org.apache.mahout.clustering.spectral

Class VectorMatrixMultiplicationJob

java.lang.Object

org.apache.mahout.clustering.spectral.VectorMatrixMultiplicationJob

public final class VectorMatrixMultiplicationJob
extends Object

This class handles the three-way multiplication of the digonal matrix and the Markov transition matrix inherent in the Eigencuts algorithm. The equation takes the form:

W = D^(1/2) * M * D^(1/2)

Since the diagonal matrix D has only n non-zero elements, it is represented as a dense vector in this job, rather than a full n-by-n matrix. This job performs the multiplications and returns the new DRM.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	VectorMatrixMultiplicationJob.VectorMatrixMultiplicationMapper

Method Summary

Modifier and Type	Method and Description
static DistributedRowMatrix	runJob(org.apache.hadoop.fs.Path markovPath, Vector diag, org.apache.hadoop.fs.Path outputPath) Invokes the job.
static DistributedRowMatrix	runJob(org.apache.hadoop.fs.Path markovPath, Vector diag, org.apache.hadoop.fs.Path outputPath, org.apache.hadoop.fs.Path tmpPath)

Methods inherited from class java.lang.Object

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

Method Detail

runJob

public static DistributedRowMatrix runJob(org.apache.hadoop.fs.Path markovPath,
                                          Vector diag,
                                          org.apache.hadoop.fs.Path outputPath)
                                   throws IOException,
                                          ClassNotFoundException,
                                          InterruptedException

Invokes the job.

Parameters:

markovPath - Path to the markov DRM's sequence files

Throws:

IOException

ClassNotFoundException

InterruptedException

runJob

public static DistributedRowMatrix runJob(org.apache.hadoop.fs.Path markovPath,
                                          Vector diag,
                                          org.apache.hadoop.fs.Path outputPath,
                                          org.apache.hadoop.fs.Path tmpPath)
                                   throws IOException,
                                          ClassNotFoundException,
                                          InterruptedException

Throws:

IOException

ClassNotFoundException

InterruptedException