org.openimaj.ml.clustering.spectral

Class GraphLaplacian

java.lang.Object

org.openimaj.ml.clustering.spectral.GraphLaplacian

Direct Known Subclasses:

GraphLaplacian.Normalised, GraphLaplacian.Unnormalised, GraphLaplacian.Warped

public abstract class GraphLaplacian
extends Object

Functions which turn a graph weight adjacency matrix into the Laplacian matrix.

Author:

Sina Samangooei (ss@ecs.soton.ac.uk)

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	GraphLaplacian.Normalised The inverted symmetric normalised Laplacian is defined as: L = D^-1/2 A D^-1/2
static class	GraphLaplacian.Unnormalised The symmetric normalised Laplacian is defined as: L = D - W
static class	GraphLaplacian.Warped The inverted symmetric normalised Laplacian is defined as: L = D^-1 .

Constructor Summary

Constructors

Constructor and Description
GraphLaplacian()

Method Summary

Modifier and Type	Method and Description
Iterator<DoubleObjectPair<ch.akuhn.matrix.Vector>>	eigenIterator(ch.akuhn.matrix.eigenvalues.Eigenvalues evd)
ch.akuhn.matrix.SparseMatrix	laplacian(ch.akuhn.matrix.SparseMatrix adj)
abstract ch.akuhn.matrix.SparseMatrix	laplacian(ch.akuhn.matrix.SparseMatrix adj, org.openimaj.math.matrix.DiagonalMatrix degree)

Methods inherited from class java.lang.Object

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

Constructor Detail

GraphLaplacian

public GraphLaplacian()

Method Detail

laplacian

public ch.akuhn.matrix.SparseMatrix laplacian(ch.akuhn.matrix.SparseMatrix adj)

Parameters:

adj - the adjanceny matrix should be square and symmetric

Returns:

the laplacian

laplacian

public abstract ch.akuhn.matrix.SparseMatrix laplacian(ch.akuhn.matrix.SparseMatrix adj,
                                                       org.openimaj.math.matrix.DiagonalMatrix degree)

Parameters:

adj - square and symmetric

degree - the sum of the adjacency for a node in the diagonals

Returns:

the laplacian

eigenIterator

public Iterator<DoubleObjectPair<ch.akuhn.matrix.Vector>> eigenIterator(ch.akuhn.matrix.eigenvalues.Eigenvalues evd)

Parameters:

evd -

Returns:

provides an iterator over the (presumeably sorted)