description
ARFFFile(java.lang.String fileName, int v) Opens an ARFF file for reading.
ARFFFile(java.lang.String fileName, int v, int csvInterval, CSVFile csvFile) Opens an ARFF file for reading in increments of a given number of lines.

Method Summary

Modifier and Type	Method and Description
ContinuousData	cdataFromARFF(java.util.TreeSet<java.lang.Integer> inputIndexes, java.util.TreeSet<java.lang.Integer> outputIndexes, int verbose) Returns the continuous data for the given high and low features.
static Observations	computeObservationsFromARFF(java.lang.String fileName, java.util.TreeSet<java.lang.String> highFeaturesSet, java.util.TreeSet<java.lang.String> lowFeaturesSet) Computes the observations from a given ARFF file and given features.
java.util.TreeSet<java.lang.Integer>	getFeatureIndices(java.util.TreeSet<java.lang.Integer> features, java.util.TreeSet<java.lang.String> featuresSet) Computes the a tree set of feature indices that appear in the given string set.
java.lang.String[][]	getSamples() Returns the samples array.
Observations	obsFromARFF(java.util.TreeSet<java.lang.Integer> inputIndexes, java.util.TreeSet<java.lang.Integer> outputIndexes) Returns the observation for specified high and low features.
void	printFeatures(java.util.TreeSet<java.lang.Integer> highFeatures, java.util.TreeSet<java.lang.Integer> lowFeatures, int verbose) Print the features that are selected to be investigated.
void	setTerminateWhenStabilised(boolean flag) Controls whether the obsFromARFF() method should terminate before it has finished reading all of the available samples, if the corrected leakage value stabilises.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

attributes

public java.lang.String[] attributes

Array of attributes

noOfTests

public int noOfTests

The number of samples

Constructor Detail

ARFFFile

public ARFFFile(java.lang.String fileName,
                int v)

Opens an ARFF file for reading.

Parameters:

fileName - ARFF file name

v - verbose

ARFFFile

public ARFFFile(java.lang.String fileName,
                int v,
                int csvInterval,
                CSVFile csvFile)

Opens an ARFF file for reading in increments of a given number of lines.

Parameters:

fileName - ARFF file name

v - Verbosity level

csvInterval - Number of lines to read from the file during each increment

csvFile - The CSV file to write intermediate leakage estimates to

Method Detail

cdataFromARFF

public ContinuousData cdataFromARFF(java.util.TreeSet<java.lang.Integer> inputIndexes,
                                    java.util.TreeSet<java.lang.Integer> outputIndexes,
                                    int verbose)

Returns the continuous data for the given high and low features.

Parameters:

inputIndexes - Tree set of high feature attribute indexes

outputIndexes - Tree set of low feature attribute indexes

verbose - verbose

Returns:

the continuous data for the given high and low features

computeObservationsFromARFF

public static Observations computeObservationsFromARFF(java.lang.String fileName,
                                                       java.util.TreeSet<java.lang.String> highFeaturesSet,
                                                       java.util.TreeSet<java.lang.String> lowFeaturesSet)

Computes the observations from a given ARFF file and given features.

Parameters:

fileName - name of an ARFF file

highFeaturesSet - set of strings representing high features

lowFeaturesSet - set of strings representing low features

Returns:

observations

getFeatureIndices

public java.util.TreeSet<java.lang.Integer> getFeatureIndices(java.util.TreeSet<java.lang.Integer> features,
                                                              java.util.TreeSet<java.lang.String> featuresSet)

Computes the a tree set of feature indices that appear in the given string set.

Parameters:

features - tree set of integers each representing a selected feature

featuresSet - tree set of strings each representing a selected feature

Returns:

feature indices

getSamples

public java.lang.String[][] getSamples()

Returns the samples array.

Returns:

the samples array

obsFromARFF

public Observations obsFromARFF(java.util.TreeSet<java.lang.Integer> inputIndexes,
                                java.util.TreeSet<java.lang.Integer> outputIndexes)

Returns the observation for specified high and low features.

Parameters:

inputIndexes - tree set of high feature attribute indexes

outputIndexes - tree set of low feature attribute indexes

Returns:

observation

printFeatures

public void printFeatures(java.util.TreeSet<java.lang.Integer> highFeatures,
                          java.util.TreeSet<java.lang.Integer> lowFeatures,
                          int verbose)

Print the features that are selected to be investigated.

Parameters:

highFeatures - the set of indeces representing high features

lowFeatures - the set of indeces representing low features

verbose - verbose

setTerminateWhenStabilised

public void setTerminateWhenStabilised(boolean flag)

Controls whether the obsFromARFF() method should terminate before it has finished reading all of the available samples, if the corrected leakage value stabilises.

Parameters:

flag - true to terminate early; false to read all of the samples regardless.