discretize.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

   * displaying in the explorer/experimenter gui
   */
  public String useEqualFrequencyTipText() {

    return "If set to true, equal-frequency binning will be used instead of" +
      " equal-width binning.";
  }
  
  /**
   * Get the value of UseEqualFrequency.
   *
   * @return Value of UseEqualFrequency.
   */
  public boolean getUseEqualFrequency() {
    
    return m_UseEqualFrequency;
  }
  
  /**
   * Set the value of UseEqualFrequency.
   *
   * @param newUseEqualFrequency Value to assign to UseEqualFrequency.
   */
  public void setUseEqualFrequency(boolean newUseEqualFrequency) {
    
    m_UseEqualFrequency = newUseEqualFrequency;
  }

  /**
   * Returns the tip text for this property
   *
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  public String binsTipText() {

    return "Number of bins.";
  }

  /**
   * Gets the number of bins numeric attributes will be divided into
   *
   * @return the number of bins.
   */
  public int getBins() {

    return m_NumBins;
  }

  /**
   * Sets the number of bins to divide each selected numeric attribute into
   *
   * @param numBins the number of bins
   */
  public void setBins(int numBins) {

    m_NumBins = numBins;
  }

  /**
   * 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"
      + " (numeric) attributes in the range will be discretized; if"
      + " true, only non-selected attributes will be discretized.";
  }

  /**
   * Gets whether the supplied columns are to be removed or kept
   *
   * @return true if the supplied columns will be kept
   */
  public boolean getInvertSelection() {

    return m_DiscretizeCols.getInvert();
  }

  /**
   * Sets whether selected columns should be removed or kept. If true the 
   * selected columns are kept and unselected columns are deleted. If false
   * selected columns are deleted and unselected columns are kept.
   *
   * @param invert the new invert setting
   */
  public void setInvertSelection(boolean invert) {

    m_DiscretizeCols.setInvert(invert);
  }

  /**
   * Returns the tip text for this property
   *
   * @return tip text for this property suitable for
   * displaying in the explorer/experimenter gui
   */
  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 \"-\". E.g: \"first-3,5,6-10,last\".";
  }

  /**
   * Gets the current range selection
   *
   * @return a string containing a comma separated list of ranges
   */
  public String getAttributeIndices() {

    return m_DiscretizeCols.getRanges();
  }

  /**
   * Sets which attributes are to be Discretized (only numeric
   * attributes among the selection will be Discretized).
   *
   * @param rangeList a string representing the list of attributes. Since
   * the string will typically come from a user, attributes are indexed from
   * 1. <br>
   * eg: first-3,5,6-last
   * @throws IllegalArgumentException if an invalid range list is supplied 
   */
  public void setAttributeIndices(String rangeList) {

    m_DiscretizeCols.setRanges(rangeList);
  }

  /**
   * Sets which attributes are to be Discretized (only numeric
   * attributes among the selection will be Discretized).
   *
   * @param attributes an array containing indexes of attributes to Discretize.
   * Since the array will typically come from a program, attributes are indexed
   * from 0.
   * @throws IllegalArgumentException if an invalid set of ranges
   * is supplied 
   */
  public void setAttributeIndicesArray(int [] attributes) {

    setAttributeIndices(Range.indicesToRangeList(attributes));
  }

  /**
   * Gets the cut points for an attribute
   *
   * @param attributeIndex the index (from 0) of the attribute to get the cut points of
   * @return an array containing the cutpoints (or null if the
   * attribute requested has been discretized into only one interval.)
   */
  public double [] getCutPoints(int attributeIndex) {

    if (m_CutPoints == null) {
      return null;
    }
    return m_CutPoints[attributeIndex];
  }

  /** Generate the cutpoints for each attribute */
  protected void calculateCutPoints() {

    m_CutPoints = new double [getInputFormat().numAttributes()] [];
    for(int i = getInputFormat().numAttributes() - 1; i >= 0; i--) {
      if ((m_DiscretizeCols.isInRange(i)) && 
	  (getInputFormat().attribute(i).isNumeric()) &&
	  (getInputFormat().classIndex() != i)) {
	if (m_FindNumBins) {
	  findNumBins(i);
	} else if (!m_UseEqualFrequency) {
	  calculateCutPointsByEqualWidthBinning(i);
	} else {
	  calculateCutPointsByEqualFrequencyBinning(i);
	}
      }
    }
  }
 
  /**
   * Set cutpoints for a single attribute.
   *
   * @param index the index of the attribute to set cutpoints for
   */
  protected void calculateCutPointsByEqualWidthBinning(int index) {

    // Scan for max and min values
    double max = 0, min = 1, currentVal;
    Instance currentInstance;
    for(int i = 0; i < getInputFormat().numInstances(); i++) {
      currentInstance = getInputFormat().instance(i);
      if (!currentInstance.isMissing(index)) {
	currentVal = currentInstance.value(index);
	if (max < min) {
	  max = min = currentVal;
	}
	if (currentVal > max) {
	  max = currentVal;
	}
	if (currentVal < min) {
	  min = currentVal;
	}
      }
    }
    double binWidth = (max - min) / m_NumBins;
    double [] cutPoints = null;
    if ((m_NumBins > 1) && (binWidth > 0)) {
      cutPoints = new double [m_NumBins - 1];
      for(int i = 1; i < m_NumBins; i++) {
	cutPoints[i - 1] = min + binWidth * i;
      }
    }
    m_CutPoints[index] = cutPoints;
  }
 
  /**
   * Set cutpoints for a single attribute.
   *
   * @param index the index of the attribute to set cutpoints for
   */
  protected void calculateCutPointsByEqualFrequencyBinning(int index) {

    // Copy data so that it can be sorted
    Instances data = new Instances(getInputFormat());

    // Sort input data
    data.sort(index);

    // Compute weight of instances without missing values
    double sumOfWeights = 0;
    for (int i = 0; i < data.numInstances(); i++) {
      if (data.instance(i).isMissing(index)) {
	break;
      } else {
	sumOfWeights += data.instance(i).weight();
      }
    }
    double freq;
    double[] cutPoints = new double[m_NumBins - 1];
    if (getDesiredWeightOfInstancesPerInterval() > 0) {
      freq = getDesiredWeightOfInstancesPerInterval();
      cutPoints = new double[(int)(sumOfWeights / freq)];
    } else {
      freq = sumOfWeights / m_NumBins;
      cutPoints = new double[m_NumBins - 1];
    }

    // Compute break points
    double counter = 0, last = 0;
    int cpindex = 0, lastIndex = -1;
    for (int i = 0; i < data.numInstances() - 1; i++) {

      // Stop if value missing
      if (data.instance(i).isMissing(index)) {
	break;
      }
      counter += data.instance(i).weight();
      sumOfWeights -= data.instance(i).weight();

      // Do we have a potential breakpoint?
      if (data.instance(i).value(index) < 
	  data.instance(i + 1).value(index)) {

	// Have we passed the ideal size?
	if (counter >= freq) {

	  // Is this break point worse than the last one?
	  if (((freq - last) < (counter - freq)) && (lastIndex != -1)) {
	    cutPoints[cpindex] = (data.instance(lastIndex).value(index) +
				  data.instance(lastIndex + 1).value(index)) / 2;
	    counter -= last;
	    last = counter;
	    lastIndex = i;
	  } else {
	    cutPoints[cpindex] = (data.instance(i).value(index) +
				  data.instance(i + 1).value(index)) / 2;
	    counter = 0;
	    last = 0;
	    lastIndex = -1;
	  }
	  cpindex++;
	  freq = (sumOfWeights + counter) / ((cutPoints.length + 1) - cpindex);
	} else {
	  lastIndex = i;
	  last = counter;
	}
      }
    }

    // Check whether there was another possibility for a cut point
    if ((cpindex < cutPoints.length) && (lastIndex != -1)) {
      cutPoints[cpindex] = (data.instance(lastIndex).value(index) +
			    data.instance(lastIndex + 1).value(index)) / 2;      
      cpindex++;
    }

    // Did we find any cutpoints?
    if (cpindex == 0) {
      m_CutPoints[index] = null;
    } else {
      double[] cp = new double[cpindex];
      for (int i = 0; i < cpindex; i++) {
	cp[i] = cutPoints[i];
      }
      m_CutPoints[index] = cp;
    }
  }

  /**
   * Optimizes the number of bins using leave-one-out cross-validation.
   *
   * @param index the attribute index
   */
  protected void findNumBins(int index) {

    double min = Double.MAX_VALUE, max = -Double.MAX_VALUE, binWidth = 0, 
      entropy, bestEntropy = Double.MAX_VALUE, currentVal;
    double[] distribution;
    int bestNumBins  = 1;
    Instance currentInstance;

    // Find minimum and maximum
    for (int i = 0; i < getInputFormat().numInstances(); i++) {
      currentInstance = getInputFormat().instance(i);
      if (!currentInstance.isMissing(index)) {
	currentVal = currentInstance.value(index);
	if (currentVal > max) {
	  max = currentVal;
	}
	if (currentVal < min) {
	  min = currentVal;
	}
      }
    }

    // Find best number of bins
    for (int i = 0; i < m_NumBins; i++) {
      distribution = new double[i + 1];
      binWidth = (max - min) / (i + 1);

      // Compute distribution
      for (int j = 0; j < getInputFormat().numInstances(); j++) {
	currentInstance = getInputFormat().instance(j);
	if (!currentInstance.isMissing(index)) {
	  for (int k = 0; k < i + 1; k++) {
	    if (currentInstance.value(index) <= 
		(min + (((double)k + 1) * binWidth))) {
	      distribution[k] += currentInstance.weight();
	      break;
	    }
	  }
	}
      }

      // Compute cross-validated entropy
      entropy = 0;
      for (int k = 0; k < i + 1; k++) {
	if (distribution[k] < 2) {
	  entropy = Double.MAX_VALUE;
	  break;
	}
	entropy -= distribution[k] * Math.log((distribution[k] - 1) / 
					      binWidth);
      }

      // Best entropy so far?
      if (entropy < bestEntropy) {
	bestEntropy = entropy;
	bestNumBins = i + 1;
      }
    }

    // Compute cut points
    double [] cutPoints = null;
    if ((bestNumBins > 1) && (binWidth > 0)) {
      cutPoints = new double [bestNumBins - 1];
      for(int i = 1; i < bestNumBins; i++) {
	cutPoints[i - 1] = min + binWidth * i;
      }
    }
    m_CutPoints[index] = cutPoints;
   }

  /**
   * Set the output format. Takes the currently defined cutpoints and 
   * m_InputFormat and calls setOutputFormat(Instances) appropriately.
   */
  protected void setOutputFormat() {

    if (m_CutPoints == null) {
      setOutputFormat(null);
      return;
    }
    FastVector attributes = new FastVector(getInputFormat().numAttributes());
    int classIndex = getInputFormat().classIndex();
    for(int i = 0; i < getInputFormat().numAttributes(); i++) {
      if ((m_DiscretizeCols.isInRange(i)) 
	  && (getInputFormat().attribute(i).isNumeric())
	  && (getInputFormat().classIndex() != i)) {
	if (!m_MakeBinary) {
	  FastVector attribValues = new FastVector(1);
	  if (m_CutPoints[i] == null) {
	    attribValues.addElement("'All'");
	  } else {
	    for(int j = 0; j <= m_CutPoints[i].length; j++) {
	      if (j == 0) {
		attribValues.addElement("'(-inf-"
			+ Utils.doubleToString(m_CutPoints[i][j], 6) + "]'");
	      } else if (j == m_CutPoints[i].length) {
		attribValues.addElement("'("
			+ Utils.doubleToString(m_CutPoints[i][j - 1], 6) 
					+ "-inf)'");
	      } else {
		attribValues.addElement("'("
			+ Utils.doubleToString(m_CutPoints[i][j - 1], 6) + "-"
			+ Utils.doubleToString(m_CutPoints[i][j], 6) + "]'");
	      }
	    }
	  }
	  attributes.addElement(new Attribute(getInputFormat().
					      attribute(i).name(),
					      attribValues));
	} else {
	  if (m_CutPoints[i] == null) {
	    FastVector attribValues = new FastVector(1);
	    attribValues.addElement("'All'");
	    attributes.addElement(new Attribute(getInputFormat().
						attribute(i).name(),
						attribValues));
	  } else {
	    if (i < getInputFormat().classIndex()) {
	      classIndex += m_CutPoints[i].length - 1;
	    }
	    for(int j = 0; j < m_CutPoints[i].length; j++) {
	      FastVector attribValues = new FastVector(2);
	      attribValues.addElement("'(-inf-"
		      + Utils.doubleToString(m_CutPoints[i][j], 6) + "]'");
	      attribValues.addElement("'("
		      + Utils.doubleToString(m_CutPoints[i][j], 6) + "-inf)'");
	      attributes.addElement(new Attribute(getInputFormat().
						  attribute(i).name(),
						  attribValues));
	    }
	  }
	}
      } else {
	attributes.addElement(getInputFormat().attribute(i).copy());
      }
    }
    Instances outputFormat = 
      new Instances(getInputFormat().relationName(), attributes, 0);
    outputFormat.setClassIndex(classIndex);
    setOutputFormat(outputFormat);
  }

  /**
   * Convert a single instance over. The converted instance is added to 
   * the end of the output queue.
   *
   * @param instance the instance to convert
   */
  protected void convertInstance(Instance instance) {

    int index = 0;
    double [] vals = new double [outputFormatPeek().numAttributes()];
    // Copy and convert the values
    for(int i = 0; i < getInputFormat().numAttributes(); i++) {
      if (m_DiscretizeCols.isInRange(i) && 
	  getInputFormat().attribute(i).isNumeric() &&
	  (getInputFormat().classIndex() != i)) {
	int j;
	double currentVal = instance.value(i);
	if (m_CutPoints[i] == null) {
	  if (instance.isMissing(i)) {
	    vals[index] = Instance.missingValue();
	  } else {
	    vals[index] = 0;
	  }
	  index++;
	} else {
	  if (!m_MakeBinary) {
	    if (instance.isMissing(i)) {
	      vals[index] = Instance.missingValue();
	    } else {
	      for (j = 0; j < m_CutPoints[i].length; j++) {
		if (currentVal <= m_CutPoints[i][j]) {
		  break;
		}
	      }
              vals[index] = j;
	    }
	    index++;
	  } else {
	    for (j = 0; j < m_CutPoints[i].length; j++) {
	      if (instance.isMissing(i)) {
                vals[index] = Instance.missingValue();
	      } else if (currentVal <= m_CutPoints[i][j]) {
                vals[index] = 0;
	      } else {
                vals[index] = 1;
	      }
	      index++;
	    }
	  }   
	}
      } else {
        vals[index] = instance.value(i);
	index++;
      }
    }
    
    Instance inst = null;
    if (instance instanceof SparseInstance) {
      inst = new SparseInstance(instance.weight(), vals);
    } else {
      inst = new Instance(instance.weight(), vals);
    }
    inst.setDataset(getOutputFormat());
    copyValues(inst, false, instance.dataset(), getOutputFormat());
    inst.setDataset(getOutputFormat());
    push(inst);
  }

  /**
   * Main method for testing this class.
   *
   * @param argv should contain arguments to the filter: use -h for help
   */
  public static void main(String [] argv) {
    runFilter(new Discretize(), argv);
  }
}