gr.demokritos.iit.jinsect.casc.console
Class casc

java.lang.Object
  gr.demokritos.iit.jinsect.casc.console.casc

public class casc
extends java.lang.Object

Helper class used in CASC evaluation of 2006 (see Computer-Assisted Stemmatology Challenge).

Field Summary

static java.lang.String	AVERAGE_LINK
static java.lang.String	COMPLETE_LINK
static java.lang.String	SINGLE_LINK

Constructor Summary

casc()

Method Summary

static void	addParent(java.util.HashMap hParents, salvo.jesus.graph.Vertex vA, salvo.jesus.graph.Vertex vB, CASCGraph gTree, double dWeightToParent)
static double	calcCASCDistanceBetween(CASCGraph gCorrect, CASCGraph g2)
static double	calcDistanceBetween(CASCGraph gCorrect, CASCGraph g2)
static double	calcDistanceBetween(CASCGraph gCorrect, CASCGraph g2, boolean bSilent)
static double	euclideanDistance(Distribution p, Distribution q)
static CASCGraph	getDefaultGraph()
static CASCGraph	getDummyGraph()
static java.util.ListIterator	getEdgeIteratorByWeight(UniqueVertexGraph g) Returns a list iterator object that runs through the edge set of a given graph in weight ascending order.
static CASCGraph	getExampleGraph()
static java.lang.String	getFileNameOnly(java.lang.String sFilePath)
static CASCGraph	getLDABasedGraph() Returns a graph based on LDA analysis of the underlying n-gram frequencies.
static CASCGraph	getLevenshteinDistanceGraph(java.lang.String sInputDir)
static CASCGraph	getNGramDistanceAgglomerativelyClusteredGraph(java.lang.String sInputDir, int iMinNGram, int iMaxNGram, int iDist, double dGraphImportance, java.lang.String sClusteringType)
static CASCGraph	getNGramDistanceGraph(java.lang.String sInputDir, int iMinNGram, int iMaxNGram, int iDist, double dGraphImportance, boolean bSilent, java.util.TreeMap hLoadedDocs)
static CASCGraph	getRandomGraph()
static java.lang.String	graphToCASCDot(CASCGraph gTree) Renders a graph to its DOT representation (See GraphViz for more info on the format).
static void	main(java.lang.String[] args)
static double	manhattanDistance(Distribution p, Distribution q)
static double	MPDDistance(Distribution p, Distribution q)
static void	testGraphDist()

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

SINGLE_LINK

public static final java.lang.String SINGLE_LINK

See Also:

Constant Field Values

COMPLETE_LINK

public static final java.lang.String COMPLETE_LINK

See Also:

Constant Field Values

AVERAGE_LINK

public static final java.lang.String AVERAGE_LINK

See Also:

Constant Field Values

Constructor Detail

casc

public casc()

Method Detail

main

public static void main(java.lang.String[] args)

getLevenshteinDistanceGraph

public static CASCGraph getLevenshteinDistanceGraph(java.lang.String sInputDir)

getEdgeIteratorByWeight

public static java.util.ListIterator getEdgeIteratorByWeight(UniqueVertexGraph g)

Returns a list iterator object that runs through the edge set of a given graph in weight ascending order.

Parameters:

g - The given UniqueVertexGraph.

Returns:

A list iterator on the sorted graph edges.

getNGramDistanceGraph

public static CASCGraph getNGramDistanceGraph(java.lang.String sInputDir,
                                              int iMinNGram,
                                              int iMaxNGram,
                                              int iDist,
                                              double dGraphImportance,
                                              boolean bSilent,
                                              java.util.TreeMap hLoadedDocs)

getNGramDistanceAgglomerativelyClusteredGraph

public static CASCGraph getNGramDistanceAgglomerativelyClusteredGraph(java.lang.String sInputDir,
                                                                      int iMinNGram,
                                                                      int iMaxNGram,
                                                                      int iDist,
                                                                      double dGraphImportance,
                                                                      java.lang.String sClusteringType)

getLDABasedGraph

public static CASCGraph getLDABasedGraph()

Returns a graph based on LDA analysis of the underlying n-gram frequencies. The file CASC/LDAcascOutput.txt is expected as input.

euclideanDistance

public static double euclideanDistance(Distribution p,
                                       Distribution q)

Parameters:

p - the first distribution, as a Distribution.

q - the second distribution, as a Distribution.

Returns:

zero if q and p are equal - the euclidean distance, viewing the distributions as feature vectors.

MPDDistance

public static double MPDDistance(Distribution p,
                                 Distribution q)

Parameters:

p - the first distribution, as a Distribution.

q - the second distribution, as a Distribution.

Returns:

zero if q and p are equal - the MPD distance, viewing the distributions as feature vectors.

manhattanDistance

public static double manhattanDistance(Distribution p,
                                       Distribution q)

Parameters:

p - the first distribution, as a Distribution.

q - the second distribution, as a Distribution.

Returns:

zero if q and p are equal - the Manhattan distance, viewing the distributions as feature vectors.

addParent

public static void addParent(java.util.HashMap hParents,
                             salvo.jesus.graph.Vertex vA,
                             salvo.jesus.graph.Vertex vB,
                             CASCGraph gTree,
                             double dWeightToParent)

calcCASCDistanceBetween

public static double calcCASCDistanceBetween(CASCGraph gCorrect,
                                             CASCGraph g2)

calcDistanceBetween

public static double calcDistanceBetween(CASCGraph gCorrect,
                                         CASCGraph g2)

calcDistanceBetween

public static double calcDistanceBetween(CASCGraph gCorrect,
                                         CASCGraph g2,
                                         boolean bSilent)

getExampleGraph

public static CASCGraph getExampleGraph()

getDummyGraph

public static CASCGraph getDummyGraph()

getDefaultGraph

public static CASCGraph getDefaultGraph()

getRandomGraph

public static CASCGraph getRandomGraph()

getFileNameOnly

public static java.lang.String getFileNameOnly(java.lang.String sFilePath)

testGraphDist

public static void testGraphDist()

graphToCASCDot

public static java.lang.String graphToCASCDot(CASCGraph gTree)

Renders a graph to its DOT representation (See GraphViz for more info on the format). The DOT file follows the CASC (Stemmatology Challenge) directives.

Parameters:

gTree - The input graph.

Returns:

The DOT formatted string representation of the graph.