lu.lippmann.cdb.lab.regression.Regression.java Source code

Introduction

Here is the source code for lu.lippmann.cdb.lab.regression.Regression.java

Source

/**
 * Copyright 2014-2016 LIST (Luxembourg Institute of Science and Technology), all right reserved.
 * Authorship : Olivier PARISOT, Yoanne DIDRY
 * Licensed under GNU General Public License version 3
 */
package lu.lippmann.cdb.lab.regression;

import java.io.File;

import lu.lippmann.cdb.common.FormatterUtil;
import lu.lippmann.cdb.weka.*;

import org.apache.commons.math3.stat.regression.OLSMultipleLinearRegression;
import weka.core.Instances;

/**
 * Regression.
 * 
 * @author the WP1 team
 */
public final class Regression {
    //
    // Instance fields
    //

    /** */
    private final Instances newds;
    /** */
    private final double r2;
    /** */
    private final double[] coe;
    /** */
    private final double[] estims;

    //
    // Cosntructors
    //

    /**
     * Constructor.
     */
    public Regression(final Instances ds, final int idx) throws Exception {
        this.newds = WekaDataProcessingUtil.buildDataSetSortedByAttribute(ds, idx);

        //System.out.println("Regression -> "+newds.toSummaryString());

        final int N = this.newds.numInstances();
        final int M = this.newds.numAttributes();

        final double[][] x = new double[N][M - 1];
        final double[] y = new double[N];
        for (int i = 0; i < N; i++) {
            y[i] = this.newds.instance(i).value(0);
        }
        for (int i = 0; i < N; i++) {
            for (int j = 1; j < M; j++) {
                x[i][j - 1] = this.newds.instance(i).value(j);
            }
        }

        final OLSMultipleLinearRegression reg = new OLSMultipleLinearRegression();
        //reg.setNoIntercept(true);
        reg.newSampleData(y, x);

        this.r2 = reg.calculateRSquared();
        //this.r2=-1d;

        this.coe = reg.estimateRegressionParameters();

        this.estims = calculateEstimations(x, y, coe);
    }

    //
    // Instance methods
    //

    public double getR2() {
        return r2;
    }

    public double[] getCoe() {
        return coe;
    }

    public double getCoef(final String attrName) {
        return coe[this.newds.attribute(attrName).index()];
    }

    public String getCoeDesc() {
        final StringBuilder sb = new StringBuilder();
        sb.append(coe[0]).append("+\n");
        for (int i = 1; i < this.newds.numAttributes(); i++) {
            sb.append(this.newds.attribute(i).name()).append("*").append(coe[i]).append("+\n");
        }
        sb.setLength(sb.length() - 2);
        return sb.toString();
    }

    public double[] getEstims() {
        return estims;
    }

    //
    // Static methods
    //

    private static double[] calculateEstimations(double[][] coords, double[] y, double[] coe) {
        if (coe == null)
            return null;

        final double[] estims = new double[coords.length];
        for (int i = 0; i < coords.length; i++) {
            estims[i] = calculateEstimation(coords[i], coe);
        }
        return estims;

    }

    private static double calculateEstimation(double[] coords, double[] coe) {
        double result = coe[0];
        for (int i = 1; i < coe.length; ++i)
            result += coe[i] * coords[i - 1]; // 1
        return result;
    }

    /**
     * Main method.
     * @param args comamnd line arguments    
     */
    public static void main(String[] args) throws Exception {

        final Instances ds = WekaDataAccessUtil.loadInstancesFromARFFOrCSVFile(new File("./samples/csv/bank.csv"));

        final int idx = 0;
        System.out.println("Regression for attribute " + ds.attribute(idx).name());
        final Regression r = new Regression(ds, idx);
        System.out.println(r.getR2());
        System.out.println(FormatterUtil.buildStringFromArrayOfDoubles(r.getCoe()));
        System.out.println(r.getCoeDesc());
        System.out.println(FormatterUtil.buildStringFromArrayOfDoubles(r.getEstims()));
    }

}