org.apache.hama.graph

Class Vertex<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

java.lang.Object

org.apache.hama.graph.Vertex<V,E,M>

Type Parameters:

V - Vertex ID object type

E - Edge cost object type

M - Vertex value object type

All Implemented Interfaces:

Comparable<VertexInterface<V,E,M>>, org.apache.hadoop.io.Writable, org.apache.hadoop.io.WritableComparable<VertexInterface<V,E,M>>, VertexInterface<V,E,M>

public abstract class Vertex<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>
extends Object
implements VertexInterface<V,E,M>

Vertex is a abstract definition of Google Pregel Vertex. For implementing a graph application, one must implement a sub-class of Vertex and define, the message passing and message processing for each vertex. Every vertex should be assigned an ID. This ID object should obey the equals-hashcode contract and would be used for partitioning. The edges for a vertex could be accessed and modified using the getEdges() call. The self value of the vertex could be changed by setValue(Writable).

Constructor Summary

Constructors

Constructor and Description
Vertex()

Method Summary

Methods

Modifier and Type	Method and Description
void	addEdge(Edge<V,E> edge)
void	addVertex(V vertexID, List<Edge<V,E>> edges, M value) Sends a message to add a new vertex through the partitioner to the appropriate BSP peer
void	aggregate(int index, M value) Provides a value to the specified aggregator.
int	compareTo(VertexInterface<V,E,M> o)
boolean	equals(Object obj)
M	getAggregatedValue(int index) Get the last aggregated value of the defined aggregator, null if nothing was configured or not returned a result.
HamaConfiguration	getConf()
String	getDestinationPeerName(Edge<V,E> edge)
String	getDestinationPeerName(V vertexId)
List<Edge<V,E>>	getEdges()
int	getMaxIteration()
org.apache.hadoop.io.IntWritable	getNumLastAggregatedVertices(int index) Get the number of aggregated vertices in the last superstep.
int	getNumPeers()
long	getNumVertices()
Partitioner<V,M>	getPartitioner()
BSPPeer<org.apache.hadoop.io.Writable,org.apache.hadoop.io.Writable,org.apache.hadoop.io.Writable,org.apache.hadoop.io.Writable,GraphJobMessage>	getPeer() Gives access to the BSP primitives and additional features by a peer.
protected GraphJobRunner<V,E,M>	getRunner()
long	getSuperstepCount()
M	getValue() Gets the vertex value
V	getVertexID()
int	hashCode()
boolean	isHalted()
void	readFields(DataInput in)
void	readState(DataInput in) Read the state of the vertex from the input stream.
void	remove() Removes current Vertex from local peer.
void	sendMessage(Edge<V,E> e, M msg) Sends a message to another vertex.
void	sendMessage(V destinationVertexID, M msg) Sends a message to the given destination vertex by ID and the message value
void	sendMessageToNeighbors(M msg) Sends a message to neighbors
void	setEdges(List<Edge<V,E>> list)
protected void	setRunner(GraphJobRunner<V,E,M> runner)
void	setup(HamaConfiguration conf) This method is called once before the Vertex computation begins.
void	setValue(M value) Sets the vertex value
void	setVertexID(V vertexID)
String	toString()
void	voteToHalt() Vote to halt.
void	write(DataOutput out)
void	writeState(DataOutput out) Writes the state of vertex to the output stream.

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Methods inherited from interface org.apache.hama.graph.VertexInterface

compute

Constructor Detail

Vertex

public Vertex()

Method Detail

getConf

public HamaConfiguration getConf()

getVertexID

public V getVertexID()

Specified by:

getVertexID in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Returns:

the unique identification for the vertex.

setup

public void setup(HamaConfiguration conf)

Description copied from interface: VertexInterface

This method is called once before the Vertex computation begins. Since the Vertex object is serializable, variables in your Vertex program always should be declared a s static.

Specified by:

setup in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

sendMessage

public void sendMessage(Edge<V,E> e,
               M msg)
                 throws IOException

Description copied from interface: VertexInterface

Sends a message to another vertex.

Specified by:

sendMessage in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Throws:

IOException

getDestinationPeerName

public String getDestinationPeerName(Edge<V,E> edge)

Returns:

the destination peer name of the destination of the given directed edge.

getDestinationPeerName

public String getDestinationPeerName(V vertexId)

Returns:

the destination peer name of the given vertex id, determined by the partitioner.

sendMessageToNeighbors

public void sendMessageToNeighbors(M msg)
                            throws IOException

Description copied from interface: VertexInterface

Sends a message to neighbors

Specified by:

sendMessageToNeighbors in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Throws:

IOException

sendMessage

public void sendMessage(V destinationVertexID,
               M msg)
                 throws IOException

Description copied from interface: VertexInterface

Sends a message to the given destination vertex by ID and the message value

Specified by:

sendMessage in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Throws:

IOException

addVertex

public void addVertex(V vertexID,
             List<Edge<V,E>> edges,
             M value)
               throws IOException

Description copied from interface: VertexInterface

Sends a message to add a new vertex through the partitioner to the appropriate BSP peer

Specified by:

addVertex in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Throws:

IOException

remove

public void remove()
            throws IOException

Description copied from interface: VertexInterface

Removes current Vertex from local peer.

Specified by:

remove in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Throws:

IOException

getSuperstepCount

public long getSuperstepCount()

Specified by:

getSuperstepCount in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Returns:

the superstep number of the current superstep (starting from 0).

setEdges

public void setEdges(List<Edge<V,E>> list)

addEdge

public void addEdge(Edge<V,E> edge)

getEdges

public List<Edge<V,E>> getEdges()

Specified by:

getEdges in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Returns:

a list of outgoing edges of this vertex in the input graph.

getValue

public M getValue()

Description copied from interface: VertexInterface

Gets the vertex value

Specified by:

getValue in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

setValue

public void setValue(M value)

Description copied from interface: VertexInterface

Sets the vertex value

Specified by:

setValue in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

setVertexID

public void setVertexID(V vertexID)

getMaxIteration

public int getMaxIteration()

getNumPeers

public int getNumPeers()

getPeer

public BSPPeer<org.apache.hadoop.io.Writable,org.apache.hadoop.io.Writable,org.apache.hadoop.io.Writable,org.apache.hadoop.io.Writable,GraphJobMessage> getPeer()

Gives access to the BSP primitives and additional features by a peer.

getPartitioner

public Partitioner<V,M> getPartitioner()

Returns:

the configured partitioner instance to message vertices.

getNumVertices

public long getNumVertices()

Specified by:

getNumVertices in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

Returns:

the number of vertices in the input graph.

voteToHalt

public void voteToHalt()

Description copied from interface: VertexInterface

Vote to halt.

Specified by:

voteToHalt in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

isHalted

public boolean isHalted()

hashCode

public int hashCode()

Overrides:

hashCode in class Object

equals

public boolean equals(Object obj)

Overrides:

equals in class Object

toString

public String toString()

Overrides:

toString in class Object

readFields

public void readFields(DataInput in)
                throws IOException

Specified by:

readFields in interface org.apache.hadoop.io.Writable

Throws:

IOException

write

public void write(DataOutput out)
           throws IOException

Specified by:

write in interface org.apache.hadoop.io.Writable

Throws:

IOException

compareTo

public final int compareTo(VertexInterface<V,E,M> o)

Specified by:

compareTo in interface Comparable<VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>>

readState

public void readState(DataInput in)
               throws IOException

Read the state of the vertex from the input stream. The framework would have already constructed and loaded the vertex-id, edges and voteToHalt state. This function is essential if there is any more properties of vertex to be read from.

Throws:

IOException

writeState

public void writeState(DataOutput out)
                throws IOException

Writes the state of vertex to the output stream. The framework writes the vertex and edge information to the output stream. This function could be used to save the state variable of the vertex added in the implementation of object.

Throws:

IOException

setRunner

protected void setRunner(GraphJobRunner<V,E,M> runner)

getRunner

protected GraphJobRunner<V,E,M> getRunner()

aggregate

public void aggregate(int index,
             M value)
               throws IOException

Description copied from interface: VertexInterface

Provides a value to the specified aggregator.

Specified by:

aggregate in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

value - value to be aggregated

Throws:

IOException

getAggregatedValue

public M getAggregatedValue(int index)

Get the last aggregated value of the defined aggregator, null if nothing was configured or not returned a result. You have to supply an index, the index is defined by the order you set the aggregator classes in GraphJob.setAggregatorClass(Class...). Index is starting at zero, so if you have a single aggregator you can retrieve it via #getLastAggregatedValue(0).

Specified by:

getAggregatedValue in interface VertexInterface<V extends org.apache.hadoop.io.WritableComparable,E extends org.apache.hadoop.io.Writable,M extends org.apache.hadoop.io.Writable>

getNumLastAggregatedVertices

public org.apache.hadoop.io.IntWritable getNumLastAggregatedVertices(int index)

Get the number of aggregated vertices in the last superstep. Or null if no aggregator is available.You have to supply an index, the index is defined by the order you set the aggregator classes in GraphJob.setAggregatorClass(Class...). Index is starting at zero, so if you have a single aggregator you can retrieve it via getNumLastAggregatedVertices(int)(0).