com.mxgraph.analysis

Class mxGraphAnalysis

java.lang.Object

com.mxgraph.analysis.mxGraphAnalysis

public class mxGraphAnalysis
extends java.lang.Object

A singleton class that provides algorithms for graphs. Assume these variables for the following examples:
mxICostFunction cf = mxDistanceCostFunction(); Object[] v = graph.getChildVertices(graph.getDefaultParent()); Object[] e = graph.getChildEdges(graph.getDefaultParent()); mxGraphAnalysis mga = mxGraphAnalysis.getInstance();

Shortest Path (Dijkstra)

For example, to find the shortest path between the first and the second selected cell in a graph use the following code:

Object[] path = mga.getShortestPath(graph, from, to, cf, v.length, true);

Minimum Spanning Tree

This algorithm finds the set of edges with the minimal length that connect all vertices. This algorithm can be used as follows:

Prim

mga.getMinimumSpanningTree(graph, v, cf, true))

Kruskal

mga.getMinimumSpanningTree(graph, v, e, cf))

Connection Components

The union find may be used as follows to determine whether two cells are connected: boolean connected = uf.differ(vertex1, vertex2).

See Also:

mxICostFunction

Field Summary

Fields

Modifier and Type	Field and Description
protected static mxGraphAnalysis	instance Holds the shared instance of this class.

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	mxGraphAnalysis()

Method Summary

Modifier and Type	Method and Description
protected mxFibonacciHeap	createPriorityQueue() Hook for subclassers to provide a custom fibonacci heap.
protected mxUnionFind	createUnionFind(java.lang.Object[] v) Hook for subclassers to provide a custom union find structure.
mxUnionFind	getConnectionComponents(mxGraph graph, java.lang.Object[] v, java.lang.Object[] e) Returns a union find structure representing the connection components of G=(E,V).
static mxGraphAnalysis	getInstance()
java.lang.Object[]	getMinimumSpanningTree(mxGraph graph, java.lang.Object[] v, mxICostFunction cf, boolean directed) Returns the minimum spanning tree (MST) for the graph defined by G=(E,V).
java.lang.Object[]	getMinimumSpanningTree(mxGraph graph, java.lang.Object[] v, java.lang.Object[] e, mxICostFunction cf) Returns the minimum spanning tree (MST) for the graph defined by G=(E,V).
java.lang.Object[]	getShortestPath(mxGraph graph, java.lang.Object from, java.lang.Object to, mxICostFunction cf, int steps, boolean directed) Returns the shortest path between two cells or their descendants represented as an array of edges in order of traversal.
static void	setInstance(mxGraphAnalysis instance) Sets the shared instance of this class.
mxCellState[]	sort(mxCellState[] states, mxICostFunction cf) Returns a sorted set for cells with respect to cf.
double	sum(mxCellState[] states, mxICostFunction cf) Returns the sum of all cost for cells with respect to cf.

Methods inherited from class java.lang.Object

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

Field Detail

instance

protected static mxGraphAnalysis instance

Holds the shared instance of this class.

Constructor Detail

mxGraphAnalysis

protected mxGraphAnalysis()

Method Detail

getInstance

public static mxGraphAnalysis getInstance()

Returns:

Returns the sharedInstance.

setInstance

public static void setInstance(mxGraphAnalysis instance)

Sets the shared instance of this class.

Parameters:

instance - The instance to set.

getShortestPath

public java.lang.Object[] getShortestPath(mxGraph graph,
                                          java.lang.Object from,
                                          java.lang.Object to,
                                          mxICostFunction cf,
                                          int steps,
                                          boolean directed)

Returns the shortest path between two cells or their descendants represented as an array of edges in order of traversal.
This implementation is based on the Dijkstra algorithm.

Parameters:

graph - The object that defines the graph structure

from - The source cell.

to - The target cell (aka sink).

cf - The cost function that defines the edge length.

steps - The maximum number of edges to traverse.

directed - If edge directions should be taken into account.

Returns:

Returns the shortest path as an alternating array of vertices and edges, starting with from and ending with to.

See Also:

createPriorityQueue()

getMinimumSpanningTree

public java.lang.Object[] getMinimumSpanningTree(mxGraph graph,
                                                 java.lang.Object[] v,
                                                 mxICostFunction cf,
                                                 boolean directed)

Returns the minimum spanning tree (MST) for the graph defined by G=(E,V). The MST is defined as the set of all vertices with minimal lengths that forms no cycles in G.
This implementation is based on the algorihm by Prim-Jarnik. It uses O(|E|+|V|log|V|) time when used with a Fibonacci heap and a graph whith a double linked-list datastructure, as is the case with the default implementation.

Parameters:

graph - the object that describes the graph

v - the vertices of the graph

cf - the cost function that defines the edge length

Returns:

Returns the MST as an array of edges

See Also:

createPriorityQueue()

getMinimumSpanningTree

public java.lang.Object[] getMinimumSpanningTree(mxGraph graph,
                                                 java.lang.Object[] v,
                                                 java.lang.Object[] e,
                                                 mxICostFunction cf)

Returns the minimum spanning tree (MST) for the graph defined by G=(E,V). The MST is defined as the set of all vertices with minimal lenths that forms no cycles in G.
This implementation is based on the algorihm by Kruskal. It uses O(|E|log|E|)=O(|E|log|V|) time for sorting the edges, O(|V|) create sets, O(|E|) find and O(|V|) union calls on the union find structure, thus yielding no more than O(|E|log|V|) steps. For a faster implementatin

Parameters:

graph - The object that contains the graph.

v - The vertices of the graph.

e - The edges of the graph.

cf - The cost function that defines the edge length.

Returns:

Returns the MST as an array of edges.

See Also:

getMinimumSpanningTree(mxGraph, Object[], mxICostFunction, boolean), createUnionFind(Object[])

getConnectionComponents

public mxUnionFind getConnectionComponents(mxGraph graph,
                                           java.lang.Object[] v,
                                           java.lang.Object[] e)

Returns a union find structure representing the connection components of G=(E,V).

Parameters:

graph - The object that contains the graph.

v - The vertices of the graph.

e - The edges of the graph.

Returns:

Returns the connection components in G=(E,V)

See Also:

createUnionFind(Object[])

sort

public mxCellState[] sort(mxCellState[] states,
                          mxICostFunction cf)

Returns a sorted set for cells with respect to cf.

Parameters:

states - the cell states to sort

cf - the cost function that defines the order

Returns:

Returns an ordered set of cells wrt. cf

sum

public double sum(mxCellState[] states,
                  mxICostFunction cf)

Returns the sum of all cost for cells with respect to cf.

Parameters:

states - the cell states to use for the sum

cf - the cost function that defines the costs

Returns:

Returns the sum of all cell cost

createUnionFind

protected mxUnionFind createUnionFind(java.lang.Object[] v)

Hook for subclassers to provide a custom union find structure.

Parameters:

v - the array of all elements

Returns:

Returns a union find structure for v

createPriorityQueue

protected mxFibonacciHeap createPriorityQueue()

Hook for subclassers to provide a custom fibonacci heap.