clusteringbuild.java

一个数据挖掘软件ALPHAMINERR的整个过程的JAVA版源代码

/*
 *    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.
 */

 /**
  * Title: XELOPES Data Mining Library
  * Description: The XELOPES library is an open platform-independent and data-source-independent library for Embedded Data Mining.
  * Copyright: Copyright (c) 2002 Prudential Systems Software GmbH
  * Company: ZSoft (www.zsoft.ru), Prudsys (www.prudsys.com)
  * @author Carsten Weisse
  * @author Michael Thess
  * @version 1.0
  */

package com.prudsys.pdm.Examples;

import java.io.FileWriter;

import com.prudsys.pdm.Core.MiningAlgorithm;
import com.prudsys.pdm.Core.MiningAlgorithmSpecification;
import com.prudsys.pdm.Core.MiningDataSpecification;
import com.prudsys.pdm.Core.MiningException;
import com.prudsys.pdm.Core.MiningModel;
import com.prudsys.pdm.Input.MiningInputStream;
import com.prudsys.pdm.Input.MiningVector;
import com.prudsys.pdm.Input.Records.Arff.MiningArffStream;
import com.prudsys.pdm.Models.Clustering.Cluster;
import com.prudsys.pdm.Models.Clustering.ClusteringMiningModel;
import com.prudsys.pdm.Models.Clustering.Distance;
import com.prudsys.pdm.Models.Clustering.CDBased.CDBasedClusteringAlgorithm;
import com.prudsys.pdm.Models.Clustering.CDBased.CDBasedClusteringSettings;
import com.prudsys.pdm.Utils.GeneralUtils;
import com.prudsys.pdm.Utils.PmmlUtils;

/**
 * Builds a clustering model using the K-Means algorithm and writes it to
 * PMML file 'ClusteringModel.xml'.
 */
public class ClusteringBuild extends BasisExample {

  /**
   * Empty constructor.
   */
  public ClusteringBuild() {
  }

  /**
   * Run the example of this class.
   *
   * @throws Exception error while example is running
   */
  public void runExample() throws Exception {

    // Open data source and get metadata:
    MiningInputStream inputData = new MiningArffStream( "data/arff/iris.arff" );
    MiningDataSpecification metaData = inputData.getMetaData();

    // Create MiningSettings object and assign metadata:
    CDBasedClusteringSettings miningSettings = new CDBasedClusteringSettings();
    miningSettings.setDataSpecification( metaData );

    // Assign settings:
    Distance dist = new Distance();
    dist.setType( Distance.TYPE_EUCLIDEAN );
    dist.setCompareFunction( Distance.COMPARISON_FUNCTION_ABS_DIFF );
    dist.setMeasureType( Distance.MEASURE_TYPE_DISTANCE );
    dist.setNormalized( false );
    miningSettings.setDistance( dist );
    miningSettings.verifySettings();

    // Get default mining algorithm specification from 'algorithms.xml':
    MiningAlgorithmSpecification miningAlgorithmSpecification =
      MiningAlgorithmSpecification.getMiningAlgorithmSpecification( "KMeans", null);
    if( miningAlgorithmSpecification == null )
      throw new MiningException( "Can't find clustering method." );

    // Get class name from algorithms specification:
    String className = miningAlgorithmSpecification.getClassname();
    if( className == null )
      throw new MiningException( "classname attribute expected." );

    // Set and display mining parameters:
    miningAlgorithmSpecification.setMAPValue("numberOfClusters", "3");
    GeneralUtils.displayMiningAlgSpecParameters(miningAlgorithmSpecification);

    // Create algorithm object with default values:
    MiningAlgorithm algorithm = (CDBasedClusteringAlgorithm)
        GeneralUtils.createMiningAlgorithmInstance(className);

    // Put it all together:
    algorithm.setMiningInputStream( inputData );
    algorithm.setMiningSettings( miningSettings );
    algorithm.setMiningAlgorithmSpecification( miningAlgorithmSpecification );
    algorithm.verify();

    // Build the mining model:
    MiningModel model = algorithm.buildModel();
    System.out.println("calculation time [s]: " + algorithm.getTimeSpentToBuildModel());

    // Show the clusters:
    showClusters( (ClusteringMiningModel) model );

    // Write to PMML:
    FileWriter writer = new FileWriter("data/pmml/ClusteringModel.xml");
    model.writePmml(writer);

    // Show in browser:
    if (debug == 2) PmmlUtils.openPmmlBrowser("ClusteringModel.xml");

    // Show relation header:
    System.out.println("Prediction:");
    for (int i = 0; i < metaData.getAttributesNumber(); i++) {
      System.out.print(metaData.getMiningAttribute(i).getName() + " ");
    };

    // Show clustering results:
    System.out.println();
    int i = 0;
    inputData.reset();
    while (inputData.next()) {
      // Make prediction:
      MiningVector vector = inputData.read();
      int predicted       = (int) model.applyModelFunction(vector);

      // Output and stats:
      System.out.println(" " + ++i +": " + vector + " -> " + predicted);
    };
  }

  /**
   * Example of building a cluster model.
   *
   * @param args arguments (ignored)
   */
  public static void main(String[] args) {

    try {
      new ClusteringBuild().runExample();
    }
    catch (Exception ex) {
      ex.printStackTrace();
    }
  }

  /**
   * Shows clusters.
   *
   * @param clustModel clustering model to show
   * @throws MiningException cannot show clusters
   */
  public static void showClusters(ClusteringMiningModel clustModel)
    throws MiningException {

    System.out.println("number of clusters: " + clustModel.getNumberOfClusters());
    Cluster[] clust = clustModel.getClusters();
    for (int i = 0; i < clust.length; i++)
      System.out.println("Clust["+i+"]: " + clust[i].toString() );
  }
}