weka.core.DistanceFunction Java Examples

/**
 * @param nearestNeighbour The nearest neighbour search algorithm to use.
 * @author Michael
 *
 */
public KNNAugSpaceSampler(final Instances preciseInsts, final Random rng, final int k, final NearestNeighbourSearch nearestNeighbour) {
	super(preciseInsts, rng);
	this.k = k;
	DistanceFunction dist = new EuclideanDistance(preciseInsts);
	String distOptionColumns = String.format("-R first-%d", preciseInsts.numAttributes() - 1);
	String[] distOptions = {distOptionColumns};

	try {
		dist.setOptions(distOptions);
		nearestNeighbour.setDistanceFunction(dist);
		nearestNeighbour.setInstances(preciseInsts);
	} catch (Exception e) {
		logger.error("Could not configure distance function or setup nearest neighbour: {}", e);
	}
	nearestNeighbour.setMeasurePerformance(false);
	this.nearestNeighbour = nearestNeighbour;
}
 

/**
 * Parses a given list of options. Valid options are:
 *
 <!-- options-start -->
 <!-- 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 nnSearchClass = Utils.getOption('A', options);
  if(nnSearchClass.length() != 0) {
    String nnSearchClassSpec[] = Utils.splitOptions(nnSearchClass);
    if(nnSearchClassSpec.length == 0) { 
      throw new Exception("Invalid DistanceFunction specification string."); 
    }
    String className = nnSearchClassSpec[0];
    nnSearchClassSpec[0] = "";

    setDistanceFunction( (DistanceFunction)
                          Utils.forName( DistanceFunction.class, 
                                         className, nnSearchClassSpec) );
  }
  else {
    setDistanceFunction(new EuclideanDistance());
  }
  
  setMeasurePerformance(Utils.getFlag('P',options));
}
 

public static double[][] createDistanceMatrix(Instances data, DistanceFunction distFunc){
    double[][] distMatrix = new double[data.numInstances()][];
    distFunc.setInstances(data);

    for (int i = 1; i < data.numInstances(); i++){
        distMatrix[i] = new double[i];
        Instance first = data.get(i);

        for (int n = 0; n < i; n++){
            distMatrix[i][n] = distFunc.distance(first, data.get(n));
        }
    }

    return distMatrix;
}
 

public static Pair<Instance, Double> findMinDistance(Instances data, Instance inst, DistanceFunction dist){
    double min = dist.distance(data.get(0), inst);
    Instance minI = data.get(0);
    for (int i = 1; i < data.numInstances(); i++) {
        double temp = dist.distance(data.get(i), inst);
        if(temp < min){
            min = temp;
            minI = data.get(i);
        }
    }
    
    return new Pair(minI, min);
}
 

/**
 * sets the distance function to use for nearest neighbour search.
 * 
 * @param df		the distance function to use
 * @throws Exception	if not EuclideanDistance
 */
public void setDistanceFunction(DistanceFunction df) throws Exception {
  if (!(df instanceof EuclideanDistance))
    throw new Exception("KDTree currently only works with "
        + "EuclideanDistanceFunction.");
  m_DistanceFunction = m_EuclideanDistance = (EuclideanDistance) df;
}
 

/**
  * Gets the distance function specification string, which contains the 
  * class name of the distance function class and any options to it.
  *
  * @return the distance function specification string
  */
 protected String getDistanceFSpec() {
   
   DistanceFunction d = getDistanceF();
   if (d instanceof OptionHandler) {
     return d.getClass().getName() + " "
+ Utils.joinOptions(((OptionHandler) d).getOptions());
   }
   return d.getClass().getName();
 }
 

/**
 * sets the distance function to use for instance comparison.
 * 
 * @param df the new distance function to use
 * @throws Exception if instances cannot be processed
 */
public void setDistanceFunction(DistanceFunction df) throws Exception {
  if (!(df instanceof EuclideanDistance)
      && !(df instanceof ManhattanDistance)) {
    throw new Exception(
        "SimpleKMeans currently only supports the Euclidean and Manhattan distances.");
  }
  m_DistanceFunction = df;
}
 

public DistanceFunction getDistanceFunction() {
    return distanceFunction;
}
 

/**
 * Sets the options
 * 
 * @param options
 *            a list of options as an array of strings
 * @throws Exception
 *             if an option is not support
 */
public void setOptions(String[] options) throws Exception {

	// Set the number of the cluster
	String optionString = Utils.getOption('N', options);
	if (optionString.length() != 0) {
		setNumClusters(Integer.parseInt(optionString));
	}

	// Set the number of the maximum iterations
	optionString = Utils.getOption("I", options);
	if (optionString.length() != 0) {
		setMaxIterations(Integer.parseInt(optionString));
	}

	// Set the repeat times
	optionString = Utils.getOption("J", options);
	if (optionString.length() != 0) {
		setRepeatTimes(Integer.parseInt(optionString));
	}

	// Set the distance function
	String distFunctionClass = Utils.getOption('A', options);
	if (distFunctionClass.length() != 0) {
		String distFunctionClassSpec[] = Utils.splitOptions(distFunctionClass);
		if (distFunctionClassSpec.length == 0) {
			throw new Exception("Invalid DistanceFunction specification string.");
		}
		String className = distFunctionClassSpec[0];
		distFunctionClassSpec[0] = "";

		setDistanceFunction(
				(DistanceFunction) Utils.forName(DistanceFunction.class, className, distFunctionClassSpec));
	} else {
		setDistanceFunction(new EuclideanDistance());
	}

	// Set whether to output the cluster result
	m_SaveClusterResult = Utils.getFlag("s", options);

	// Other options
	super.setOptions(options);
}
 

public void setDistanceFunction(DistanceFunction distFunc){
    this.distFunc = distFunc;
}
 

/**
 * Calculates the radius of a node.
 * 
 * @param start The start index of the portion in indices array 
 * that belongs to the node.
 * @param end The end index of the portion in indices array 
 * that belongs to the node. 
 * @param instList The indices array holding indices of 
 * instances. 
 * @param insts The actual instances. instList points to 
 * instances in this object. 
 * @param pivot The centre/pivot of the node. 
 * @param distanceFunction The distance function to use to 
 * calculate the radius. 
 * @return The radius of the node. 
 * @throws Exception If there is some problem calculating the 
 * radius. 
 */
public static double calcRadius(int start, int end, int[] instList, 
                               Instances insts, Instance pivot, 
                               DistanceFunction distanceFunction) 
                                                           throws Exception {
  double radius = Double.NEGATIVE_INFINITY;
  
  for(int i=start; i<=end; i++) {
    double dist = distanceFunction.distance(pivot, 
                                            insts.instance(instList[i]), Double.POSITIVE_INFINITY);
    
    if(dist>radius)
      radius = dist;
  }
  return Math.sqrt(radius);
}
 

/**
 * Parses a given list of options. <p/>
 *
  <!-- options-start -->
 * Valid options are: <p/>
 * 
  <!-- 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 {
  m_bPrintNewick = Utils.getFlag('P', options);

  String optionString = Utils.getOption('N', options); 
  if (optionString.length() != 0) {
    Integer temp = new Integer(optionString);
    setNumClusters(temp);
  }
  else {
    setNumClusters(2);
  }

  setDebug(Utils.getFlag('D', options));
  setDistanceIsBranchLength(Utils.getFlag('B', options));

  String sLinkType = Utils.getOption('L', options);


  if (sLinkType.compareTo("SINGLE") == 0) {setLinkType(new SelectedTag(SINGLE, TAGS_LINK_TYPE));}
  if (sLinkType.compareTo("COMPLETE") == 0) {setLinkType(new SelectedTag(COMPLETE, TAGS_LINK_TYPE));}
  if (sLinkType.compareTo("AVERAGE") == 0) {setLinkType(new SelectedTag(AVERAGE, TAGS_LINK_TYPE));}
  if (sLinkType.compareTo("MEAN") == 0) {setLinkType(new SelectedTag(MEAN, TAGS_LINK_TYPE));}
  if (sLinkType.compareTo("CENTROID") == 0) {setLinkType(new SelectedTag(CENTROID, TAGS_LINK_TYPE));}
  if (sLinkType.compareTo("WARD") == 0) {setLinkType(new SelectedTag(WARD, TAGS_LINK_TYPE));}
  if (sLinkType.compareTo("ADJCOMLPETE") == 0) {setLinkType(new SelectedTag(ADJCOMLPETE, TAGS_LINK_TYPE));}
  if (sLinkType.compareTo("NEIGHBOR_JOINING") == 0) {setLinkType(new SelectedTag(NEIGHBOR_JOINING, TAGS_LINK_TYPE));}

  String nnSearchClass = Utils.getOption('A', options);
  if(nnSearchClass.length() != 0) {
    String nnSearchClassSpec[] = Utils.splitOptions(nnSearchClass);
    if(nnSearchClassSpec.length == 0) { 
      throw new Exception("Invalid DistanceFunction specification string."); 
    }
    String className = nnSearchClassSpec[0];
    nnSearchClassSpec[0] = "";

    setDistanceFunction( (DistanceFunction)
        Utils.forName( DistanceFunction.class, 
            className, nnSearchClassSpec) );
  }
  else {
    setDistanceFunction(new EuclideanDistance());
  }

  Utils.checkForRemainingOptions(options);
}
 

@Override public void setParams(final ParamSet params) {
    ParamHandler.setParam(params, DistanceMeasureable.getDistanceFunctionFlag(), this::setDistanceFunction, DistanceFunction.class);
    ParamHandler.setParam(params, getKFlag(), this::setK, Integer.class);
    ParamHandler.setParam(params, getEarlyAbandonFlag(), this::setEarlyAbandon, Boolean.class);
    ParamHandler.setParam(params, getRandomTieBreakFlag(), this::setRandomTieBreak, Boolean.class);
}
 

@Override public void setParams(final ParamSet param) {
    super.setParams(param);
    ParamHandler.setParam(param, getTransformerFlag(), this::setTransformer, Filter.class);
    ParamHandler.setParam(param, DistanceMeasureable.getDistanceFunctionFlag(), this::setDistanceFunction,
                          DistanceFunction.class);
}
 

@Override
public void setDistanceFunction(DistanceFunction distanceFunction) {
    super.setDistanceFunction(distanceFunction);
}
 

public TransformDistanceMeasure(String name,
    Filter transformer, DistanceFunction distanceFunction) {
    super(name, transformer, distanceFunction);
}
 

public KNNLOOCV(DistanceFunction df) {
    super(df);
    setAbleToEstimateOwnPerformance(true);
}
 

public KNN(DistanceFunction df) {
    this();
    setDistanceFunction(df);
}
 

protected void setDistanceFunction(DistanceFunction distanceFunction) {
    if(distanceFunction == null) throw new NullPointerException();
    this.distanceFunction = distanceFunction;
}
 

public TransformedDistanceMeasure(String name, Filter transformer,
    DistanceFunction distanceFunction) {
    setName(name);
    setDistanceFunction(distanceFunction);
    setTransformer(transformer);
}
 

public DistanceFunction getDistanceFunction() {
    return distanceFunction;
}
 

public void setDistanceFunction(final DistanceFunction distanceFunction) {
    this.distanceFunction = distanceFunction;
}
 

/**
 * Sets the distance function to use for instance comparison.
 * 
 * @param df
 *            the new distance function to use
 * @throws Exception
 *             if df is not EuclideanDistance or ManhattanDistance
 */
public void setDistanceFunction(DistanceFunction df) throws Exception {
	if ((df instanceof EuclideanDistance) || (df instanceof ManhattanDistance)) {
		m_DistanceFunction = df;
	} else {
		throw new Exception("MyPAM only support the Euclidean or Manhattan distance.");
	}
}
 

/**
 * sets the distance function to use for instance comparison.
 * 
 * @param df
 *            the new distance function to use
 * @throws Exception
 *             if instances cannot be processed
 */
public void setDistanceFunction(DistanceFunction df) throws Exception {
	if (!(df instanceof EuclideanDistance) && !(df instanceof ManhattanDistance)) {
		throw new Exception("SimpleKMeans currently only supports the Euclidean and Manhattan distances.");
	}
	m_DistanceFunction = df;
}
 

/**
 * Calculates the radius of node.
 *  
 * @param instList The indices array containing the indices of the 
 * instances inside the node. 
 * @param insts The actual instances object. instList points to 
 * instances in this object.
 * @param pivot The centre/pivot of the node.
 * @param distanceFunction The distance fuction to use to calculate 
 * the radius. 
 * @return The radius of the node. 
 * @throws Exception If there is some problem in calculating the 
 * radius. 
 */
public static double calcRadius(int[] instList, Instances insts,Instance pivot, 
                               DistanceFunction distanceFunction) 
                                                throws Exception {
  return calcRadius(0, instList.length-1, instList, insts, 
                    pivot, distanceFunction);
}
 

/**
 * Calculates the radius of a node based on its two 
 * child nodes (if merging two nodes).
 * @param child1 The first child of the node.
 * @param child2 The second child of the node.
 * @param pivot The centre/pivot of the node. 
 * @param distanceFunction The distance function to 
 * use to calculate the radius
 * @return The radius of the node. 
 * @throws Exception If there is some problem 
 * in calculating the radius.
 */
public static double calcRadius(BallNode child1, BallNode child2, 
                                Instance pivot, 
                                DistanceFunction distanceFunction) 
                                                           throws Exception {
  Instance p1 = child1.getPivot(), p2 = child2.getPivot();                                                               
  
  double radius = child1.getRadius() + distanceFunction.distance(p1, p2) + 
                  child2.getRadius();
  
  return radius/2;
}
 

/**
 * Sets the distance function to use for nearest neighbour search.
 * Currently only EuclideanDistance is supported.
 * 
 * @param df 		the distance function to use 
 * @throws Exception 	if not EuclideanDistance
 */
public void setDistanceFunction(DistanceFunction df) throws Exception {
  if (!(df instanceof EuclideanDistance))
    throw new Exception("CoverTree currently only works with "
 + "EuclideanDistanceFunction.");
  m_DistanceFunction = m_EuclideanDistance = (EuclideanDistance) df;
}
 

/**
 * Returns the distance function currently in use.
 * 
 * @return the distance function currently in use
 */
public DistanceFunction getDistanceFunction() {
	return m_DistanceFunction;
}
 

/** 
 * sets the distance function to use 
 * 
 * @param value	the distance function to use
 */
public void setDistanceFunction(DistanceFunction value) {
  m_DistanceFunction = value;
}
 

/**
 * returns the distance function currently in use.
 * 
 * @return the distance function
 */
public DistanceFunction getDistanceFunction() {
	return m_DistanceFunction;
}