decisiontreebuild.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.MiningAlgorithmSpecification;
import com.prudsys.pdm.Core.MiningAttribute;
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.Records.Arff.MiningArffStream;
import com.prudsys.pdm.Models.Classification.DecisionTree.DecisionTreeAlgorithm;
import com.prudsys.pdm.Models.Classification.DecisionTree.DecisionTreeMiningModel;
import com.prudsys.pdm.Models.Classification.DecisionTree.DecisionTreeNode;
import com.prudsys.pdm.Models.Classification.DecisionTree.DecisionTreeSettings;
import com.prudsys.pdm.Models.Classification.DecisionTree.Algorithms.GenTree.GenTreeAlgorithm;
import com.prudsys.pdm.Utils.GeneralUtils;
import com.prudsys.pdm.Utils.PmmlUtils;

/**
 * Builds a decision tree using the General tree algorithm and writes it to
 * PMML file 'DecisionTreeModel.xml'.
 */
public class DecisionTreeBuild extends BasisExample {

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

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

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

    // Get target attribute:
    MiningAttribute targetAttribute = (MiningAttribute)metaData.getMiningAttribute( "class" );

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

    // Assign settings:
    miningSettings.setTarget(targetAttribute);
    miningSettings.setMinNodeSize(0, DecisionTreeSettings.SIZE_UNIT_COUNT);
    miningSettings.setMaxDepth(100);
    miningSettings.setMinDecreaseInImpurity(0.1);
    miningSettings.verifySettings();

    // Get default mining algorithm specification from 'algorithms.xml':
    MiningAlgorithmSpecification miningAlgorithmSpecification =
      MiningAlgorithmSpecification.getMiningAlgorithmSpecification( "Decision Tree (ID3)", null );
    if( miningAlgorithmSpecification == null )
      throw new MiningException( "Can't find application Decision Tree." );

    // 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("impurityMeasureType", String.valueOf(GenTreeAlgorithm.FS_GainRatio) );
    GeneralUtils.displayMiningAlgSpecParameters(miningAlgorithmSpecification);

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

    // Put it all together:
    algorithm.setMiningInputStream( inputData );
    algorithm.setMiningSettings( miningSettings );
    algorithm.setMiningAlgorithmSpecification( miningAlgorithmSpecification );
    // Parameter specific for DecisionTreeAlgorithm but not in MAS:
    algorithm.setStoreScoreDistribution(true);
    algorithm.verify();

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

    // Show results:
    showTree((DecisionTreeMiningModel) model);

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

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

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

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

  /**
   * Show decision tree model.
   *
   * @param treeModel model of decision tree
   */
  public static void showTree(DecisionTreeMiningModel treeModel) {

    System.out.println("Show Decision Tree:");

    DecisionTreeNode root = (DecisionTreeNode) treeModel.getClassifier();
    System.out.println("Total number of nodes = " + (root.getTotalNumberOfChildren()+1));

    showTreeRecursively(root);
  }

  /**
   * Show tree model recursively.
   *
   * @param node tree node
   */
  private static void showTreeRecursively(DecisionTreeNode node) {

    // Loop over all childs:
    for (int i = 0; i < node.getChildCount(); i++) {
      DecisionTreeNode child = (DecisionTreeNode) node.getChildAt(i);
      System.out.println( "parent: " + node.toString() + " ==> child: " + child.toString());

      // Get child's childs:
      showTreeRecursively(child);
    };
  }

}