kstar.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

   * @param first the test instance.
   * @param second the train instance.
   * @param col the index of the attribute in the instance.
   * @return the value of the transformation probability.
   */
  private double attrTransProb(Instance first, Instance second, int col) {
    String debug = "(KStar.attrTransProb)";
    double transProb = 0.0;
    KStarNominalAttribute ksNominalAttr;
    KStarNumericAttribute ksNumericAttr;
    switch ( m_Train.attribute(col).type() )
      {
      case Attribute.NOMINAL:
	ksNominalAttr = new KStarNominalAttribute(first, second, col, m_Train, 
						  m_RandClassCols, 
						  m_Cache[col]);
	ksNominalAttr.setOptions(m_MissingMode, m_BlendMethod, m_GlobalBlend);
	transProb = ksNominalAttr.transProb();
	ksNominalAttr = null;
	break;

      case Attribute.NUMERIC:
	ksNumericAttr = new KStarNumericAttribute(first, second, col, 
						  m_Train, m_RandClassCols, 
						  m_Cache[col]);
	ksNumericAttr.setOptions(m_MissingMode, m_BlendMethod, m_GlobalBlend);
	transProb = ksNumericAttr.transProb();
	ksNumericAttr = null;
	break;
      }
    return transProb;
  }
   
  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String missingModeTipText() {
    return "Determines how missing attribute values are treated.";
  }

  /**
   * Gets the method to use for handling missing values. Will be one of
   * M_NORMAL, M_AVERAGE, M_MAXDIFF or M_DELETE.
   *
   * @return the method used for handling missing values.
   */
  public SelectedTag getMissingMode() {

    return new SelectedTag(m_MissingMode, TAGS_MISSING);
  }
  
  /**
   * Sets the method to use for handling missing values. Values other than
   * M_NORMAL, M_AVERAGE, M_MAXDIFF and M_DELETE will be ignored.
   *
   * @param newMode the method to use for handling missing values.
   */
  public void setMissingMode(SelectedTag newMode) {
    
    if (newMode.getTags() == TAGS_MISSING) {
      m_MissingMode = newMode.getSelectedTag().getID();
    }
  }

  /**
   * Returns an enumeration describing the available options.
   *
   * @return an enumeration of all the available options.
   */
  public Enumeration listOptions() {

    Vector optVector = new Vector( 3 );
    optVector.addElement(new Option(
	      "\tManual blend setting (default 20%)\n",
	      "B", 1, "-B <num>"));
    optVector.addElement(new Option(
	      "\tEnable entropic auto-blend setting (symbolic class only)\n",
	      "E", 0, "-E"));
    optVector.addElement(new Option(
	      "\tSpecify the missing value treatment mode (default a)\n"
	      +"\tValid options are: a(verage), d(elete), m(axdiff), n(ormal)\n",
	      "M", 1,"-M <char>"));
    return optVector.elements();
  }
   
  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String globalBlendTipText() {
    return "The parameter for global blending. Values are restricted to [0,100].";
  }

  /**
   * Set the global blend parameter
   * @param b the value for global blending
   */
  public void setGlobalBlend(int b) {
     m_GlobalBlend = b;
      if ( m_GlobalBlend > 100 ) {
	m_GlobalBlend = 100;
      }
      if ( m_GlobalBlend < 0 ) {
	m_GlobalBlend = 0;
      }
  }

  /**
   * Get the value of the global blend parameter
   * @return the value of the global blend parameter
   */
  public int getGlobalBlend() {
    return m_GlobalBlend;
  }
   
  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String entropicAutoBlendTipText() {
    return "Whether entropy-based blending is to be used.";
  }

  /**
   * Set whether entropic blending is to be used.
   * @param e true if entropic blending is to be used
   */
  public void setEntropicAutoBlend(boolean e) {
    if (e) {
      m_BlendMethod = B_ENTROPY;
    } else {
      m_BlendMethod = B_SPHERE;
    }
  }

  /**
   * Get whether entropic blending being used
   * @return true if entropic blending is used
   */
  public boolean getEntropicAutoBlend() {
    if (m_BlendMethod == B_ENTROPY) {
      return true;
    }

    return false;
  }

  /**
   * Parses a given list of options. <p/>
   *
   <!-- options-start -->
   * Valid options are: <p/>
   * 
   * <pre> -B &lt;num&gt;
   *  Manual blend setting (default 20%)
   * </pre>
   * 
   * <pre> -E
   *  Enable entropic auto-blend setting (symbolic class only)
   * </pre>
   * 
   * <pre> -M &lt;char&gt;
   *  Specify the missing value treatment mode (default a)
   *  Valid options are: a(verage), d(elete), m(axdiff), n(ormal)
   * </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 debug = "(KStar.setOptions)";
    String blendStr = Utils.getOption('B', options);
    if (blendStr.length() != 0) {
      setGlobalBlend(Integer.parseInt(blendStr));
    }

    setEntropicAutoBlend(Utils.getFlag('E', options));
    
    String missingModeStr = Utils.getOption('M', options);
    if (missingModeStr.length() != 0) {
      switch ( missingModeStr.charAt(0) ) {
      case 'a':
	setMissingMode(new SelectedTag(M_AVERAGE, TAGS_MISSING));
	break;
      case 'd':
	setMissingMode(new SelectedTag(M_DELETE, TAGS_MISSING));
	break;
      case 'm':
	setMissingMode(new SelectedTag(M_MAXDIFF, TAGS_MISSING));
	break;
      case 'n':
	setMissingMode(new SelectedTag(M_NORMAL, TAGS_MISSING));
	break;
      default:
	setMissingMode(new SelectedTag(M_AVERAGE, TAGS_MISSING));
      }
    }
    Utils.checkForRemainingOptions(options);
  }


  /**
   * Gets the current settings of K*.
   *
   * @return an array of strings suitable for passing to setOptions()
   */
  public String [] getOptions() {
    // -B <num> -E -M <char>
    String [] options = new String [ 5 ];
    int itr = 0;
    options[itr++] = "-B";
    options[itr++] = "" + m_GlobalBlend;

    if (getEntropicAutoBlend()) {
      options[itr++] = "-E";
    }

    options[itr++] = "-M";
    if (m_MissingMode == M_AVERAGE) {
      options[itr++] = "" + "a";
    }
    else if (m_MissingMode == M_DELETE) {
      options[itr++] = "" + "d";
    }
    else if (m_MissingMode == M_MAXDIFF) {
      options[itr++] = "" + "m";
    }
    else if (m_MissingMode == M_NORMAL) {
      options[itr++] = "" + "n";
    }
    while (itr < options.length) {
      options[itr++] = "";
    }
    return options;
  }

  /**
   * Returns a description of this classifier.
   *
   * @return a description of this classifier as a string.
   */
  public String toString() {
    StringBuffer st = new StringBuffer();
    st.append("KStar Beta Verion (0.1b).\n"
	      +"Copyright (c) 1995-97 by Len Trigg ([e-mail]).\n"
	      +"Java port to Weka by Abdelaziz Mahoui "
	      +"([e-mail]).\n\nKStar options : ");
    String [] ops = getOptions();
    for (int i=0;i<ops.length;i++) {
      st.append(ops[i]+' ');
    }
    return st.toString();
  }
  
  /**
   * Main method for testing this class.
   *
   * @param argv should contain command line options (see setOptions)
   */
  public static void main(String [] argv) {
    runClassifier(new KStar(), argv);
  }

  /**
   * Initializes the m_Attributes of the class.
   */
  private void init_m_Attributes() {
    try {
      m_NumInstances = m_Train.numInstances();
      m_NumClasses = m_Train.numClasses();
      m_NumAttributes = m_Train.numAttributes();
      m_ClassType = m_Train.classAttribute().type();
      m_InitFlag = ON;
    } catch(Exception e) {
      e.printStackTrace();
    }
  }

  /**
   * Updates the m_attributes of the class.
   */
  private void update_m_Attributes() {
    m_NumInstances = m_Train.numInstances();
    m_InitFlag = ON;
  }

  /**
   * Note: for Nominal Class Only!
   * Generates a set of random versions of the class colomn.
   */
  private void generateRandomClassColomns() {
    String debug = "(KStar.generateRandomClassColomns)";
    Random generator = new Random(42);
    //    Random generator = new Random();
    m_RandClassCols = new int [NUM_RAND_COLS+1][];
    int [] classvals = classValues();
    for (int i=0; i < NUM_RAND_COLS; i++) {
      // generate a randomized version of the class colomn
      m_RandClassCols[i] = randomize(classvals, generator);
    }
    // original colomn is preserved in colomn NUM_RAND_COLS
    m_RandClassCols[NUM_RAND_COLS] = classvals;
  }

  /**
   * Note: for Nominal Class Only!
   * Returns an array of the class values
   *
   * @return an array of class values
   */
  private int [] classValues() {
    String debug = "(KStar.classValues)";
    int [] classval = new int[m_NumInstances];
    for (int i=0; i < m_NumInstances; i++) {
      try {
	classval[i] = (int)m_Train.instance(i).classValue();
      } catch (Exception ex) {
	ex.printStackTrace();
      }
    }
    return classval;
  }

  /**
   * Returns a copy of the array with its elements randomly redistributed.
   *
   * @param array the array to randomize.
   * @param generator the random number generator to use
   * @return a copy of the array with its elements randomly redistributed.
   */
  private int [] randomize(int [] array, Random generator) {
    String debug = "(KStar.randomize)";
    int index;
    int temp;
    int [] newArray = new int[array.length];
    System.arraycopy(array, 0, newArray, 0, array.length);
    for (int j = newArray.length - 1; j > 0; j--) {
      index = (int) ( generator.nextDouble() * (double)j );
      temp = newArray[j];
      newArray[j] = newArray[index];
      newArray[index] = temp;
    }
    return newArray;
  }

} // class end