meka.experiment.evaluationstatistics

Class EvaluationStatisticsUtils

java.lang.Object

meka.experiment.evaluationstatistics.EvaluationStatisticsUtils

public class EvaluationStatisticsUtils
extends Object

Helper class for stats related operations.

Version:

$Revision$

Author:

FracPete (fracpete at waikato dot ac dot nz)

Constructor Summary

Constructors

Constructor and Description
EvaluationStatisticsUtils()

Method Summary

Modifier and Type	Method and Description
static List<MultiLabelClassifier>	classifiers(List<EvaluationStatistics> stats, boolean sort) Returns all the unique classifiers of all the statistics.
static List<String>	commandLines(List<EvaluationStatistics> stats, boolean sort) Returns all the unique classifiers of all the statistics.
static List<String>	headers(List<EvaluationStatistics> stats, boolean moveRunFold, boolean addClassifierRelation) Creates a list of headers (= stats keys) from the provided statistics.
static List<String>	keys(List<EvaluationStatistics> stats, boolean sort) Returns all the keys of all the statistics.
static List<String>	measurements(List<EvaluationStatistics> stats) Returns all available measurements.
static List<Number>	measurements(List<EvaluationStatistics> stats, MultiLabelClassifier classifier, weka.core.Instances dataset, String measurement) Returns all the values of a specific measurement for the specified classifier/dataset combination.
static List<Number>	measurements(List<EvaluationStatistics> stats, String classifier, String dataset, String measurement) Returns all the values of a specific measurement for the specified classifier/dataset combination.
static int[][]	rankMatrix(List<EvaluationStatistics> stats, String measurement) Rank Matrix
static List<String>	relations(List<EvaluationStatistics> stats, boolean sort) Returns all the unique relations of all the statistics.
static double[][]	valueMatrix(List<EvaluationStatistics> stats, String measurement) Value Matrix

Methods inherited from class java.lang.Object

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

Constructor Detail

EvaluationStatisticsUtils

public EvaluationStatisticsUtils()

Method Detail

keys

public static List<String> keys(List<EvaluationStatistics> stats,
                                boolean sort)

Returns all the keys of all the statistics.

Parameters:

stats - the stats to inspect

sort - whether to sort the keys alphabetically

Returns:

the keys

classifiers

public static List<MultiLabelClassifier> classifiers(List<EvaluationStatistics> stats,
                                                     boolean sort)

Returns all the unique classifiers of all the statistics.

Parameters:

stats - the stats to inspect

sort - whether to sort the classifiers alphabetically

Returns:

the classifiers

commandLines

public static List<String> commandLines(List<EvaluationStatistics> stats,
                                        boolean sort)

Returns all the unique classifiers of all the statistics.

Parameters:

stats - the stats to inspect

sort - whether to sort the classifiers alphabetically

Returns:

the command-lines

measurements

public static List<Number> measurements(List<EvaluationStatistics> stats,
                                        MultiLabelClassifier classifier,
                                        weka.core.Instances dataset,
                                        String measurement)

Returns all the values of a specific measurement for the specified classifier/dataset combination.

Parameters:

stats - the stats to inspect

classifier - the classifier to look for

dataset - the dataset to look for

measurement - the measurement to retrieve

Returns:

the values

measurements

public static List<Number> measurements(List<EvaluationStatistics> stats,
                                        String classifier,
                                        String dataset,
                                        String measurement)

Returns all the values of a specific measurement for the specified classifier/dataset combination.

Parameters:

stats - the stats to inspect

classifier - the classifier to look for (commandline)

dataset - the dataset to look for (relation name)

measurement - the measurement to retrieve

Returns:

the values

valueMatrix

public static double[][] valueMatrix(List<EvaluationStatistics> stats,
                                     String measurement)

Value Matrix

rankMatrix

public static int[][] rankMatrix(List<EvaluationStatistics> stats,
                                 String measurement)

Rank Matrix

measurements

public static List<String> measurements(List<EvaluationStatistics> stats)

Returns all available measurements.

Parameters:

stats - the stats to inspect

Returns:

the values

relations

public static List<String> relations(List<EvaluationStatistics> stats,
                                     boolean sort)

Returns all the unique relations of all the statistics.

Parameters:

stats - the stats to inspect

sort - whether to sort the relations alphabetically

Returns:

the relations

headers

public static List<String> headers(List<EvaluationStatistics> stats,
                                   boolean moveRunFold,
                                   boolean addClassifierRelation)

Creates a list of headers (= stats keys) from the provided statistics.

Parameters:

stats - the stats to use

moveRunFold - whether to moved "Fold" and "Run" to the start

addClassifierRelation - whether to add "Classifier" and "Relation"

Returns:

the generated header list