attributeselection.java

代码是一个分类器的实现,其中使用了部分weka的源代码。可以将项目导入eclipse运行

/*
 *    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.
 */

/*
 *    AttributeSelection.java
 *    Copyright (C) 1999 Mark Hall
 *
 */

package weka.filters.supervised.attribute;

import weka.attributeSelection.ASEvaluation;
import weka.attributeSelection.ASSearch;
import weka.attributeSelection.AttributeEvaluator;
import weka.attributeSelection.AttributeTransformer;
import weka.attributeSelection.BestFirst;
import weka.attributeSelection.CfsSubsetEval;
import weka.attributeSelection.Ranker;
import weka.attributeSelection.UnsupervisedAttributeEvaluator;
import weka.attributeSelection.UnsupervisedSubsetEvaluator;
import weka.core.Capabilities;
import weka.core.FastVector;
import weka.core.Instance;
import weka.core.Instances;
import weka.core.Option;
import weka.core.OptionHandler;
import weka.core.SparseInstance;
import weka.core.Utils;
import weka.core.Capabilities.Capability;
import weka.filters.Filter;
import weka.filters.SupervisedFilter;

import java.util.Enumeration;
import java.util.Vector;

/** 
 <!-- globalinfo-start -->
 * A supervised attribute filter that can be used to select attributes. It is very flexible and allows various search and evaluation methods to be combined.
 * <p/>
 <!-- globalinfo-end -->
 * 
 <!-- options-start -->
 * Valid options are: <p/>
 * 
 * <pre> -S &lt;"Name of search class [search options]"&gt;
 *  Sets search method for subset evaluators.
 *  eg. -S "weka.attributeSelection.BestFirst -S 8"</pre>
 * 
 * <pre> -E &lt;"Name of attribute/subset evaluation class [evaluator options]"&gt;
 *  Sets attribute/subset evaluator.
 *  eg. -E "weka.attributeSelection.CfsSubsetEval -L"</pre>
 * 
 * <pre> 
 * Options specific to evaluator weka.attributeSelection.CfsSubsetEval:
 * </pre>
 * 
 * <pre> -M
 *  Treat missing values as a seperate
 *  value.</pre>
 * 
 * <pre> -L
 *  Don't include locally predictive attributes.</pre>
 * 
 * <pre> 
 * Options specific to search weka.attributeSelection.BestFirst:
 * </pre>
 * 
 * <pre> -P &lt;start set&gt;
 *  Specify a starting set of attributes.
 *  Eg. 1,3,5-7.</pre>
 * 
 * <pre> -D &lt;0 = backward | 1 = forward | 2 = bi-directional&gt;
 *  Direction of search. (default = 1).</pre>
 * 
 * <pre> -N &lt;num&gt;
 *  Number of non-improving nodes to
 *  consider before terminating search.</pre>
 * 
 * <pre> -S &lt;num&gt;
 *  Size of lookup cache for evaluated subsets.
 *  Expressed as a multiple of the number of
 *  attributes in the data set. (default = 1)</pre>
 * 
 <!-- options-end -->
 *
 * @author Mark Hall ([e-mail])
 * @version $Revision: 1.6 $
 */
public class AttributeSelection 
  extends Filter
  implements SupervisedFilter, OptionHandler {
  
  /** for serialization */
  static final long serialVersionUID = -296211247688169716L;

  /** the attribute selection evaluation object */
  private weka.attributeSelection.AttributeSelection m_trainSelector;

  /** the attribute evaluator to use */
  private ASEvaluation m_ASEvaluator;

  /** the search method if any */
  private ASSearch m_ASSearch;

  /** holds a copy of the full set of valid  options passed to the filter */
  private String [] m_FilterOptions;

  /** holds the selected attributes  */
  private int [] m_SelectedAttributes;

  /**
   * Returns a string describing this filter
   *
   * @return a description of the filter suitable for
   * displaying in the explorer/experimenter gui
   */
  public String globalInfo() {

    return "A supervised attribute filter that can be used to select " 
      + "attributes. It is very flexible and allows various search " 
      + "and evaluation methods to be combined.";
  }

  /**
   * Constructor
   */
  public AttributeSelection () {
    
    resetOptions();
  }

  /**
   * Returns an enumeration describing the available options.
   * @return an enumeration of all the available options.
   */
  public Enumeration listOptions() {
    
    Vector newVector = new Vector(6);

    newVector.addElement(new Option(
	"\tSets search method for subset evaluators.\n"
	+ "\teg. -S \"weka.attributeSelection.BestFirst -S 8\"", 
	"S", 1,
	"-S <\"Name of search class [search options]\">"));

    newVector.addElement(new Option(
	"\tSets attribute/subset evaluator.\n"
	+ "\teg. -E \"weka.attributeSelection.CfsSubsetEval -L\"",
	"E", 1,
	"-E <\"Name of attribute/subset evaluation class [evaluator options]\">"));
    
    if ((m_ASEvaluator != null) && (m_ASEvaluator instanceof OptionHandler)) {
      Enumeration enu = ((OptionHandler)m_ASEvaluator).listOptions();
      
      newVector.addElement(new Option("", "", 0, "\nOptions specific to "
	   + "evaluator " + m_ASEvaluator.getClass().getName() + ":"));
      while (enu.hasMoreElements()) {
	newVector.addElement((Option)enu.nextElement());
      }
    }
  
    if ((m_ASSearch != null) && (m_ASSearch instanceof OptionHandler)) {
      Enumeration enu = ((OptionHandler)m_ASSearch).listOptions();
    
      newVector.addElement(new Option("", "", 0, "\nOptions specific to "
	      + "search " + m_ASSearch.getClass().getName() + ":"));
      while (enu.hasMoreElements()) {
	newVector.addElement((Option)enu.nextElement());
      }
    }
    return newVector.elements();
  }

  /**
   * Parses a given list of options. <p/>
   * 
   <!-- options-start -->
   * Valid options are: <p/>
   * 
   * <pre> -S &lt;"Name of search class [search options]"&gt;
   *  Sets search method for subset evaluators.
   *  eg. -S "weka.attributeSelection.BestFirst -S 8"</pre>
   * 
   * <pre> -E &lt;"Name of attribute/subset evaluation class [evaluator options]"&gt;
   *  Sets attribute/subset evaluator.
   *  eg. -E "weka.attributeSelection.CfsSubsetEval -L"</pre>
   * 
   * <pre> 
   * Options specific to evaluator weka.attributeSelection.CfsSubsetEval:
   * </pre>
   * 
   * <pre> -M
   *  Treat missing values as a seperate
   *  value.</pre>
   * 
   * <pre> -L
   *  Don't include locally predictive attributes.</pre>
   * 
   * <pre> 
   * Options specific to search weka.attributeSelection.BestFirst:
   * </pre>
   * 
   * <pre> -P &lt;start set&gt;
   *  Specify a starting set of attributes.
   *  Eg. 1,3,5-7.</pre>
   * 
   * <pre> -D &lt;0 = backward | 1 = forward | 2 = bi-directional&gt;
   *  Direction of search. (default = 1).</pre>
   * 
   * <pre> -N &lt;num&gt;
   *  Number of non-improving nodes to
   *  consider before terminating search.</pre>
   * 
   * <pre> -S &lt;num&gt;
   *  Size of lookup cache for evaluated subsets.
   *  Expressed as a multiple of the number of
   *  attributes in the data set. (default = 1)</pre>
   * 
   <!-- options-end -->
   *
   * @param options the list of options as an array of strings
   * @throws Exception if an option is not supported
   */
  public void setOptions(String[] options) throws Exception {
    
    String optionString;
    resetOptions();

    if (Utils.getFlag('X',options)) {
	throw new Exception("Cross validation is not a valid option"
			    + " when using attribute selection as a Filter.");
    }

    optionString = Utils.getOption('E',options);
    if (optionString.length() != 0) {
      optionString = optionString.trim();
      // split a quoted evaluator name from its options (if any)
      int breakLoc = optionString.indexOf(' ');
      String evalClassName = optionString;
      String evalOptionsString = "";
      String [] evalOptions=null;
      if (breakLoc != -1) {
	evalClassName = optionString.substring(0, breakLoc);
	evalOptionsString = optionString.substring(breakLoc).trim();
	evalOptions = Utils.splitOptions(evalOptionsString);
      }
      setEvaluator(ASEvaluation.forName(evalClassName, evalOptions));
    }

    if (m_ASEvaluator instanceof AttributeEvaluator) {
      setSearch(new Ranker());
    }

    optionString = Utils.getOption('S',options);
    if (optionString.length() != 0) {
      optionString = optionString.trim();
      int breakLoc = optionString.indexOf(' ');
      String SearchClassName = optionString;
      String SearchOptionsString = "";
      String [] SearchOptions=null;
      if (breakLoc != -1) {
	SearchClassName = optionString.substring(0, breakLoc);
	SearchOptionsString = optionString.substring(breakLoc).trim();
	SearchOptions = Utils.splitOptions(SearchOptionsString);
      }
      setSearch(ASSearch.forName(SearchClassName, SearchOptions));
    }

    Utils.checkForRemainingOptions(options);
  }