You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mahout.apache.org by ss...@apache.org on 2013/03/26 14:24:46 UTC
svn commit: r1461119 [2/4] - in /mahout/trunk: ./
buildtools/src/main/resources/
core/src/main/java/org/apache/mahout/cf/taste/hadoop/als/
core/src/main/java/org/apache/mahout/cf/taste/hadoop/item/
core/src/main/java/org/apache/mahout/cf/taste/hadoop/s...
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/clustering/meanshift/MeanShiftCanopyDriver.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/meanshift/MeanShiftCanopyDriver.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/clustering/meanshift/MeanShiftCanopyDriver.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/clustering/meanshift/MeanShiftCanopyDriver.java Tue Mar 26 13:24:40 2013
@@ -160,7 +160,7 @@ public class MeanShiftCanopyDriver exten
DistanceMeasure measure, IKernelProfile kernelProfile, double t1,
double t2, double convergenceDelta, int maxIterations,
boolean inputIsCanopies, boolean runClustering, boolean runSequential)
- throws IOException, InterruptedException, ClassNotFoundException {
+ throws IOException, InterruptedException, ClassNotFoundException {
Path clustersIn = new Path(output, Cluster.INITIAL_CLUSTERS_DIR);
if (inputIsCanopies) {
clustersIn = input;
@@ -172,9 +172,8 @@ public class MeanShiftCanopyDriver exten
kernelProfile, t1, t2, convergenceDelta, maxIterations, runSequential,
runClustering);
if (runClustering) {
- clusterData(inputIsCanopies ? input : new Path(output,
- Cluster.INITIAL_CLUSTERS_DIR), clustersOut, new Path(output,
- Cluster.CLUSTERED_POINTS_DIR), runSequential);
+ clusterData(inputIsCanopies ? input : new Path(output, Cluster.INITIAL_CLUSTERS_DIR), clustersOut,
+ new Path(output, Cluster.CLUSTERED_POINTS_DIR), runSequential);
}
}
@@ -183,7 +182,7 @@ public class MeanShiftCanopyDriver exten
*/
public static void createCanopyFromVectors(Configuration conf, Path input,
Path output, DistanceMeasure measure, boolean runSequential)
- throws IOException, InterruptedException, ClassNotFoundException {
+ throws IOException, InterruptedException, ClassNotFoundException {
if (runSequential) {
createCanopyFromVectorsSeq(input, output, measure);
} else {
@@ -212,13 +211,11 @@ public class MeanShiftCanopyDriver exten
SequenceFile.Writer writer = new SequenceFile.Writer(fs, conf, new Path(
output, "part-m-" + part++), Text.class, ClusterWritable.class);
try {
- for (VectorWritable value : new SequenceFileValueIterable<VectorWritable>(
- s.getPath(), conf)) {
- MeanShiftCanopy initialCanopy = MeanShiftCanopy.initialCanopy(value.get(),
- id++, measure);
+ for (VectorWritable value : new SequenceFileValueIterable<VectorWritable>(s.getPath(), conf)) {
+ MeanShiftCanopy initialCanopy = MeanShiftCanopy.initialCanopy(value.get(), id++, measure);
ClusterWritable clusterWritable = new ClusterWritable();
clusterWritable.setValue(initialCanopy);
- writer.append(new Text(), clusterWritable);
+ writer.append(new Text(), clusterWritable);
}
} finally {
Closeables.closeQuietly(writer);
@@ -360,8 +357,7 @@ public class MeanShiftCanopyDriver exten
int iteration = 1;
while (!converged && iteration <= maxIterations) {
int numReducers = Integer.valueOf(conf.get(MAPRED_REDUCE_TASKS, "1"));
- log.info("Mean Shift Iteration: {}, numReducers {}", new Object[] {
- iteration, numReducers });
+ log.info("Mean Shift Iteration: {}, numReducers {}", new Object[] { iteration, numReducers });
// point the output to a new directory per iteration
Path clustersOut = new Path(output, Cluster.CLUSTERS_DIR + iteration);
Path controlOut = new Path(output, CONTROL_CONVERGED);
@@ -469,14 +465,13 @@ public class MeanShiftCanopyDriver exten
/**
* Cluster the data sequentially
*/
- private static void clusterDataSeq(Path input, Path clustersIn, Path output)
- throws IOException {
+ private static void clusterDataSeq(Path input, Path clustersIn, Path output) throws IOException {
Collection<MeanShiftCanopy> clusters = Lists.newArrayList();
Configuration conf = new Configuration();
- for (ClusterWritable clusterWritable : new SequenceFileDirValueIterable<ClusterWritable>(
- clustersIn, PathType.LIST, PathFilters.logsCRCFilter(), conf)) {
+ for (ClusterWritable clusterWritable : new SequenceFileDirValueIterable<ClusterWritable>(clustersIn, PathType.LIST,
+ PathFilters.logsCRCFilter(), conf)) {
MeanShiftCanopy cluster = (MeanShiftCanopy) clusterWritable.getValue();
- clusters.add(cluster);
+ clusters.add(cluster);
}
// iterate over all points, assigning each to the closest canopy and
// outputting that clustering
@@ -488,14 +483,12 @@ public class MeanShiftCanopyDriver exten
output, "part-m-" + part++), IntWritable.class,
WeightedVectorWritable.class);
try {
- for (Pair<Writable, ClusterWritable> record : new SequenceFileIterable<Writable, ClusterWritable>(
- s.getPath(), conf)) {
+ for (Pair<Writable, ClusterWritable> record
+ : new SequenceFileIterable<Writable, ClusterWritable>(s.getPath(), conf)) {
ClusterWritable clusterWritable = record.getSecond();
- MeanShiftCanopy canopy = (MeanShiftCanopy) clusterWritable.getValue();
- MeanShiftCanopy closest = MeanShiftCanopyClusterer
- .findCoveringCanopy(canopy, clusters);
- writer.append(new IntWritable(closest.getId()),
- new WeightedVectorWritable(1, canopy.getCenter()));
+ MeanShiftCanopy canopy = (MeanShiftCanopy) clusterWritable.getValue();
+ MeanShiftCanopy closest = MeanShiftCanopyClusterer.findCoveringCanopy(canopy, clusters);
+ writer.append(new IntWritable(closest.getId()), new WeightedVectorWritable(1, canopy.getCenter()));
}
} finally {
Closeables.closeQuietly(writer);
@@ -507,7 +500,7 @@ public class MeanShiftCanopyDriver exten
* Cluster the data using Hadoop
*/
private static void clusterDataMR(Path input, Path clustersIn, Path output)
- throws IOException, InterruptedException, ClassNotFoundException {
+ throws IOException, InterruptedException, ClassNotFoundException {
Configuration conf = new Configuration();
conf.set(STATE_IN_KEY, clustersIn.toString());
Job job = new Job(conf,
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/clustering/minhash/HashFactory.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/minhash/HashFactory.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/clustering/minhash/HashFactory.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/clustering/minhash/HashFactory.java Tue Mar 26 13:24:40 2013
@@ -127,7 +127,7 @@ public final class HashFactory {
@Override
public int hash(byte[] bytes) {
- long hashValue = MurmurHash3.murmurhash3_x86_32(bytes, 0, bytes.length, seed);
+ long hashValue = MurmurHash3.murmurhash3x8632(bytes, 0, bytes.length, seed);
return Math.abs((int) (hashValue % RandomUtils.MAX_INT_SMALLER_TWIN_PRIME));
}
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/clustering/spectral/kmeans/SpectralKMeansDriver.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/spectral/kmeans/SpectralKMeansDriver.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/clustering/spectral/kmeans/SpectralKMeansDriver.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/clustering/spectral/kmeans/SpectralKMeansDriver.java Tue Mar 26 13:24:40 2013
@@ -62,7 +62,7 @@ public class SpectralKMeansDriver extend
@Override
public int run(String[] arg0)
- throws IOException, ClassNotFoundException, InstantiationException, IllegalAccessException, InterruptedException {
+ throws IOException, ClassNotFoundException, InstantiationException, IllegalAccessException, InterruptedException {
Configuration conf = getConf();
addInputOption();
@@ -99,15 +99,14 @@ public class SpectralKMeansDriver extend
Path tempdir = new Path(getOption("tempDir"));
boolean ssvd = parsedArgs.containsKey("--usessvd");
if (ssvd) {
- int reducers = Integer.parseInt(getOption("reduceTasks"));
- int blockheight = Integer.parseInt(getOption("outerProdBlockHeight"));
- int oversampling = Integer.parseInt(getOption("oversampling"));
- int poweriters = Integer.parseInt(getOption("powerIter"));
- run(conf, input, output, numDims, clusters, measure, convergenceDelta,
- maxIterations, tempdir, true, reducers, blockheight, oversampling, poweriters);
+ int reducers = Integer.parseInt(getOption("reduceTasks"));
+ int blockheight = Integer.parseInt(getOption("outerProdBlockHeight"));
+ int oversampling = Integer.parseInt(getOption("oversampling"));
+ int poweriters = Integer.parseInt(getOption("powerIter"));
+ run(conf, input, output, numDims, clusters, measure, convergenceDelta, maxIterations, tempdir, true, reducers,
+ blockheight, oversampling, poweriters);
} else {
- run(conf, input, output, numDims, clusters, measure, convergenceDelta,
- maxIterations, tempdir, false);
+ run(conf, input, output, numDims, clusters, measure, convergenceDelta, maxIterations, tempdir, false);
}
return 0;
@@ -124,8 +123,8 @@ public class SpectralKMeansDriver extend
int maxIterations,
Path tempDir,
boolean ssvd) throws IOException, InterruptedException, ClassNotFoundException {
- run(conf, input, output, numDims, clusters, measure, convergenceDelta,
- maxIterations, tempDir, ssvd, REDUCERS, BLOCKHEIGHT, OVERSAMPLING, POWERITERS);
+ run(conf, input, output, numDims, clusters, measure, convergenceDelta, maxIterations, tempDir, ssvd, REDUCERS,
+ BLOCKHEIGHT, OVERSAMPLING, POWERITERS);
}
/**
@@ -161,7 +160,7 @@ public class SpectralKMeansDriver extend
int blockHeight,
int oversampling,
int poweriters)
- throws IOException, InterruptedException, ClassNotFoundException {
+ throws IOException, InterruptedException, ClassNotFoundException {
Path outputCalc = new Path(tempDir, "calculations");
Path outputTmp = new Path(tempDir, "temporary");
@@ -264,7 +263,6 @@ public class SpectralKMeansDriver extend
// Run the KMeansDriver
Path answer = new Path(output, "kmeans_out");
- KMeansDriver.run(conf, data, initialclusters, answer,
- measure,convergenceDelta, maxIterations, true, 0.0, false);
- }
+ KMeansDriver.run(conf, data, initialclusters, answer, measure,convergenceDelta, maxIterations, true, 0.0, false);
+ }
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/clustering/topdown/postprocessor/ClusterOutputPostProcessor.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/topdown/postprocessor/ClusterOutputPostProcessor.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/clustering/topdown/postprocessor/ClusterOutputPostProcessor.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/clustering/topdown/postprocessor/ClusterOutputPostProcessor.java Tue Mar 26 13:24:40 2013
@@ -73,8 +73,8 @@ public final class ClusterOutputPostProc
*/
public void process() throws IOException {
createPostProcessDirectory();
- for (Pair<?,WeightedVectorWritable> record :
- new SequenceFileDirIterable<Writable,WeightedVectorWritable>(clusteredPoints,
+ for (Pair<?,WeightedVectorWritable> record
+ : new SequenceFileDirIterable<Writable,WeightedVectorWritable>(clusteredPoints,
PathType.GLOB,
PathFilters.partFilter(),
null,
@@ -91,8 +91,8 @@ public final class ClusterOutputPostProc
* Creates the directory to put post processed clusters.
*/
private void createPostProcessDirectory() throws IOException {
- if (!fileSystem.exists(clusterPostProcessorOutput) &&
- !fileSystem.mkdirs(clusterPostProcessorOutput)) {
+ if (!fileSystem.exists(clusterPostProcessorOutput)
+ && !fileSystem.mkdirs(clusterPostProcessorOutput)) {
throw new IOException("Error creating cluster post processor directory");
}
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/AbstractJob.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/AbstractJob.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/AbstractJob.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/AbstractJob.java Tue Mar 26 13:24:40 2013
@@ -498,8 +498,8 @@ public abstract class AbstractJob extend
// the option appeared on the command-line, or it has a value
// (which is likely a default value).
- if (cmdLine.hasOption(o) || cmdLine.getValue(o) != null ||
- (cmdLine.getValues(o) != null && !cmdLine.getValues(o).isEmpty())) {
+ if (cmdLine.hasOption(o) || cmdLine.getValue(o) != null
+ || (cmdLine.getValues(o) != null && !cmdLine.getValues(o).isEmpty())) {
// nulls are ok, for cases where options are simple flags.
List<?> vo = cmdLine.getValues(o);
@@ -601,7 +601,7 @@ public abstract class AbstractJob extend
* obsolete when MultipleInputs is available again
*/
public static void setS3SafeCombinedInputPath(Job job, Path referencePath, Path inputPathOne, Path inputPathTwo)
- throws IOException {
+ throws IOException {
FileSystem fs = FileSystem.get(referencePath.toUri(), job.getConfiguration());
FileInputFormat.setInputPaths(job, inputPathOne.makeQualified(fs), inputPathTwo.makeQualified(fs));
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/HadoopUtil.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/HadoopUtil.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/HadoopUtil.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/HadoopUtil.java Tue Mar 26 13:24:40 2013
@@ -71,7 +71,7 @@ public final class HadoopUtil {
Configuration jobConf = job.getConfiguration();
if (mapper.equals(Mapper.class)) {
- throw new IllegalStateException("Can't figure out the user class jar file from mapper/reducer");
+ throw new IllegalStateException("Can't figure out the user class jar file from mapper/reducer");
}
job.setJarByClass(mapper);
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/commandline/DefaultOptionCreator.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/commandline/DefaultOptionCreator.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/commandline/DefaultOptionCreator.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/commandline/DefaultOptionCreator.java Tue Mar 26 13:24:40 2013
@@ -240,7 +240,7 @@ public final class DefaultOptionCreator
/**
* @return a DefaultOptionBuilder for the clusterFilter option
*/
-public static DefaultOptionBuilder clusterFilterOption() {
+ public static DefaultOptionBuilder clusterFilterOption() {
return new DefaultOptionBuilder()
.withLongName(CLUSTER_FILTER_OPTION)
.withShortName("cf")
@@ -336,13 +336,10 @@ public static DefaultOptionBuilder clust
return new DefaultOptionBuilder()
.withLongName(ANALYZER_NAME_OPTION)
.withRequired(false)
- .withDescription(
- "If present, the name of a Lucene analyzer class to use")
- .withArgument(
- new ArgumentBuilder().withName(ANALYZER_NAME_OPTION).withDefault(StandardAnalyzer.class.getName())
- .withMinimum(1).withMaximum(1).create()
- )
- .withShortName("an");
+ .withDescription("If present, the name of a Lucene analyzer class to use")
+ .withArgument(new ArgumentBuilder().withName(ANALYZER_NAME_OPTION).withDefault(StandardAnalyzer.class.getName())
+ .withMinimum(1).withMaximum(1).create())
+ .withShortName("an");
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/StableFixedSizeSamplingIterator.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/StableFixedSizeSamplingIterator.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/StableFixedSizeSamplingIterator.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/StableFixedSizeSamplingIterator.java Tue Mar 26 13:24:40 2013
@@ -56,12 +56,12 @@ public class StableFixedSizeSamplingIter
Collections.sort(buf);
delegate = Iterators.transform(buf.iterator(),
- new Function<Pair<Integer,T>,T>() {
- @Override
- public T apply(Pair<Integer,T> from) {
- return from.getSecond();
- }
- });
+ new Function<Pair<Integer,T>,T>() {
+ @Override
+ public T apply(Pair<Integer,T> from) {
+ return from.getSecond();
+ }
+ });
}
@Override
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirIterator.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirIterator.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirIterator.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirIterator.java Tue Mar 26 13:24:40 2013
@@ -103,23 +103,19 @@ public final class SequenceFileDirIterat
Iterator<Iterator<Pair<K, V>>> fsIterators =
Iterators.transform(fileStatusIterator,
- new Function<FileStatus, Iterator<Pair<K, V>>>() {
- @Override
- public Iterator<Pair<K, V>> apply(FileStatus from) {
- try {
- SequenceFileIterator<K, V> iterator =
- new SequenceFileIterator<K, V>(from.getPath(),
- reuseKeyValueInstances,
- conf);
- iterators.add(iterator);
- return iterator;
- } catch (IOException ioe) {
- throw new IllegalStateException(from.getPath()
- .toString(),
- ioe);
- }
- }
- });
+ new Function<FileStatus, Iterator<Pair<K, V>>>() {
+ @Override
+ public Iterator<Pair<K, V>> apply(FileStatus from) {
+ try {
+ SequenceFileIterator<K, V> iterator = new SequenceFileIterator<K, V>(from.getPath(),
+ reuseKeyValueInstances, conf);
+ iterators.add(iterator);
+ return iterator;
+ } catch (IOException ioe) {
+ throw new IllegalStateException(from.getPath().toString(), ioe);
+ }
+ }
+ });
Collections.reverse(iterators); // close later in reverse order
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirValueIterator.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirValueIterator.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirValueIterator.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/iterator/sequencefile/SequenceFileDirValueIterator.java Tue Mar 26 13:24:40 2013
@@ -117,23 +117,19 @@ public final class SequenceFileDirValueI
Iterator<Iterator<V>> fsIterators =
Iterators.transform(fileStatusIterator,
- new Function<FileStatus, Iterator<V>>() {
- @Override
- public Iterator<V> apply(FileStatus from) {
- try {
- SequenceFileValueIterator<V> iterator =
- new SequenceFileValueIterator<V>(from.getPath(),
- reuseKeyValueInstances,
- conf);
- iterators.add(iterator);
- return iterator;
- } catch (IOException ioe) {
- throw new IllegalStateException(from.getPath()
- .toString(),
- ioe);
- }
- }
- });
+ new Function<FileStatus, Iterator<V>>() {
+ @Override
+ public Iterator<V> apply(FileStatus from) {
+ try {
+ SequenceFileValueIterator<V> iterator = new SequenceFileValueIterator<V>(from.getPath(),
+ reuseKeyValueInstances, conf);
+ iterators.add(iterator);
+ return iterator;
+ } catch (IOException ioe) {
+ throw new IllegalStateException(from.getPath().toString(), ioe);
+ }
+ }
+ });
Collections.reverse(iterators); // close later in reverse order
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsCombiner.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsCombiner.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsCombiner.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsCombiner.java Tue Mar 26 13:24:40 2013
@@ -28,7 +28,7 @@ public class MergeVectorsCombiner
@Override
public void reduce(WritableComparable<?> key, Iterable<VectorWritable> vectors, Context ctx)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
ctx.write(key, VectorWritable.merge(vectors.iterator()));
}
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsReducer.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsReducer.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsReducer.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/common/mapreduce/MergeVectorsReducer.java Tue Mar 26 13:24:40 2013
@@ -32,7 +32,7 @@ public class MergeVectorsReducer extends
@Override
public void reduce(WritableComparable<?> key, Iterable<VectorWritable> vectors, Context ctx)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
Vector merged = VectorWritable.merge(vectors.iterator()).get();
result.set(new SequentialAccessSparseVector(merged));
ctx.write(key, result);
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/fpm/pfpgrowth/convertors/TransactionIterator.java Tue Mar 26 13:24:40 2013
@@ -43,8 +43,8 @@ public class TransactionIterator<T> exte
@Override
public Pair<int[],Long> apply(Pair<List<T>,Long> from) {
if (from == null) {
- return null;
- }
+ return null;
+ }
int index = 0;
for (T attribute : from.getFirst()) {
if (attributeIdMapping.containsKey(attribute)) {
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/MatrixUtils.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/MatrixUtils.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/MatrixUtils.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/MatrixUtils.java Tue Mar 26 13:24:40 2013
@@ -36,7 +36,7 @@ public final class MatrixUtils {
}
public static void write(Path outputDir, Configuration conf, VectorIterable matrix)
- throws IOException {
+ throws IOException {
FileSystem fs = outputDir.getFileSystem(conf);
fs.delete(outputDir, true);
SequenceFile.Writer writer = SequenceFile.createWriter(fs, conf, outputDir,
@@ -57,8 +57,8 @@ public final class MatrixUtils {
boolean sparse = false;
List<Pair<Integer, Vector>> rows = Lists.newArrayList();
for (Path modelPath : modelPaths) {
- for (Pair<IntWritable, VectorWritable> row :
- new SequenceFileIterable<IntWritable, VectorWritable>(modelPath, true, conf)) {
+ for (Pair<IntWritable, VectorWritable> row
+ : new SequenceFileIterable<IntWritable, VectorWritable>(modelPath, true, conf)) {
rows.add(Pair.of(row.getFirst().get(), row.getSecond().get()));
numRows = Math.max(numRows, row.getFirst().get());
sparse = !row.getSecond().get().isDense();
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/decomposer/HdfsBackedLanczosState.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/decomposer/HdfsBackedLanczosState.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/decomposer/HdfsBackedLanczosState.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/decomposer/HdfsBackedLanczosState.java Tue Mar 26 13:24:40 2013
@@ -109,8 +109,8 @@ public class HdfsBackedLanczosState exte
nextBasisVectorPath = new Path(basisPath, BASIS_PREFIX + '_' + ++numBasisVectorsOnDisk);
}
Vector nextVector;
- while (numBasisVectorsOnDisk < iterationNumber &&
- (nextVector = getBasisVector(numBasisVectorsOnDisk)) != null) {
+ while (numBasisVectorsOnDisk < iterationNumber
+ && (nextVector = getBasisVector(numBasisVectorsOnDisk)) != null) {
persistVector(nextBasisVectorPath, numBasisVectorsOnDisk, nextVector);
nextBasisVectorPath = new Path(basisPath, BASIS_PREFIX + '_' + ++numBasisVectorsOnDisk);
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/SeedVectorUtil.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/SeedVectorUtil.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/SeedVectorUtil.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/SeedVectorUtil.java Tue Mar 26 13:24:40 2013
@@ -51,8 +51,8 @@ final class SeedVectorUtil {
List<NamedVector> seedVectors = Lists.newArrayList();
long item = 0;
- for (Writable value :
- new SequenceFileDirValueIterable<Writable>(new Path(seedPathStr),
+ for (Writable value
+ : new SequenceFileDirValueIterable<Writable>(new Path(seedPathStr),
PathType.LIST,
PathFilters.partFilter(),
conf)) {
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceMapper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceMapper.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceMapper.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceMapper.java Tue Mar 26 13:24:40 2013
@@ -53,10 +53,10 @@ public final class VectorDistanceMapper
for (NamedVector seedVector : seedVectors) {
double distance = measure.distance(seedVector, valVec);
if (!usesThreshold || distance <= maxDistance) {
- StringTuple outKey = new StringTuple();
- outKey.add(seedVector.getName());
- outKey.add(keyName);
- context.write(outKey, new DoubleWritable(distance));
+ StringTuple outKey = new StringTuple();
+ outKey.add(seedVector.getName());
+ outKey.add(keyName);
+ context.write(outKey, new DoubleWritable(distance));
}
}
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceSimilarityJob.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceSimilarityJob.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceSimilarityJob.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/VectorDistanceSimilarityJob.java Tue Mar 26 13:24:40 2013
@@ -63,11 +63,11 @@ public class VectorDistanceSimilarityJob
addOutputOption();
addOption(DefaultOptionCreator.distanceMeasureOption().create());
addOption(SEEDS, "s", "The set of vectors to compute distances against. Must fit in memory on the mapper");
- addOption(MAX_DISTANCE, "mx", "set an upper-bound on distance (double) such that any pair of vectors with a" +
- " distance greater than this value is ignored in the output. Ignored for non pairwise output!");
+ addOption(MAX_DISTANCE, "mx", "set an upper-bound on distance (double) such that any pair of vectors with a"
+ + " distance greater than this value is ignored in the output. Ignored for non pairwise output!");
addOption(DefaultOptionCreator.overwriteOption().create());
- addOption(OUT_TYPE_KEY, "ot", "[pw|v] -- Define the output style: pairwise, the default, (pw) or vector (v). " +
- "Pairwise is a tuple of <seed, other, distance>, vector is <other, <Vector of size the number of seeds>>.",
+ addOption(OUT_TYPE_KEY, "ot", "[pw|v] -- Define the output style: pairwise, the default, (pw) or vector (v). "
+ + "Pairwise is a tuple of <seed, other, distance>, vector is <other, <Vector of size the number of seeds>>.",
"pw");
if (parseArguments(args) == null) {
@@ -110,7 +110,7 @@ public class VectorDistanceSimilarityJob
Path output,
DistanceMeasure measure, String outType)
throws IOException, ClassNotFoundException, InterruptedException {
- run(conf, input, seeds, output, measure, outType, null);
+ run(conf, input, seeds, output, measure, outType, null);
}
public static void run(Configuration conf,
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/RowSimilarityJob.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/RowSimilarityJob.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/RowSimilarityJob.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/RowSimilarityJob.java Tue Mar 26 13:24:40 2013
@@ -120,8 +120,8 @@ public class RowSimilarityJob extends Ab
int maxSimilaritiesPerRow = Integer.parseInt(getOption("maxSimilaritiesPerRow"));
boolean excludeSelfSimilarity = Boolean.parseBoolean(getOption("excludeSelfSimilarity"));
- double threshold = hasOption("threshold") ?
- Double.parseDouble(getOption("threshold")) : NO_THRESHOLD;
+ double threshold = hasOption("threshold")
+ ? Double.parseDouble(getOption("threshold")) : NO_THRESHOLD;
Path weightsPath = getTempPath("weights");
Path normsPath = getTempPath("norms.bin");
@@ -445,7 +445,7 @@ public class RowSimilarityJob extends Ab
@Override
protected void reduce(IntWritable row, Iterable<VectorWritable> partials, Context ctx)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
Vector allSimilarities = Vectors.merge(partials);
Vector topKSimilarities = Vectors.topKElements(maxSimilaritiesPerRow, allSimilarities);
ctx.write(row, new VectorWritable(topKSimilarities));
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/CosineSimilarity.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/CosineSimilarity.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/CosineSimilarity.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/CosineSimilarity.java Tue Mar 26 13:24:40 2013
@@ -44,7 +44,7 @@ public class CosineSimilarity implements
@Override
public boolean consider(int numNonZeroEntriesA, int numNonZeroEntriesB, double maxValueA, double maxValueB,
double threshold) {
- return numNonZeroEntriesB >= threshold / maxValueA &&
- numNonZeroEntriesA >= threshold / maxValueB;
+ return numNonZeroEntriesB >= threshold / maxValueA
+ && numNonZeroEntriesA >= threshold / maxValueB;
}
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/TanimotoCoefficientSimilarity.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/TanimotoCoefficientSimilarity.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/TanimotoCoefficientSimilarity.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/similarity/cooccurrence/measures/TanimotoCoefficientSimilarity.java Tue Mar 26 13:24:40 2013
@@ -28,7 +28,7 @@ public class TanimotoCoefficientSimilari
@Override
public boolean consider(int numNonZeroEntriesA, int numNonZeroEntriesB, double maxValueA, double maxValueB,
double threshold) {
- return numNonZeroEntriesA >= numNonZeroEntriesB * threshold &&
- numNonZeroEntriesB >= numNonZeroEntriesA * threshold;
+ return numNonZeroEntriesA >= numNonZeroEntriesB * threshold
+ && numNonZeroEntriesB >= numNonZeroEntriesA * threshold;
}
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/BasicStats.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/BasicStats.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/BasicStats.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/BasicStats.java Tue Mar 26 13:24:40 2013
@@ -49,10 +49,9 @@ public final class BasicStats {
* @return The variance (based on sample estimation)
*/
public static double variance(Path input, Path output,
- Configuration baseConf) throws IOException, InterruptedException,
- ClassNotFoundException {
-
- VarianceTotals varianceTotals = computeVarianceTotals(input, output, baseConf);
+ Configuration baseConf)
+ throws IOException, InterruptedException, ClassNotFoundException {
+ VarianceTotals varianceTotals = computeVarianceTotals(input, output, baseConf);
return varianceTotals.computeVariance();
}
@@ -66,10 +65,9 @@ public final class BasicStats {
* @return The variance (based on sample estimation)
*/
public static double varianceForGivenMean(Path input, Path output, double mean,
- Configuration baseConf) throws IOException, InterruptedException,
- ClassNotFoundException {
-
- VarianceTotals varianceTotals = computeVarianceTotals(input, output, baseConf);
+ Configuration baseConf)
+ throws IOException, InterruptedException, ClassNotFoundException {
+ VarianceTotals varianceTotals = computeVarianceTotals(input, output, baseConf);
return varianceTotals.computeVarianceForGivenMean(mean);
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/StandardDeviationCalculatorMapper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/StandardDeviationCalculatorMapper.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/StandardDeviationCalculatorMapper.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stats/StandardDeviationCalculatorMapper.java Tue Mar 26 13:24:40 2013
@@ -33,7 +33,7 @@ public class StandardDeviationCalculator
@Override
protected void map(IntWritable key, Writable value, Context context)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
if (key.get() == -1) {
return;
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/ABtDenseOutJob.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/ABtDenseOutJob.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/ABtDenseOutJob.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/ABtDenseOutJob.java Tue Mar 26 13:24:40 2013
@@ -347,8 +347,7 @@ public final class ABtDenseOutJob {
*
*/
public static class QRReducer
- extends
- Reducer<SplitPartitionedWritable, DenseBlockWritable, SplitPartitionedWritable, VectorWritable> {
+ extends Reducer<SplitPartitionedWritable, DenseBlockWritable, SplitPartitionedWritable, VectorWritable> {
/*
* HACK: partition number formats in hadoop, copied. this may stop working
@@ -476,12 +475,10 @@ public final class ABtDenseOutJob {
* @param <K>
* bogus
*/
- private <K, V> OutputCollector<K, V>
- createOutputCollector(String name,
+ private <K, V> OutputCollector<K, V> createOutputCollector(String name,
final SplitPartitionedWritable spw,
Context ctx,
- Class<V> valueClass) throws IOException,
- InterruptedException {
+ Class<V> valueClass) throws IOException, InterruptedException {
Path outputPath = getSplitFilePath(name, spw, ctx);
final SequenceFile.Writer w =
SequenceFile.createWriter(FileSystem.get(outputPath.toUri(), ctx.getConfiguration()),
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/BtJob.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/BtJob.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/BtJob.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/BtJob.java Tue Mar 26 13:24:40 2013
@@ -298,7 +298,7 @@ public final class BtJob {
}
@SuppressWarnings("unchecked")
- private void outputQRow(Writable key, Vector qRow, Vector aRow ) throws IOException {
+ private void outputQRow(Writable key, Vector qRow, Vector aRow) throws IOException {
if (nv && (aRow instanceof NamedVector)) {
qRowValue.set(new NamedVector(qRow, ((NamedVector) aRow).getName()));
} else {
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/VJob.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/VJob.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/VJob.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/hadoop/stochasticsvd/VJob.java Tue Mar 26 13:24:40 2013
@@ -167,7 +167,7 @@ public class VJob {
Path outputPath,
int k,
int numReduceTasks,
- SSVDSolver.OutputScalingEnum outputScaling ) throws ClassNotFoundException,
+ SSVDSolver.OutputScalingEnum outputScaling) throws ClassNotFoundException,
InterruptedException, IOException {
job = new Job(conf);
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyMapper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyMapper.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyMapper.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyMapper.java Tue Mar 26 13:24:40 2013
@@ -33,7 +33,7 @@ public class SpecificConditionalEntropyM
@Override
protected void map(StringTuple key, VarIntWritable value, Context context)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
resultKey.set(key.stringAt(0));
context.write(resultKey, value);
}
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyReducer.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyReducer.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyReducer.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/SpecificConditionalEntropyReducer.java Tue Mar 26 13:24:40 2013
@@ -47,7 +47,7 @@ public final class SpecificConditionalEn
@Override
protected void reduce(Text key, Iterable<VarIntWritable> values, Context context)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
double sum = 0.0;
double entropy = 0.0;
for (VarIntWritable value : values) {
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/VarIntSumReducer.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/VarIntSumReducer.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/VarIntSumReducer.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/math/stats/entropy/VarIntSumReducer.java Tue Mar 26 13:24:40 2013
@@ -32,7 +32,7 @@ public final class VarIntSumReducer exte
@Override
protected void reduce(Writable key, Iterable<VarIntWritable> values, Context context)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
int sum = 0;
for (VarIntWritable value : values) {
sum += value.get();
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/EncodedVectorsFromSequenceFiles.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/EncodedVectorsFromSequenceFiles.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/EncodedVectorsFromSequenceFiles.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/EncodedVectorsFromSequenceFiles.java Tue Mar 26 13:24:40 2013
@@ -43,16 +43,17 @@ public final class EncodedVectorsFromSeq
addOutputOption();
addOption(DefaultOptionCreator.analyzerOption().create());
addOption(buildOption("sequentialAccessVector", "seq",
- "(Optional) Whether output vectors should be SequentialAccessVectors. " +
- "If set true else false",
+ "(Optional) Whether output vectors should be SequentialAccessVectors. "
+ + "If set true else false",
false, false, null));
addOption(buildOption("namedVector", "nv",
"Create named vectors using the key. False by default", false, false, null));
addOption("cardinality", "c",
"The cardinality to use for creating the vectors. Default is 5000", "5000");
addOption("encoderFieldName", "en",
- "The name of the encoder to be passed to the FeatureVectorEncoder constructor. Default is text. " +
- "Note this is not the class name of a FeatureValueEncoder, but is instead the construction argument.",
+ "The name of the encoder to be passed to the FeatureVectorEncoder constructor. Default is text. "
+ + "Note this is not the class name of a FeatureValueEncoder, but is instead the construction "
+ + "argument.",
"text");
addOption("encoderClass", "ec",
"The class name of the encoder to be used. Default is " + LuceneTextValueEncoder.class.getName(),
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/pruner/WordsPrunerReducer.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/pruner/WordsPrunerReducer.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/pruner/WordsPrunerReducer.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/vectorizer/pruner/WordsPrunerReducer.java Tue Mar 26 13:24:40 2013
@@ -43,7 +43,7 @@ public class WordsPrunerReducer extends
@Override
protected void reduce(WritableComparable<?> key, Iterable<VectorWritable> values, Context context)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
Iterator<VectorWritable> it = values.iterator();
if (!it.hasNext()) {
return;
@@ -81,8 +81,8 @@ public class WordsPrunerReducer extends
Path dictionaryFile = new Path(localFiles[0].getPath());
// key is feature, value is the document frequency
- for (Pair<IntWritable, LongWritable> record :
- new SequenceFileIterable<IntWritable, LongWritable>(dictionaryFile, true, conf)) {
+ for (Pair<IntWritable, LongWritable> record
+ : new SequenceFileIterable<IntWritable, LongWritable>(dictionaryFile, true, conf)) {
dictionary.put(record.getFirst().get(), record.getSecond().get());
}
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToPrefsDriver.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToPrefsDriver.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToPrefsDriver.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToPrefsDriver.java Tue Mar 26 13:24:40 2013
@@ -85,12 +85,12 @@ public final class MailToPrefsDriver ext
addOption("chunkSize", "cs", "The size of chunks to write. Default is 100 mb", "100");
addOption("separator", "sep", "The separator used in the input file to separate to, from, subject. Default is \\n",
"\n");
- addOption("from", "f", "The position in the input text (value) where the from email is located, starting from " +
- "zero (0).", "0");
- addOption("refs", "r", "The position in the input text (value) where the reference ids are located, " +
- "starting from zero (0).", "1");
- addOption(buildOption("useCounts", "u", "If set, then use the number of times the user has interacted with a " +
- "thread as an indication of their preference. Otherwise, use boolean preferences.", false, false,
+ addOption("from", "f", "The position in the input text (value) where the from email is located, starting from "
+ + "zero (0).", "0");
+ addOption("refs", "r", "The position in the input text (value) where the reference ids are located, "
+ + "starting from zero (0).", "1");
+ addOption(buildOption("useCounts", "u", "If set, then use the number of times the user has interacted with a "
+ + "thread as an indication of their preference. Otherwise, use boolean preferences.", false, false,
String.valueOf(true)));
Map<String, List<String>> parsedArgs = parseArguments(args);
@@ -231,7 +231,7 @@ public final class MailToPrefsDriver ext
String name,
Configuration baseConf,
int chunkSizeInMegabytes, int[] maxTermDimension)
- throws IOException {
+ throws IOException {
List<Path> chunkPaths = Lists.newArrayList();
Configuration conf = new Configuration(baseConf);
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToRecReducer.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToRecReducer.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToRecReducer.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToRecReducer.java Tue Mar 26 13:24:40 2013
@@ -39,7 +39,7 @@ public class MailToRecReducer extends Re
@Override
protected void reduce(Text key, Iterable<LongWritable> values, Context context)
- throws IOException, InterruptedException {
+ throws IOException, InterruptedException {
if (useCounts) {
long sum = 0;
for (LongWritable value : values) {
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/Track1RecommenderEvaluator.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/Track1RecommenderEvaluator.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/Track1RecommenderEvaluator.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/Track1RecommenderEvaluator.java Tue Mar 26 13:24:40 2013
@@ -71,8 +71,8 @@ public final class Track1RecommenderEval
Collection<Callable<Void>> estimateCallables = Lists.newArrayList();
AtomicInteger noEstimateCounter = new AtomicInteger();
- for (Pair<PreferenceArray,long[]> userData :
- new DataFileIterable(KDDCupDataModel.getValidationFile(dataFileDirectory))) {
+ for (Pair<PreferenceArray,long[]> userData
+ : new DataFileIterable(KDDCupDataModel.getValidationFile(dataFileDirectory))) {
PreferenceArray validationPrefs = userData.getFirst();
long userID = validationPrefs.get(0).getUserID();
estimateCallables.add(
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/KDDCupFactorizablePreferences.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/KDDCupFactorizablePreferences.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/KDDCupFactorizablePreferences.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/KDDCupFactorizablePreferences.java Tue Mar 26 13:24:40 2013
@@ -50,12 +50,12 @@ public class KDDCupFactorizablePreferenc
public Iterable<Preference> getPreferences() {
Iterable<Iterable<Preference>> prefIterators =
Iterables.transform(new DataFileIterable(dataFile),
- new Function<Pair<PreferenceArray,long[]>,Iterable<Preference>>() {
- @Override
- public Iterable<Preference> apply(Pair<PreferenceArray,long[]> from) {
- return from.getFirst();
- }
- });
+ new Function<Pair<PreferenceArray,long[]>,Iterable<Preference>>() {
+ @Override
+ public Iterable<Preference> apply(Pair<PreferenceArray,long[]> from) {
+ return from.getFirst();
+ }
+ });
return Iterables.concat(prefIterators);
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/ParallelArraysSGDFactorizer.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/ParallelArraysSGDFactorizer.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/ParallelArraysSGDFactorizer.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/ParallelArraysSGDFactorizer.java Tue Mar 26 13:24:40 2013
@@ -171,7 +171,7 @@ public class ParallelArraysSGDFactorizer
for (int feature = 0; feature < numFeatures; feature++) {
log.info("Shuffling preferences...");
shufflePreferences();
- log.info("Starting training of feature {} ...", feature);
+ log.info("Starting training of feature {} ...", feature);
for (int currentIteration = 0; currentIteration < numIterations; currentIteration++) {
if (currentIteration == numIterations - 1) {
double rmse = trainingIterationWithRmse(feature);
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/Track1SVDRunner.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/Track1SVDRunner.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/Track1SVDRunner.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/kddcup/track1/svd/Track1SVDRunner.java Tue Mar 26 13:24:40 2013
@@ -84,8 +84,8 @@ public final class Track1SVDRunner {
log.info("Estimating validation preferences...");
int prefsProcessed = 0;
RunningAverage average = new FullRunningAverage();
- for (Pair<PreferenceArray,long[]> validationPair :
- new DataFileIterable(KDDCupDataModel.getValidationFile(dataFileDirectory))) {
+ for (Pair<PreferenceArray,long[]> validationPair
+ : new DataFileIterable(KDDCupDataModel.getValidationFile(dataFileDirectory))) {
for (Preference validationPref : validationPair.getFirst()) {
double estimate = estimatePreference(factorization, validationPref.getUserID(), validationPref.getItemID(),
factorizablePreferences.getMinPreference(), factorizablePreferences.getMaxPreference());
@@ -107,8 +107,8 @@ public final class Track1SVDRunner {
try {
out = new BufferedOutputStream(new FileOutputStream(resultFile));
- for (Pair<PreferenceArray,long[]> testPair :
- new DataFileIterable(KDDCupDataModel.getTestFile(dataFileDirectory))) {
+ for (Pair<PreferenceArray,long[]> testPair
+ : new DataFileIterable(KDDCupDataModel.getTestFile(dataFileDirectory))) {
for (Preference testPref : testPair.getFirst()) {
double estimate = estimatePreference(factorization, testPref.getUserID(), testPref.getItemID(),
factorizablePreferences.getMinPreference(), factorizablePreferences.getMaxPreference());
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/hadoop/example/als/netflix/NetflixDatasetConverter.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/hadoop/example/als/netflix/NetflixDatasetConverter.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/hadoop/example/als/netflix/NetflixDatasetConverter.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/hadoop/example/als/netflix/NetflixDatasetConverter.java Tue Mar 26 13:24:40 2013
@@ -55,8 +55,8 @@ public final class NetflixDatasetConvert
public static void main(String[] args) throws IOException {
if (args.length != 4) {
- System.err.println("Usage: NetflixDatasetConverter /path/to/training_set/ /path/to/qualifying.txt " +
- "/path/to/judging.txt /path/to/destination");
+ System.err.println("Usage: NetflixDatasetConverter /path/to/training_set/ /path/to/qualifying.txt "
+ + "/path/to/judging.txt /path/to/destination");
return;
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/similarity/precompute/example/BatchItemSimilaritiesGroupLens.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/similarity/precompute/example/BatchItemSimilaritiesGroupLens.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/similarity/precompute/example/BatchItemSimilaritiesGroupLens.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/similarity/precompute/example/BatchItemSimilaritiesGroupLens.java Tue Mar 26 13:24:40 2013
@@ -59,8 +59,8 @@ public final class BatchItemSimilarities
int numSimilarities = batch.computeItemSimilarities(Runtime.getRuntime().availableProcessors(), 1,
new FileSimilarItemsWriter(resultFile));
- System.out.println("Computed " + numSimilarities + " similarities for " + dataModel.getNumItems() + " items " +
- "and saved them to " + resultFile.getAbsolutePath());
+ System.out.println("Computed " + numSimilarities + " similarities for " + dataModel.getNumItems() + " items "
+ + "and saved them to " + resultFile.getAbsolutePath());
}
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/NewsgroupHelper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/NewsgroupHelper.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/NewsgroupHelper.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/NewsgroupHelper.java Tue Mar 26 13:24:40 2013
@@ -48,10 +48,11 @@ import java.util.Random;
public final class NewsgroupHelper {
private static final SimpleDateFormat[] DATE_FORMATS = {
- new SimpleDateFormat("", Locale.ENGLISH),
- new SimpleDateFormat("MMM-yyyy", Locale.ENGLISH),
- new SimpleDateFormat("dd-MMM-yyyy HH:mm:ss", Locale.ENGLISH)
+ new SimpleDateFormat("", Locale.ENGLISH),
+ new SimpleDateFormat("MMM-yyyy", Locale.ENGLISH),
+ new SimpleDateFormat("dd-MMM-yyyy HH:mm:ss", Locale.ENGLISH)
};
+
public static final int FEATURES = 10000;
// 1997-01-15 00:01:00 GMT
private static final long DATE_REFERENCE = 853286460;
@@ -87,8 +88,8 @@ public final class NewsgroupHelper {
countWords(analyzer, words, dateString, overallCounts);
while (line != null && !line.isEmpty()) {
boolean countHeader = (
- line.startsWith("From:") || line.startsWith("Subject:") ||
- line.startsWith("Keywords:") || line.startsWith("Summary:")) && leakType < 6;
+ line.startsWith("From:") || line.startsWith("Subject:")
+ || line.startsWith("Keywords:") || line.startsWith("Summary:")) && leakType < 6;
do {
Reader in = new StringReader(line);
if (countHeader) {
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/df/mapreduce/BuildForest.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/df/mapreduce/BuildForest.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/df/mapreduce/BuildForest.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/df/mapreduce/BuildForest.java Tue Mar 26 13:24:40 2013
@@ -91,8 +91,8 @@ public class BuildForest extends Configu
Option selectionOpt = obuilder.withLongName("selection").withShortName("sl").withRequired(false)
.withArgument(abuilder.withName("m").withMinimum(1).withMaximum(1).create())
.withDescription("Optional, Number of variables to select randomly at each tree-node.\n"
- + "For classification problem, the default is square root of the number of explanatory variables.\n" +
- "For regression problem, the default is 1/3 of the number of explanatory variables.").create();
+ + "For classification problem, the default is square root of the number of explanatory variables.\n"
+ + "For regression problem, the default is 1/3 of the number of explanatory variables.").create();
Option noCompleteOpt = obuilder.withLongName("no-complete").withShortName("nc").withRequired(false)
.withDescription("Optional, The tree is not complemented").create();
@@ -105,8 +105,8 @@ public class BuildForest extends Configu
Option minPropOpt = obuilder.withLongName("minprop").withShortName("mp").withRequired(false)
.withArgument(abuilder.withName("minprop").withMinimum(1).withMaximum(1).create())
.withDescription("Optional, The tree-node is not divided, if the proportion of the "
- + "variance of branching data is smaller than this value.\n" +
- "In the case of a regression problem, this value is used. "
+ + "variance of branching data is smaller than this value.\n"
+ + "In the case of a regression problem, this value is used. "
+ "The default is 1/1000(0.001).").create();
Option seedOpt = obuilder.withLongName("seed").withShortName("sd").withRequired(false)
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/email/PrepEmailVectorsDriver.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/email/PrepEmailVectorsDriver.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/email/PrepEmailVectorsDriver.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/email/PrepEmailVectorsDriver.java Tue Mar 26 13:24:40 2013
@@ -51,10 +51,10 @@ public class PrepEmailVectorsDriver exte
addInputOption();
addOutputOption();
addOption(DefaultOptionCreator.overwriteOption().create());
- addOption("maxItemsPerLabel", "mipl", "The maximum number of items per label. Can be useful for making the " +
- "training sets the same size", String.valueOf(100000));
- addOption(buildOption("useListName", "ul", "Use the name of the list as part of the label. If not set, then " +
- "just use the project name", false, false, "false"));
+ addOption("maxItemsPerLabel", "mipl", "The maximum number of items per label. Can be useful for making the "
+ + "training sets the same size", String.valueOf(100000));
+ addOption(buildOption("useListName", "ul", "Use the name of the list as part of the label. If not set, then "
+ + "just use the project name", false, false, "false"));
Map<String,List<String>> parsedArgs = parseArguments(args);
if (parsedArgs == null) {
return -1;
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/SimpleCsvExamples.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/SimpleCsvExamples.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/SimpleCsvExamples.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/SimpleCsvExamples.java Tue Mar 26 13:24:40 2013
@@ -142,7 +142,7 @@ public final class SimpleCsvExamples {
private static final Splitter ON_TABS = Splitter.on(SEPARATOR_CHAR).trimResults();
public static final Joiner WITH_COMMAS = Joiner.on(SEPARATOR_CHAR);
- public static final Random rand = RandomUtils.getRandom();
+ public static final Random RAND = RandomUtils.getRandom();
private final List<String> data;
@@ -180,7 +180,7 @@ public final class SimpleCsvExamples {
* @return
*/
private static int randomValue(double mean) {
- return (int) (-mean * Math.log1p(-rand.nextDouble()));
+ return (int) (-mean * Math.log1p(-RAND.nextDouble()));
}
@Override
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/TrainAdaptiveLogistic.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/TrainAdaptiveLogistic.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/TrainAdaptiveLogistic.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/sgd/TrainAdaptiveLogistic.java Tue Mar 26 13:24:40 2013
@@ -229,13 +229,11 @@ public final class TrainAdaptiveLogistic
.create();
Option targetCategories = builder
- .withLongName("categories")
- .withDescription("the number of target categories to be considered")
- .withRequired(true)
- .withArgument(
- argumentBuilder.withName("categories").withMaximum(1)
- .create())
- .create();
+ .withLongName("categories")
+ .withDescription("the number of target categories to be considered")
+ .withRequired(true)
+ .withArgument(argumentBuilder.withName("categories").withMaximum(1).create())
+ .create();
Option features = builder
@@ -373,7 +371,7 @@ public final class TrainAdaptiveLogistic
return lmp;
}
- static BufferedReader open(String inputFile) throws IOException {
+ static BufferedReader open(String inputFile) throws IOException {
InputStream in;
try {
in = Resources.getResource(inputFile).openStream();
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayClustering.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayClustering.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayClustering.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayClustering.java Tue Mar 26 13:24:40 2013
@@ -75,8 +75,8 @@ public class DisplayClustering extends F
protected static final List<List<Cluster>> CLUSTERS = Lists.newArrayList();
- static final Color[] COLORS = {Color.red, Color.orange, Color.yellow, Color.green, Color.blue, Color.magenta,
- Color.lightGray};
+ static final Color[] COLORS = { Color.red, Color.orange, Color.yellow, Color.green, Color.blue, Color.magenta,
+ Color.lightGray };
protected static final double T1 = 3.0;
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayDirichlet.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayDirichlet.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayDirichlet.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayDirichlet.java Tue Mar 26 13:24:40 2013
@@ -80,7 +80,7 @@ public class DisplayDirichlet extends Di
private static void runSequentialDirichletClassifier(Path input, Path output,
ModelDistribution<VectorWritable> modelDist, int numClusters, int numIterations, double alpha0)
- throws IOException {
+ throws IOException {
List<Cluster> models = Lists.newArrayList();
for (Model<VectorWritable> cluster : modelDist.sampleFromPrior(numClusters)) {
models.add((Cluster) cluster);
@@ -94,7 +94,7 @@ public class DisplayDirichlet extends Di
private static void runSequentialDirichletClusterer(Path input, Path output,
ModelDistribution<VectorWritable> modelDist, int numClusters, int numIterations, double alpha0)
- throws IOException, ClassNotFoundException, InterruptedException {
+ throws IOException, ClassNotFoundException, InterruptedException {
DistributionDescription description = new DistributionDescription(modelDist.getClass().getName(),
RandomAccessSparseVector.class.getName(), ManhattanDistanceMeasure.class.getName(), 2);
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayMinHash.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayMinHash.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayMinHash.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/display/DisplayMinHash.java Tue Mar 26 13:24:40 2013
@@ -363,8 +363,8 @@ public class DisplayMinHash extends Disp
private static void runMinHash(Configuration conf, Path samples, Path output) throws Exception {
ToolRunner.run(conf, new MinHashDriver(), new String[] { "--input", samples.toString(),
- "--hashType", HashFactory.HashType.MURMUR3.toString(), "--output", output.toString(),
- "--minVectorSize", "1", "--debugOutput"
+ "--hashType", HashFactory.HashType.MURMUR3.toString(), "--output", output.toString(),
+ "--minVectorSize", "1", "--debugOutput"
});
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/meanshift/Job.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/meanshift/Job.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/meanshift/Job.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/meanshift/Job.java Tue Mar 26 13:24:40 2013
@@ -133,7 +133,7 @@ public final class Job extends AbstractJ
double t2,
double convergenceDelta,
int maxIterations)
- throws Exception {
+ throws Exception {
Path directoryContainingConvertedInput = new Path(output,
DIRECTORY_CONTAINING_CONVERTED_INPUT);
InputDriver.runJob(input, directoryContainingConvertedInput);
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/cassandra/CassandraDataModel.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/cassandra/CassandraDataModel.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/cassandra/CassandraDataModel.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/cassandra/CassandraDataModel.java Tue Mar 26 13:24:40 2013
@@ -246,9 +246,9 @@ public final class CassandraDataModel im
public int getNumUsersWithPreferenceFor(long itemID1, long itemID2) throws TasteException {
FastIDSet userIDs1 = userIDsFromItemCache.get(itemID1);
FastIDSet userIDs2 = userIDsFromItemCache.get(itemID2);
- return userIDs1.size() < userIDs2.size() ?
- userIDs2.intersectionSize(userIDs1) :
- userIDs1.intersectionSize(userIDs2);
+ return userIDs1.size() < userIDs2.size()
+ ? userIDs2.intersectionSize(userIDs1)
+ : userIDs1.intersectionSize(userIDs2);
}
@Override
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/hbase/HBaseDataModel.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/hbase/HBaseDataModel.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/hbase/HBaseDataModel.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/hbase/HBaseDataModel.java Tue Mar 26 13:24:40 2013
@@ -95,7 +95,7 @@ public final class HBaseDataModel implem
private final HTablePool pool;
private final String tableName;
- public final boolean tableWasCreated;
+ private final boolean tableWasCreated;
// Cache of user and item ids
private volatile FastIDSet itemIDs;
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/mongodb/MongoDBDataModel.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/mongodb/MongoDBDataModel.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/mongodb/MongoDBDataModel.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/cf/taste/impl/model/mongodb/MongoDBDataModel.java Tue Mar 26 13:24:40 2013
@@ -577,8 +577,8 @@ public final class MongoDBDataModel impl
userIDPrefMap.put(userID, userPrefs);
}
userPrefs.add(new GenericPreference(userID, itemID, ratingValue));
- if (user.containsKey("created_at") &&
- mongoTimestamp.compareTo(getDate(user.get("created_at"))) < 0) {
+ if (user.containsKey("created_at")
+ && mongoTimestamp.compareTo(getDate(user.get("created_at"))) < 0) {
mongoTimestamp = getDate(user.get("created_at"));
}
}
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/classifier/ConfusionMatrixDumper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/classifier/ConfusionMatrixDumper.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/classifier/ConfusionMatrixDumper.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/classifier/ConfusionMatrixDumper.java Tue Mar 26 13:24:40 2013
@@ -64,18 +64,18 @@ public final class ConfusionMatrixDumper
+ "}\n"
+ "th.normalHeader\n"
+ "{\n"
- + "border:1px solid black;border-collapse:collapse;text-align:center;" +
- "background-color:white\n"
+ + "border:1px solid black;border-collapse:collapse;text-align:center;"
+ + "background-color:white\n"
+ "}\n"
+ "th.tallHeader\n"
+ "{\n"
- + "border:1px solid black;border-collapse:collapse;text-align:center;" +
- "background-color:white; height:6em\n"
+ + "border:1px solid black;border-collapse:collapse;text-align:center;"
+ + "background-color:white; height:6em\n"
+ "}\n"
+ "tr.label\n"
+ "{\n"
- + "border:1px solid black;border-collapse:collapse;text-align:center;" +
- "background-color:white\n"
+ + "border:1px solid black;border-collapse:collapse;text-align:center;"
+ + "background-color:white\n"
+ "}\n"
+ "tr.row\n"
+ "{\n"
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/evaluation/RepresentativePointsDriver.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/evaluation/RepresentativePointsDriver.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/evaluation/RepresentativePointsDriver.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/evaluation/RepresentativePointsDriver.java Tue Mar 26 13:24:40 2013
@@ -184,8 +184,8 @@ public final class RepresentativePointsD
Map<Integer,WeightedVectorWritable> mostDistantPoints = Maps.newHashMap();
FileSystem fs = FileSystem.get(clusteredPointsIn.toUri(), conf);
for (Pair<IntWritable,WeightedVectorWritable> record
- : new SequenceFileDirIterable<IntWritable,WeightedVectorWritable>(
- clusteredPointsIn, PathType.LIST, PathFilters.logsCRCFilter(), null, true, conf)) {
+ : new SequenceFileDirIterable<IntWritable,WeightedVectorWritable>(clusteredPointsIn, PathType.LIST,
+ PathFilters.logsCRCFilter(), null, true, conf)) {
RepresentativePointsMapper.mapPoint(record.getFirst(), record.getSecond(), measure, repPoints, mostDistantPoints);
}
int part = 0;
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/lda/LDAPrintTopics.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/lda/LDAPrintTopics.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/lda/LDAPrintTopics.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/clustering/lda/LDAPrintTopics.java Tue Mar 26 13:24:40 2013
@@ -180,7 +180,7 @@ public final class LDAPrintTopics {
});
for (Pair<String,Double> wordWithScore : topKasList) {
out.write(wordWithScore.getFirst() + " [p(" + wordWithScore.getFirst() + "|topic_" + i + ") = "
- + wordWithScore.getSecond());
+ + wordWithScore.getSecond());
out.write('\n');
}
} finally {
@@ -199,9 +199,9 @@ public final class LDAPrintTopics {
int numWordsToPrint) {
List<Queue<Pair<String,Double>>> queues = Lists.newArrayList();
Map<Integer,Double> expSums = Maps.newHashMap();
- for (Pair<IntPairWritable,DoubleWritable> record :
- new SequenceFileDirIterable<IntPairWritable, DoubleWritable>(
- new Path(dir, "part-*"), PathType.GLOB, null, null, true, job)) {
+ for (Pair<IntPairWritable,DoubleWritable> record
+ : new SequenceFileDirIterable<IntPairWritable, DoubleWritable>(
+ new Path(dir, "part-*"), PathType.GLOB, null, null, true, job)) {
IntPairWritable key = record.getFirst();
int topic = key.getFirst();
int word = key.getSecond();
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/text/MailArchivesClusteringAnalyzer.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/text/MailArchivesClusteringAnalyzer.java?rev=1461119&r1=1461118&r2=1461119&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/text/MailArchivesClusteringAnalyzer.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/text/MailArchivesClusteringAnalyzer.java Tue Mar 26 13:24:40 2013
@@ -47,7 +47,7 @@ public final class MailArchivesClusterin
// HTML tags, and Java keywords asmany of the messages in the archives
// are subversion check-in notifications
- private static final CharArraySet stopSet = new CharArraySet(LUCENE_VERSION, Arrays.asList(
+ private static final CharArraySet STOP_SET = new CharArraySet(LUCENE_VERSION, Arrays.asList(
"3d","7bit","a0","about","above","abstract","across","additional","after",
"afterwards","again","against","align","all","almost","alone","along",
"already","also","although","always","am","among","amongst","amoungst",
@@ -103,11 +103,11 @@ public final class MailArchivesClusterin
), false);
// Regex used to exclude non-alpha-numeric tokens
- private static final Pattern alphaNumeric = Pattern.compile("^[a-z][a-z0-9_]+$");
- private static final Matcher matcher = alphaNumeric.matcher("");
+ private static final Pattern ALPHA_NUMERIC = Pattern.compile("^[a-z][a-z0-9_]+$");
+ private static final Matcher MATCHER = ALPHA_NUMERIC.matcher("");
public MailArchivesClusteringAnalyzer() {
- super(LUCENE_VERSION, stopSet);
+ super(LUCENE_VERSION, STOP_SET);
}
public MailArchivesClusteringAnalyzer(CharArraySet stopSet) {
@@ -122,7 +122,7 @@ public final class MailArchivesClusterin
result = new LowerCaseFilter(LUCENE_VERSION, result);
result = new ASCIIFoldingFilter(result);
result = new AlphaNumericMaxLengthFilter(result);
- result = new StopFilter(LUCENE_VERSION, result, stopSet);
+ result = new StopFilter(LUCENE_VERSION, result, STOP_SET);
result = new PorterStemFilter(result);
return new TokenStreamComponents(tokenizer, result);
}
@@ -154,8 +154,8 @@ public final class MailArchivesClusterin
}
}
String term = new String(output, 0, at);
- matcher.reset(term);
- if (matcher.matches() && !term.startsWith("a0")) {
+ MATCHER.reset(term);
+ if (MATCHER.matches() && !term.startsWith("a0")) {
termAtt.setEmpty();
termAtt.append(term);
return true;