classifiersplitevaluator.java

Java 编写的多种数据挖掘算法 包括聚类、分类、预处理等

    testTimeStart = System.currentTimeMillis();
    if(canMeasureCPUTime) 
      CPUStartTime = thMonitor.getThreadUserTime(thID);
    predictions = eval.evaluateModel(m_Classifier, test);
    if(canMeasureCPUTime)
      testCPUTimeElapsed = thMonitor.getThreadUserTime(thID) - CPUStartTime;
    testTimeElapsed = System.currentTimeMillis() - testTimeStart;
    thMonitor = null;
    
    m_result = eval.toSummaryString();
    // The results stored are all per instance -- can be multiplied by the
    // number of instances to get absolute numbers
    int current = 0;
    result[current++] = new Double(train.numInstances());
    result[current++] = new Double(eval.numInstances());
    result[current++] = new Double(eval.correct());
    result[current++] = new Double(eval.incorrect());
    result[current++] = new Double(eval.unclassified());
    result[current++] = new Double(eval.pctCorrect());
    result[current++] = new Double(eval.pctIncorrect());
    result[current++] = new Double(eval.pctUnclassified());
    result[current++] = new Double(eval.kappa());
    
    result[current++] = new Double(eval.meanAbsoluteError());
    result[current++] = new Double(eval.rootMeanSquaredError());
    result[current++] = new Double(eval.relativeAbsoluteError());
    result[current++] = new Double(eval.rootRelativeSquaredError());
    
    result[current++] = new Double(eval.SFPriorEntropy());
    result[current++] = new Double(eval.SFSchemeEntropy());
    result[current++] = new Double(eval.SFEntropyGain());
    result[current++] = new Double(eval.SFMeanPriorEntropy());
    result[current++] = new Double(eval.SFMeanSchemeEntropy());
    result[current++] = new Double(eval.SFMeanEntropyGain());
    
    // K&B stats
    result[current++] = new Double(eval.KBInformation());
    result[current++] = new Double(eval.KBMeanInformation());
    result[current++] = new Double(eval.KBRelativeInformation());
    
    // IR stats
    result[current++] = new Double(eval.truePositiveRate(m_IRclass));
    result[current++] = new Double(eval.numTruePositives(m_IRclass));
    result[current++] = new Double(eval.falsePositiveRate(m_IRclass));
    result[current++] = new Double(eval.numFalsePositives(m_IRclass));
    result[current++] = new Double(eval.trueNegativeRate(m_IRclass));
    result[current++] = new Double(eval.numTrueNegatives(m_IRclass));
    result[current++] = new Double(eval.falseNegativeRate(m_IRclass));
    result[current++] = new Double(eval.numFalseNegatives(m_IRclass));
    result[current++] = new Double(eval.precision(m_IRclass));
    result[current++] = new Double(eval.recall(m_IRclass));
    result[current++] = new Double(eval.fMeasure(m_IRclass));
    result[current++] = new Double(eval.areaUnderROC(m_IRclass));
    
    // Timing stats
    result[current++] = new Double(trainTimeElapsed / 1000.0);
    result[current++] = new Double(testTimeElapsed / 1000.0);
    if(canMeasureCPUTime) {
      result[current++] = new Double((trainCPUTimeElapsed/1000000.0) / 1000.0);
      result[current++] = new Double((testCPUTimeElapsed /1000000.0) / 1000.0);
    }
    else {
      result[current++] = new Double(Instance.missingValue());
      result[current++] = new Double(Instance.missingValue());
    }
    
    // IDs
    if (getAttributeID() >= 0){
      String idsString = "";
      if (test.attribute(m_attID).isNumeric()){
        if (test.numInstances() > 0)
          idsString += test.instance(0).value(m_attID);
        for(int i=1;i<test.numInstances();i++){
          idsString += "|" + test.instance(i).value(m_attID);
        }
      } else {
        if (test.numInstances() > 0)
          idsString += test.instance(0).stringValue(m_attID);
        for(int i=1;i<test.numInstances();i++){
          idsString += "|" + test.instance(i).stringValue(m_attID);
        }
      }
      result[current++] = idsString;
    }
    
    if (getPredTargetColumn()){
      if (test.classAttribute().isNumeric()){
        // Targets
        if (test.numInstances() > 0){
          String targetsString = "";
          targetsString += test.instance(0).value(test.classIndex());
          for(int i=1;i<test.numInstances();i++){
            targetsString += "|" + test.instance(i).value(test.classIndex());
          }
          result[current++] = targetsString;
        }
        
        // Predictions
        if (predictions.length > 0){
          String predictionsString = "";
          predictionsString += predictions[0];
          for(int i=1;i<predictions.length;i++){
            predictionsString += "|" + predictions[i];
          }
          result[current++] = predictionsString;
        }
      } else {
        // Targets
        if (test.numInstances() > 0){
          String targetsString = "";
          targetsString += test.instance(0).stringValue(test.classIndex());
          for(int i=1;i<test.numInstances();i++){
            targetsString += "|" + test.instance(i).stringValue(test.classIndex());
          }
          result[current++] = targetsString;
        }
        
        // Predictions
        if (predictions.length > 0){
          String predictionsString = "";
          predictionsString += test.classAttribute().value((int) predictions[0]);
          for(int i=1;i<predictions.length;i++){
            predictionsString += "|" + test.classAttribute().value((int) predictions[i]);
          }
          result[current++] = predictionsString;
        }
      }
    }
    
    if (m_Classifier instanceof Summarizable) {
      result[current++] = ((Summarizable)m_Classifier).toSummaryString();
    } else {
      result[current++] = null;
    }
    
    for (int i=0;i<addm;i++) {
      if (m_doesProduce[i]) {
        try {
          double dv = ((AdditionalMeasureProducer)m_Classifier).
          getMeasure(m_AdditionalMeasures[i]);
          if (!Instance.isMissingValue(dv)) {
            Double value = new Double(dv);
            result[current++] = value;
          } else {
            result[current++] = null;
          }
        } catch (Exception ex) {
          System.err.println(ex);
        }
      } else {
        result[current++] = null;
      }
    }
    
    if (current != overall_length) {
      throw new Error("Results didn't fit RESULT_SIZE");
    }
    return result;
  }

  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String classifierTipText() {
    return "The classifier to use.";
  }

  /**
   * Get the value of Classifier.
   *
   * @return Value of Classifier.
   */
  public Classifier getClassifier() {
    
    return m_Template;
  }
  
  /**
   * Sets the classifier.
   *
   * @param newClassifier the new classifier to use.
   */
  public void setClassifier(Classifier newClassifier) {
    
    m_Template = newClassifier;
    updateOptions();
  }
  
  /**
   * Get the value of ClassForIRStatistics.
   * @return Value of ClassForIRStatistics.
   */
  public int getClassForIRStatistics() {
    return m_IRclass;
  }
  
  /**
   * Set the value of ClassForIRStatistics.
   * @param v  Value to assign to ClassForIRStatistics.
   */
  public void setClassForIRStatistics(int v) {
    m_IRclass = v;
  }

  /**
   * Get the index of Attibute Identifying the instances
   * @return index of outputed Attribute.
   */
  public int getAttributeID() {
    return m_attID;
  }
  
  /**
   * Set the index of Attibute Identifying the instances
   * @param v index the attribute to output
   */
  public void setAttributeID(int v) {
    m_attID = v;
  }
    
  /**
   *@return true if the prediction and target columns must be outputed.
   */
  public boolean getPredTargetColumn(){
      return m_predTargetColumn;
  }

  /**
   * Set the flag for prediction and target output.
   *@param v true if the 2 columns have to be outputed. false otherwise.
   */
  public void setPredTargetColumn(boolean v){
      m_predTargetColumn = v;
  }
  
  /**
   * Updates the options that the current classifier is using.
   */
  protected void updateOptions() {
    
    if (m_Template instanceof OptionHandler) {
      m_ClassifierOptions = Utils.joinOptions(((OptionHandler)m_Template)
					      .getOptions());
    } else {
      m_ClassifierOptions = "";
    }
    if (m_Template instanceof Serializable) {
      ObjectStreamClass obs = ObjectStreamClass.lookup(m_Template
						       .getClass());
      m_ClassifierVersion = "" + obs.getSerialVersionUID();
    } else {
      m_ClassifierVersion = "";
    }
  }

  /**
   * Set the Classifier to use, given it's class name. A new classifier will be
   * instantiated.
   *
   * @param newClassifierName the Classifier class name.
   * @throws Exception if the class name is invalid.
   */
  public void setClassifierName(String newClassifierName) throws Exception {

    try {
      setClassifier((Classifier)Class.forName(newClassifierName)
		    .newInstance());
    } catch (Exception ex) {
      throw new Exception("Can't find Classifier with class name: "
			  + newClassifierName);
    }
  }

  /**
   * Gets the raw output from the classifier
   * @return the raw output from the classifier
   */
  public String getRawResultOutput() {
    StringBuffer result = new StringBuffer();

    if (m_Classifier == null) {
      return "<null> classifier";
    }
    result.append(toString());
    result.append("Classifier model: \n"+m_Classifier.toString()+'\n');

    // append the performance statistics
    if (m_result != null) {
      result.append(m_result);
      
      if (m_doesProduce != null) {
	for (int i=0;i<m_doesProduce.length;i++) {
	  if (m_doesProduce[i]) {
	    try {
	      double dv = ((AdditionalMeasureProducer)m_Classifier).
		getMeasure(m_AdditionalMeasures[i]);
	      if (!Instance.isMissingValue(dv)) {
		Double value = new Double(dv);
		result.append(m_AdditionalMeasures[i]+" : "+value+'\n');
	      } else {
		result.append(m_AdditionalMeasures[i]+" : "+'?'+'\n');
	      }
	    } catch (Exception ex) {
	      System.err.println(ex);
	    }
	  } 
	}
      }
    }
    return result.toString();
  }

  /**
   * Returns a text description of the split evaluator.
   *
   * @return a text description of the split evaluator.
   */
  public String toString() {

    String result = "ClassifierSplitEvaluator: ";
    if (m_Template == null) {
      return result + "<null> classifier";
    }
    return result + m_Template.getClass().getName() + " " 
      + m_ClassifierOptions + "(version " + m_ClassifierVersion + ")";
  }
} // ClassifierSplitEvaluator