org.apache.mahout.math.hadoop

Class MatrixColumnMeansJob.MatrixColumnMeansMapper

java.lang.Object

org.apache.hadoop.mapreduce.Mapper<org.apache.hadoop.io.Writable,VectorWritable,org.apache.hadoop.io.NullWritable,VectorWritable>

org.apache.mahout.math.hadoop.MatrixColumnMeansJob.MatrixColumnMeansMapper

Enclosing class:

MatrixColumnMeansJob

public static class MatrixColumnMeansJob.MatrixColumnMeansMapper
extends org.apache.hadoop.mapreduce.Mapper<org.apache.hadoop.io.Writable,VectorWritable,org.apache.hadoop.io.NullWritable,VectorWritable>

Mapper for calculation of column-wise mean.

Nested Class Summary

Nested classes/interfaces inherited from class org.apache.hadoop.mapreduce.Mapper

org.apache.hadoop.mapreduce.Mapper.Context

Constructor Summary

Constructors

Constructor and Description
MatrixColumnMeansMapper()

Method Summary

Modifier and Type	Method and Description
void	cleanup(org.apache.hadoop.mapreduce.Mapper.Context context) The column-wise sum is written at the cleanup stage.
void	map(org.apache.hadoop.io.Writable r, VectorWritable v, org.apache.hadoop.mapreduce.Mapper.Context context) The mapper computes a running sum of the vectors the task has seen.
void	setup(org.apache.hadoop.mapreduce.Mapper.Context context)

Methods inherited from class org.apache.hadoop.mapreduce.Mapper

run

Methods inherited from class java.lang.Object

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

Constructor Detail

MatrixColumnMeansMapper

public MatrixColumnMeansMapper()

Method Detail

setup

public void setup(org.apache.hadoop.mapreduce.Mapper.Context context)

Overrides:

setup in class org.apache.hadoop.mapreduce.Mapper<org.apache.hadoop.io.Writable,VectorWritable,org.apache.hadoop.io.NullWritable,VectorWritable>

map

public void map(org.apache.hadoop.io.Writable r,
                VectorWritable v,
                org.apache.hadoop.mapreduce.Mapper.Context context)
         throws IOException

The mapper computes a running sum of the vectors the task has seen. Element 0 of the running sum vector contains a count of the number of vectors that have been seen. The remaining elements contain the column-wise running sum. Nothing is written at this stage

Overrides:

map in class org.apache.hadoop.mapreduce.Mapper<org.apache.hadoop.io.Writable,VectorWritable,org.apache.hadoop.io.NullWritable,VectorWritable>

Throws:

IOException

cleanup

public void cleanup(org.apache.hadoop.mapreduce.Mapper.Context context)
             throws InterruptedException,
                    IOException

The column-wise sum is written at the cleanup stage. A single reducer is forced so null can be used for the key

Overrides:

cleanup in class org.apache.hadoop.mapreduce.Mapper<org.apache.hadoop.io.Writable,VectorWritable,org.apache.hadoop.io.NullWritable,VectorWritable>

Throws:

InterruptedException

IOException