semisupincompletelabelcurvecvresultproducer.java

wekaUT是 university texas austin 开发的基于weka的半指导学习(semi supervised learning)的分类器

	    else 
	      System.out.println("Run:" + run + " Fold:" + fold + " Size:" + m_CurrentSize  + " NumMissingLablels:" + m_NumMissingLabels);
	    Instances test = new Instances(testSet);
	    Instances labeledTrainSubset = new Instances(train, 0, m_CurrentSize);
	    Instances unlabeledTrainSubsetWithLabels = new Instances(train, m_CurrentSize, maxTrainSize()-m_CurrentSize);
	    System.out.println("labeledTrain: " + m_CurrentSize + ", unlabeledTrain: " + unlabeledTrainSubsetWithLabels.numInstances() + ", maxTrain: " + maxTrainSize());
	    if (m_IsTransductive) {
	      for (int i=0; i<test.numInstances(); i++) {
		unlabeledTrainSubsetWithLabels.add(test.instance(i));
	      }
	    }
	    
	    // Select random set of missing class labels to remove from training data 
	    Random random = new Random(fold);
	    int [] marks = new int[numClasses];
	    for (int i=0; i<numClasses; i++) 
	      marks[i] = 0;
	    for (int rand=0; rand<m_NumMissingLabels; rand++) {
	      int classToRemove = random.nextInt(numClasses);
	      while (marks[classToRemove] == 1) {
		classToRemove = random.nextInt(numClasses);
	      }
	      marks[classToRemove] = 1;
	    }
	    
	    // Remove labeled data with missing class labels from training set
	    /*
	    for (int i=0; i<numClasses; i++) {
	      if (marks[i] == 1)
		System.out.println("Removing class: " + i + " from train");
	      else
		System.out.println("Keeping class: " + i + " in train");
	    }
	    */

	    for (int num=0; num<labeledTrainSubset.numInstances(); num++) {
	      Instance inst = labeledTrainSubset.instance(num);
	      if (marks[(int) inst.classValue()] == 1) {
		//		System.out.println("Removing train instance with class " + inst.classValue());
		inst.setClassMissing();
	      }
	    }
	    labeledTrainSubset.deleteWithMissingClass();
	    System.out.println("NumInstances in train: " + labeledTrainSubset.numInstances());

	    // Keep data only having missing class labels in test set
	    
	    if (m_NumMissingLabels != 0) { // if no classes removed, do normal clustering
	      for (int num=0; num<test.numInstances(); num++) {
		Instance inst = test.instance(num);
		if (marks[(int) inst.classValue()] == 0) {
		  //		  System.out.println("Removing test instance with class " + inst.classValue());
		  inst.setClassMissing();
		}
	      }
	      test.deleteWithMissingClass();
	    }
	    System.out.println("NumInstances in test: " + test.numInstances());

	    int classIndex = unlabeledTrainSubsetWithLabels.numAttributes(); // assuming that the last attribute is always the class

	    // Need to remove the class labels from the unlabeledTrainSubsetWithLabels data before training learner

	    Instances unlabeledTrainSubset = new Instances(unlabeledTrainSubsetWithLabels);
	    unlabeledTrainSubset.deleteClassAttribute();
		
	    Object [] seResults;
	    if (m_SplitEvaluator instanceof SemiSupClustererSplitEvaluator) {
	      seResults = ((SemiSupClustererSplitEvaluator) m_SplitEvaluator).getResult(labeledTrainSubset, unlabeledTrainSubset, test, numClasses); 
	    }
	    else {
	      throw new Exception("SplitEvaluator should be SemiSupClustererSplitEvaluator - SemiSupClassifierSplitEvaluator not yet implemented");
	    }
	    
	    Object [] results = new Object [seResults.length + 1];
	    results[0] = getTimestamp();
	    System.arraycopy(seResults, 0, results, 1,
			     seResults.length);
	    if (m_debugOutput) {
	      String resultName = (""+run+"."+(fold+1)+"."+ m_CurrentSize + "." 
				   + Utils.backQuoteChars(runInstances.relationName())
				   +"."
				   +m_SplitEvaluator.toString()).replace(' ','_');
	      resultName = Utils.removeSubstring(resultName, 
						 "weka.clusterers.");
	      resultName = Utils.removeSubstring(resultName, 
						 "weka.filters.");
	      resultName = Utils.removeSubstring(resultName, 
						 "weka.attributeSelection.");
	      m_ZipDest.zipit(m_SplitEvaluator.getRawResultOutput(), resultName);
	    }
	    m_ResultListener.acceptResult(this, key, results);
	  } catch (Exception ex) {
	    // Save the train and test datasets for debugging purposes?
	    throw ex;
	  }
	}
	if (m_PlotPoints != null) {
	  pointNum ++;
	  m_NumMissingLabels = plotPoint(pointNum);
	}
	else {
	  m_NumMissingLabels += m_StepSize;
	}
      }
    }
  }

  /** Determines if the points specified are fractions of the total number of examples */
  protected boolean setIsFraction(){
    if (m_PlotPoints != null){
      if(!isInteger(m_PlotPoints[0]))//if the first point is not an integer
	m_IsFraction = true;
      else
	m_IsFraction = false;
    }
    return m_IsFraction;
  }
  
  /** Return the number of training examples for the ith point on the
   * curve for plotPoints as specified.
   */
  protected int plotPoint(int i) {
    // If i beyond number of given plot points return a value greater than maximum training size
    int length = m_PlotPoints.length;  
    double point = 0;
    if (i >= length) {
      point = m_Instances.numClasses() + 1;
    }
    else {
      point = m_PlotPoints[i];
    }
    return (int)point;
  }
  
  /** Return true if the given double represents an integer value */
  protected static boolean isInteger(double val) {
    return Utils.eq(Math.floor(val), Math.ceil(val));
  }
  
  /**
   * Gets the names of each of the columns produced for a single run.
   * This method should really be static.
   *
   * @return an array containing the name of each column
   */
  public String [] getKeyNames() {

    String [] keyNames = m_SplitEvaluator.getKeyNames();
    // Add in the names of our extra key fields
    int numExtraKeys;
    if(m_IsFraction)
      numExtraKeys = 5;
    else 
      numExtraKeys = 4;
    String [] newKeyNames = new String [keyNames.length + numExtraKeys];
    newKeyNames[0] = DATASET_FIELD_NAME;
    newKeyNames[1] = RUN_FIELD_NAME;
    newKeyNames[2] = FOLD_FIELD_NAME;
    newKeyNames[3] = STEP_FIELD_NAME;
    if(m_IsFraction) newKeyNames[4] = FRACTION_FIELD_NAME;
    System.arraycopy(keyNames, 0, newKeyNames, numExtraKeys, keyNames.length);
    return newKeyNames;
  }

  /**
   * Gets the data types of each of the columns produced for a single run.
   * This method should really be static.
   *
   * @return an array containing objects of the type of each column. The 
   * objects should be Strings, or Doubles.
   */
  public Object [] getKeyTypes() {
    Object [] keyTypes = m_SplitEvaluator.getKeyTypes();
    int numExtraKeys;
    if(m_IsFraction)
      numExtraKeys = 5;
    else 
      numExtraKeys = 4;
    // Add in the types of our extra fields
    Object [] newKeyTypes = new String [keyTypes.length + numExtraKeys];
    newKeyTypes[0] = new String();
    newKeyTypes[1] = new String();
    newKeyTypes[2] = new String();
    newKeyTypes[3] = new String();
    if(m_IsFraction) newKeyTypes[4] = new String();
    System.arraycopy(keyTypes, 0, newKeyTypes, numExtraKeys, keyTypes.length);
    return newKeyTypes;
  }

  /**
   * Gets the names of each of the columns produced for a single run.
   * This method should really be static.
   *
   * @return an array containing the name of each column
   */
  public String [] getResultNames() {

    String [] resultNames = m_SplitEvaluator.getResultNames();
    // Add in the names of our extra Result fields
    String [] newResultNames = new String [resultNames.length + 1];
    newResultNames[0] = TIMESTAMP_FIELD_NAME;
    System.arraycopy(resultNames, 0, newResultNames, 1, resultNames.length);
    return newResultNames;
  }

  /**
   * Gets the data types of each of the columns produced for a single run.
   * This method should really be static.
   *
   * @return an array containing objects of the type of each column. The 
   * objects should be Strings, or Doubles.
   */
  public Object [] getResultTypes() {

    Object [] resultTypes = m_SplitEvaluator.getResultTypes();
    // Add in the types of our extra Result fields
    Object [] newResultTypes = new Object [resultTypes.length + 1];
    newResultTypes[0] = new Double(0);
    System.arraycopy(resultTypes, 0, newResultTypes, 1, resultTypes.length);
    return newResultTypes;
  }

  /**
   * Gets a description of the internal settings of the result
   * producer, sufficient for distinguishing a ResultProducer
   * instance from another with different settings (ignoring
   * those settings set through this interface). For example,
   * a cross-validation ResultProducer may have a setting for the
   * number of folds. For a given state, the results produced should
   * be compatible. Typically if a ResultProducer is an OptionHandler,
   * this string will represent the command line arguments required
   * to set the ResultProducer to that state.
   *
   * @return the description of the ResultProducer state, or null
   * if no state is defined
   */
  public String getCompatibilityState() {

    String result = "-X " + m_NumFolds + " -S " + getStepSize() + 
	" -L " + getLowerSize() + " -U " + getUpperSize() + " ";
    if (m_SplitEvaluator == null) {
      result += "<null SplitEvaluator>";
    } else {
      result += "-W " + m_SplitEvaluator.getClass().getName();
    }
    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 outputFileTipText() {
    return "Set the destination for saving raw output. If the rawOutput "
      +"option is selected, then output from the splitEvaluator for "
      +"individual folds is saved. If the destination is a directory, "
      +"then each output is saved to an individual gzip file; if the "
      +"destination is a file, then each output is saved as an entry "
      +"in a zip file.";
  }

  /**
   * Get the value of OutputFile.
   *
   * @return Value of OutputFile.
   */
  public File getOutputFile() {
    
    return m_OutputFile;
  }
  
  /**
   * Set the value of OutputFile.
   *
   * @param newOutputFile Value to assign to OutputFile.
   */
  public void setOutputFile(File newOutputFile) {
    
    m_OutputFile = newOutputFile;
  }  

  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String numFoldsTipText() {
    return "Number of folds to use in cross validation.";
  }

  /**
   * Get the value of NumFolds.
   *
   * @return Value of NumFolds.
   */
  public int getNumFolds() {
    
    return m_NumFolds;
  }

  
  /**
   * Set the value of NumFolds.
   *
   * @param newNumFolds Value to assign to NumFolds.
   */
  public void setNumFolds(int newNumFolds) {
    
    m_NumFolds = newNumFolds;
  }

  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String isTransductiveTipText() {
    return "Whether evaluation is transductive or not.";
  }

  /**
   * Get the value of IsTransductive.
   *
   * @return Value of IsTransductive.
   */
  public boolean getIsTransductive() {
    
    return m_IsTransductive;
  }

  
  /**
   * Set the value of IsTransductive.
   *
   * @param flag Value to assign to IsTransductive.
   */
  public void setIsTransductive(boolean flag) {
    
    m_IsTransductive = flag;
  }


  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String lowerSizeTipText() {
    return "Set the minimum number of categories to drop in a training set. Setting zero "
      + "here will actually drop <stepSize> number of categories at the first "
      + "step ";
  }

  /**
   * Get the value of LowerSize.
   *
   * @return Value of LowerSize.
   */
  public int getLowerSize() {
    
    return m_LowerSize;
  }
  
  /**
   * Set the value of LowerSize.
   *
   * @param newLowerSize Value to assign to
   * LowerSize.
   */
  public void setLowerSize(int newLowerSize) {
    
    m_LowerSize = newLowerSize;
  }

  /**
   * Returns the tip text for this property
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String upperSizeTipText() {
    return "Set the maximum number of labeled categories to drop in a training set. Setting -1 "
      + "sets no upper limit (other than the total number of categories "
      + "in the full training set)";
  }

  /**
   * Get the value of UpperSize.
   *
   * @return Value of UpperSize.
   */
  public int getUpperSize() {
    
    return m_UpperSize;
  }
  
  /**
   * Set the value of UpperSize.