relaggs.java

来自「Weka」· Java 代码 · 共 582 行 · 第 1/2 页

   */
  public String attributeIndicesTipText() {
    return 
        "Specify range of attributes to act on; "
      + "this is a comma separated list of attribute indices, with "
      + "\"first\" and \"last\" valid values; Specify an inclusive "
      + "range with \"-\"; eg: \"first-3,5,6-10,last\".";
  }
  
  /**
   * Set the range of attributes to process.
   *
   * @param value 	the new range.
   */
  public void setSelectedRange(String value) {
    m_SelectedRange = new Range(value);
  }

  /**
   * Gets the current range selection.
   *
   * @return 		current selection.
   */
  public Range getSelectedRange() {
    return m_SelectedRange;
  }

  /**
   * Returns the tip text for this property
   *
   * @return 		tip text for this property suitable for
   * 			displaying in the explorer/experimenter gui
   */
  public String invertSelectionTipText() {
    return 
        "Set attribute selection mode. If false, only selected "
      + "attributes in the range will be worked on; if "
      + "true, only non-selected attributes will be processed.";
  }

  /**
   * Sets whether selected columns should be processed or skipped.
   *
   * @param value 	the new invert setting
   */
  public void setInvertSelection(boolean value) {
    m_SelectedRange.setInvert(value);
  }

  /**
   * Gets whether the supplied columns are to be processed or skipped
   *
   * @return 		true if the supplied columns will be kept
   */
  public boolean getInvertSelection() {
    return m_SelectedRange.getInvert();
  }

  /** 
   * Returns the Capabilities of this filter.
   *
   * @return            the capabilities of this object
   * @see               Capabilities
   */
  public Capabilities getCapabilities() {
    Capabilities result = super.getCapabilities();

    // attributes
    result.enable(Capability.NOMINAL_ATTRIBUTES);
    result.enable(Capability.NUMERIC_ATTRIBUTES);
    result.enable(Capability.DATE_ATTRIBUTES);
    result.enable(Capability.RELATIONAL_ATTRIBUTES);
    result.enable(Capability.MISSING_VALUES);
    
    // class
    result.enable(Capability.NOMINAL_CLASS);
    result.enable(Capability.NUMERIC_CLASS);
    result.enable(Capability.DATE_CLASS);
    result.enable(Capability.MISSING_CLASS_VALUES);
    result.enable(Capability.NO_CLASS);
    
    return result;
  }

  /**
   * Determines the output format based on the input format and returns 
   * this. In case the output format cannot be returned immediately, i.e.,
   * immediateOutputFormat() returns false, then this method will be called
   * from batchFinished().
   *
   * @param inputFormat     the input format to base the output format on
   * @return                the output format
   * @throws Exception      in case the determination goes wrong
   * @see   #hasImmediateOutputFormat()
   * @see   #batchFinished()
   */
  protected Instances determineOutputFormat(Instances inputFormat)
      throws Exception {

    Instances 	result;
    Instances	relFormat;
    FastVector	atts;
    int		i;
    int		n;
    int		m;
    int		clsIndex;
    Attribute	att;
    String	prefix;

    m_SelectedRange.setUpper(inputFormat.numAttributes() - 1);
    
    atts     = new FastVector();
    clsIndex = -1;
    for (i = 0; i < inputFormat.numAttributes(); i++) {
      // we don't process the class
      if (i == inputFormat.classIndex()) {
	clsIndex = atts.size();
	atts.addElement(inputFormat.attribute(i).copy());
	continue;
      }
      
      if (!inputFormat.attribute(i).isRelationValued()) {
	atts.addElement(inputFormat.attribute(i).copy());
	continue;
      }
      
      if (!m_SelectedRange.isInRange(i)) {
	if (getDebug())
	  System.out.println(
	      "Attribute " + (i+1) + " (" + inputFormat.attribute(i).name() 
	      + ") skipped.");
	continue;
      }

      // process relational attribute
      prefix    = inputFormat.attribute(i).name() + "_";
      relFormat = inputFormat.attribute(i).relation();
      for (n = 0; n < relFormat.numAttributes(); n++) {
	att = relFormat.attribute(n);
	
	if (att.isNumeric()) {
	  atts.addElement(new Attribute(prefix + att.name() + "_MIN"));
	  atts.addElement(new Attribute(prefix + att.name() + "_MAX"));
	  atts.addElement(new Attribute(prefix + att.name() + "_AVG"));
	  atts.addElement(new Attribute(prefix + att.name() + "_STDEV"));
	  atts.addElement(new Attribute(prefix + att.name() + "_SUM"));
	}
	else if (att.isNominal()) {
	  if (att.numValues() <= m_MaxCardinality) {
	    for (m = 0; m < att.numValues(); m++)
	      atts.addElement(new Attribute(prefix + att.name() + "_" + att.value(m) + "_CNT"));
	  }
	  else {
	    if (getDebug())
	      System.out.println(
		  "Attribute " + (i+1) + "/" + (n+1) 
		  + " (" + inputFormat.attribute(i).name() + "/" + att.name()
		  + ") skipped, " + att.numValues() + " > " + m_MaxCardinality + ".");
	  }
	}
	else {
	  if (getDebug())
	    System.out.println(
		"Attribute " + (i+1) + "/" + (n+1) 
		+ " (" + inputFormat.attribute(i).name() + "/" + att.name()
		+ ") skipped.");
	}
      }
    }
    
    // generate new format
    result = new Instances(inputFormat.relationName(), atts, 0);
    result.setClassIndex(clsIndex);
    
    // neither string nor relational attributes need to be copied to the 
    // output:
    initOutputLocators(result, new int[0]);
    
    return result;
  }

  /**
   * Processes the given data (may change the provided dataset) and returns
   * the modified version. This method is called in batchFinished().
   *
   * @param instances   the data to process
   * @return            the modified data
   * @throws Exception  in case the processing goes wrong
   * @see               #batchFinished()
   */
  protected Instances process(Instances instances) throws Exception {
    Instances		result;
    Instance		inst;
    Instance		newInst;
    Instances		relInstances;
    int			k;
    int			l;
    int			i;
    int			n;
    int			m;
    AttributeStats	stats;
    Attribute		att;
    
    result = getOutputFormat();

    // initialize attribute statistics
    m_AttStats.clear();

    // collect data for all relational attributes
    for (i = 0; i < instances.numAttributes(); i++) {
      if (i == instances.classIndex())
	continue;

      if (!instances.attribute(i).isRelationValued())
	continue;

      if (!m_SelectedRange.isInRange(i))
	continue;

      // compute statistics
      for (k = 0; k < instances.numInstances(); k++) {
	relInstances = instances.instance(k).relationalValue(i);

	for (n = 0; n < relInstances.numAttributes(); n++) {
	  att   = relInstances.attribute(n);
	  stats = null;

	  if (    att.isNumeric() 
	      || (att.isNominal() && att.numValues() <= m_MaxCardinality) ) {
	    stats = relInstances.attributeStats(n);
	    m_AttStats.put(k + "-" + i + "-" + n, stats);
	  }
	}
      }
    }
    
    // convert data
    for (k = 0; k < instances.numInstances(); k++) {
      inst    = instances.instance(k);
      newInst = new Instance(result.numAttributes());
      newInst.setWeight(inst.weight());

      l = 0;
      for (i = 0; i < instances.numAttributes(); i++) {
	if (!instances.attribute(i).isRelationValued()) {
	  newInst.setValue(l, inst.value(i));
	  l++;
	}
	else {
	  if (!m_SelectedRange.isInRange(i))
	    continue;
	  
	  // replace relational data with statistics
	  relInstances = inst.relationalValue(i);
	  for (n = 0; n < relInstances.numAttributes(); n++) {
	    att   = relInstances.attribute(n);
	    stats = (AttributeStats) m_AttStats.get(k + "-" + i + "-" + n);
	    
	    if (att.isNumeric()) {
	      newInst.setValue(l, stats.numericStats.min); l++;
	      newInst.setValue(l, stats.numericStats.max); l++;
	      newInst.setValue(l, stats.numericStats.mean); l++;
	      newInst.setValue(l, stats.numericStats.stdDev); l++;
	      newInst.setValue(l, stats.numericStats.sum); l++;
	    }
	    else if (att.isNominal() && att.numValues() <= m_MaxCardinality) {
	      for (m = 0; m < att.numValues(); m++) {
		newInst.setValue(l, stats.nominalCounts[m]);
		l++;
	      }
	    }
	  }
	}
      }
      
      result.add(newInst);
    }
    
    return result;
  }

  /**
   * runs the filter with the given arguments
   *
   * @param args      the commandline arguments
   */
  public static void main(String[] args) {
    runFilter(new RELAGGS(), args);
  }
}