com.evaluation.ConfidenceLabelBasedMeasures.java Source code

Introduction

Here is the source code for com.evaluation.ConfidenceLabelBasedMeasures.java

Source

/*
 *    This program is free software; you can redistribute it and/or modify
 *    it under the terms of the GNU General Public License as published by
 *    the Free Software Foundation; either version 2 of the License, or
 *    (at your option) any later version.
 *
 *    This program is distributed in the hope that it will be useful,
 *    but WITHOUT ANY WARRANTY; without even the implied warranty of
 *    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *    GNU General Public License for more details.
 *
 *    You should have received a copy of the GNU General Public License
 *    along with this program; if not, write to the Free Software
 *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

/*
 *    ConfidenceLabelBasedMeasures.java
 *    Copyright (C) 2009 Aristotle University of Thessaloniki, Thessaloniki, Greece
 *
 */package com.evaluation;

import com.output.MultiLabelOutput;
import weka.classifiers.evaluation.NominalPrediction;
import weka.classifiers.evaluation.ThresholdCurve;
import weka.core.FastVector;
import weka.core.Instances;
import weka.core.Utils;

public class ConfidenceLabelBasedMeasures {

    double[] auc = new double[2];
    double[] labelAUC;

    public ConfidenceLabelBasedMeasures(MultiLabelOutput[] output, boolean[][] trueLabels) {
        computeMeasures(output, trueLabels);
    }

    ConfidenceLabelBasedMeasures(ConfidenceLabelBasedMeasures[] arrayOfMeasures) {
        int numLabels = arrayOfMeasures[0].labelAUC.length;
        labelAUC = new double[numLabels];

        for (ConfidenceLabelBasedMeasures measures : arrayOfMeasures) {
            for (Averaging type : Averaging.values()) {
                auc[type.ordinal()] += measures.getAUC(type);
            }

            for (int labelIndex = 0; labelIndex < numLabels; labelIndex++) {
                labelAUC[labelIndex] += measures.getLabelAUC(labelIndex);
            }
        }

        int arrayLength = arrayOfMeasures.length;
        for (Averaging type : Averaging.values()) {
            auc[type.ordinal()] /= arrayLength;
        }

        for (int labelIndex = 0; labelIndex < numLabels; labelIndex++) {
            labelAUC[labelIndex] /= arrayLength;
        }

    }

    private void computeMeasures(MultiLabelOutput[] output, boolean[][] trueLabels) {
        int numLabels = trueLabels[0].length;

        // AUC
        FastVector[] m_Predictions = new FastVector[numLabels];
        for (int j = 0; j < numLabels; j++)
            m_Predictions[j] = new FastVector();
        FastVector all_Predictions = new FastVector();

        int numInstances = output.length;
        for (int instanceIndex = 0; instanceIndex < numInstances; instanceIndex++) {
            double[] confidences = output[instanceIndex].getConfidences();
            for (int labelIndex = 0; labelIndex < numLabels; labelIndex++) {

                int classValue;
                boolean actual = trueLabels[instanceIndex][labelIndex];
                if (actual)
                    classValue = 1;
                else
                    classValue = 0;

                double[] dist = new double[2];
                dist[1] = confidences[labelIndex];
                dist[0] = 1 - dist[1];

                m_Predictions[labelIndex].addElement(new NominalPrediction(classValue, dist, 1));
                all_Predictions.addElement(new NominalPrediction(classValue, dist, 1));
            }
        }

        labelAUC = new double[numLabels];
        for (int i = 0; i < numLabels; i++) {
            ThresholdCurve tc = new ThresholdCurve();
            Instances result = tc.getCurve(m_Predictions[i], 1);
            labelAUC[i] = ThresholdCurve.getROCArea(result);
        }
        auc[Averaging.MACRO.ordinal()] = Utils.mean(labelAUC);
        ThresholdCurve tc = new ThresholdCurve();
        Instances result = tc.getCurve(all_Predictions, 1);
        auc[Averaging.MICRO.ordinal()] = ThresholdCurve.getROCArea(result);
    }

    public double getLabelAUC(int label) {
        return labelAUC[label];
    }

    public double getAUC(Averaging averagingType) {
        return auc[averagingType.ordinal()];
    }

}