source: trunk/CrossPare/src/de/ugoe/cs/cpdp/dataprocessing/TCAPlusNormalization.java @ 51

package de.ugoe.cs.cpdp.dataprocessing;

import org.apache.commons.math3.ml.distance.EuclideanDistance;

import weka.core.Instances;

// normalization selected according to TCA+ rules (TCA has to be applied separately
public class TCAPlusNormalization implements IProcessesingStrategy {

    private class DistChar {
        private final double mean;
        private final double std;
        private final double min;
        private final double max;
        private int num;
        private DistChar(double mean, double std, double min, double max, int num) {
            this.mean = mean;
            this.std = std;
            this.min = min;
            this.max = max;
            this.num = num;
        }
    }
    
    /**
     * Does not have parameters. String is ignored.
     * 
     * @param parameters
     *            ignored
     */
    @Override
    public void setParameter(String parameters) {
        // TODO Auto-generated method stub
        
    }

    @Override
    public void apply(Instances testdata, Instances traindata) {
        applyTCAPlus(testdata, traindata);
    }
    
    private void applyTCAPlus(Instances testdata, Instances traindata) {
        DistChar dcTest = datasetDistance(testdata);
        DistChar dcTrain = datasetDistance(traindata);
        
        // RULE 1:
        if( 0.9*dcTrain.mean<=dcTest.mean && 1.1*dcTrain.mean>=dcTest.mean &&
            0.9*dcTrain.std<=dcTest.std && 1.1*dcTrain.std>=dcTest.std) {
            // do nothing
        }
        // RULE 2:
        else if((0.4*dcTrain.min>dcTest.min || 1.6*dcTrain.min<dcTest.min) &&
                (0.4*dcTrain.max>dcTest.max || 1.6*dcTrain.min<dcTest.max) &&
                (0.4*dcTrain.min>dcTest.num || 1.6*dcTrain.min<dcTest.num)) {
            NormalizationUtil.minMax(testdata);
            NormalizationUtil.minMax(traindata);
        }
        // RULE 3:
        else if((0.4*dcTrain.std>dcTest.std && dcTrain.num<dcTest.num) || 
                (1.6*dcTrain.std<dcTest.std)&& dcTrain.num>dcTest.num) {
            NormalizationUtil.zScoreTraining(testdata, traindata);
        }
        // RULE 4:
        else if((0.4*dcTrain.std>dcTest.std && dcTrain.num>dcTest.num) || 
                (1.6*dcTrain.std<dcTest.std)&& dcTrain.num<dcTest.num) {
            NormalizationUtil.zScoreTarget(testdata, traindata);
        }
        //RULE 5:
        else {
            NormalizationUtil.zScore(testdata);
            NormalizationUtil.zScore(traindata);
        }
    }
    
    private DistChar datasetDistance(Instances data) {
        double distance;
        double sumAll = 0.0;
        double sumAllQ = 0.0;
        double min = Double.MAX_VALUE;
        double max = Double.MIN_VALUE;
        int numCmp = 0;
        int l = 0;
        double[] inst1 = new double[data.numAttributes()-1];
        double[] inst2 = new double[data.numAttributes()-1];
        EuclideanDistance euclideanDistance = new EuclideanDistance();
        for( int i=0; i<data.numInstances(); i++ ) {
            l=0;
            for( int k=0; k<data.numAttributes(); k++ ) {
                if( k!=data.classIndex() ) {
                    inst1[l] = data.instance(i).value(k);
                }
            }
            for( int j=0; j<data.numInstances(); j++ ) {
                l=0;
                for( int k=0; k<data.numAttributes(); k++ ) {
                    if( k!=data.classIndex() ) {
                        inst2[l] = data.instance(j).value(k);
                    }
                }
                distance = euclideanDistance.compute(inst1, inst2);
                sumAll += distance;
                sumAllQ += distance*distance;
                numCmp++;
                if( distance < min ) {
                    min = distance;
                }
                if( distance > max ) {
                    max = distance;
                }
            }
        }
        double mean = sumAll / numCmp;
        double std = Math.sqrt((sumAllQ-(sumAll*sumAll)/numCmp) *
                                  (1.0d / (numCmp - 1)));
        return new DistChar(mean, std, min, max, data.numInstances());
    }

}