You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mahout.apache.org by gs...@apache.org on 2012/02/13 16:14:43 UTC
svn commit: r1243557 - in /mahout/trunk:
core/src/main/java/org/apache/mahout/clustering/
core/src/main/java/org/apache/mahout/common/
examples/src/main/java/org/apache/mahout/cf/taste/example/email/
examples/src/main/java/org/apache/mahout/classifier/...
Author: gsingers
Date: Mon Feb 13 15:14:42 2012
New Revision: 1243557
URL: http://svn.apache.org/viewvc?rev=1243557&view=rev
Log:
MAHOUT-947: add in support for multiple options
Modified:
mahout/trunk/core/src/main/java/org/apache/mahout/clustering/WeightedPropertyVectorWritable.java
mahout/trunk/core/src/main/java/org/apache/mahout/common/AbstractJob.java
mahout/trunk/examples/src/main/java/org/apache/mahout/cf/taste/example/email/MailToPrefsDriver.java
mahout/trunk/examples/src/main/java/org/apache/mahout/classifier/email/PrepEmailVectorsDriver.java
mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/canopy/Job.java
mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/dirichlet/Job.java
mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/fuzzykmeans/Job.java
mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/kmeans/Job.java
mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/meanshift/Job.java
mahout/trunk/integration/src/main/java/org/apache/mahout/classifier/ConfusionMatrixDumper.java
mahout/trunk/integration/src/main/java/org/apache/mahout/utils/MatrixDumper.java
mahout/trunk/integration/src/main/java/org/apache/mahout/utils/SequenceFileDumper.java
mahout/trunk/integration/src/main/java/org/apache/mahout/utils/clustering/ClusterDumper.java
mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/RowIdJob.java
mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/VectorDumper.java
Modified: mahout/trunk/core/src/main/java/org/apache/mahout/clustering/WeightedPropertyVectorWritable.java
URL: http://svn.apache.org/viewvc/mahout/trunk/core/src/main/java/org/apache/mahout/clustering/WeightedPropertyVectorWritable.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/core/src/main/java/org/apache/mahout/clustering/WeightedPropertyVectorWritable.java (original)
+++ mahout/trunk/core/src/main/java/org/apache/mahout/clustering/WeightedPropertyVectorWritable.java Mon Feb 13 15:14:42 2012
@@ -79,7 +79,7 @@ public class WeightedPropertyVectorWrita
@Override
public String toString() {
Vector vector = getVector();
- StringBuilder bldr = new StringBuilder("wt: ").append(getWeight());
+ StringBuilder bldr = new StringBuilder("wt: ").append(getWeight()).append(" ");
if (properties != null && !properties.isEmpty()) {
for (Map.Entry<Text, Text> entry : properties.entrySet()) {
bldr.append(entry.getKey().toString()).append(": ").append(entry.getValue().toString()).append(' ');
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=1243557&r1=1243556&r2=1243557&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 Mon Feb 13 15:14:42 2012
@@ -94,17 +94,17 @@ public abstract class AbstractJob extend
private Option outputOption;
/** input path, populated by {@link #parseArguments(String[])} */
- private Path inputPath;
- private File inputFile;//the input represented as a file
+ protected Path inputPath;
+ protected File inputFile;//the input represented as a file
/** output path, populated by {@link #parseArguments(String[]) */
- private Path outputPath;
- private File outputFile;//the output represented as a file
+ protected Path outputPath;
+ protected File outputFile;//the output represented as a file
/** temp path, populated by {@link #parseArguments(String[]) */
- private Path tempPath;
+ protected Path tempPath;
- private Map<String, List<String>> argMap;
+ protected Map<String, List<String>> argMap;
/** internal list of options that have been added */
private final List<Option> options;
@@ -308,10 +308,22 @@ public abstract class AbstractJob extend
* argument values can be retrieved using {@code get(optionName)}. The
* names used for keys are the option name parameter prefixed by '--'.
*
+ * @see #parseArguments(String[], boolean, boolean) -- passes in false, false for the optional args.
*
*/
public Map<String, List<String>> parseArguments(String[] args) throws IOException {
+ return parseArguments(args, false, false);
+ }
+ /**
+ *
+ * @param args The args to parse
+ * @param inputOptional if false, then the input option, if set, need not be present. If true and input is an option and there is no input, then throw an error
+ * @param outputOptional if false, then the output option, if set, need not be present. If true and output is an option and there is no output, then throw an error
+ * @return the args parsed into a map.
+ * @throws IOException
+ */
+ public Map<String, List<String>> parseArguments(String[] args, boolean inputOptional, boolean outputOptional) throws IOException{
Option helpOpt = addOption(DefaultOptionCreator.helpOption());
addOption("tempDir", null, "Intermediate output directory", "temp");
addOption("startPhase", null, "First phase to run", "0");
@@ -344,7 +356,7 @@ public abstract class AbstractJob extend
}
try {
- parseDirectories(cmdLine);
+ parseDirectories(cmdLine, inputOptional, outputOptional);
} catch (IllegalArgumentException e) {
log.error(e.getMessage());
CommandLineUtil.printHelpWithGenericOptions(group);
@@ -424,7 +436,7 @@ public abstract class AbstractJob extend
* specified or outputOption is present and neither {@code --output}
* nor {@code -Dmapred.output.dir} are specified.
*/
- protected void parseDirectories(CommandLine cmdLine) {
+ protected void parseDirectories(CommandLine cmdLine, boolean inputOptional, boolean outputOptional) {
Configuration conf = getConf();
@@ -444,9 +456,9 @@ public abstract class AbstractJob extend
this.outputPath = new Path(conf.get("mapred.output.dir"));
}
- Preconditions.checkArgument(inputOption == null || inputPath != null,
+ Preconditions.checkArgument(inputOptional == true || inputOption == null || inputPath != null,
"No input specified or -Dmapred.input.dir must be provided to specify input directory");
- Preconditions.checkArgument(outputOption == null || outputPath != null,
+ Preconditions.checkArgument(outputOptional == true || outputOption == null || outputPath != null,
"No output specified: or -Dmapred.output.dir must be provided to specify output directory");
}
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=1243557&r1=1243556&r2=1243557&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 Mon Feb 13 15:14:42 2012
@@ -87,18 +87,18 @@ public final class MailToPrefsDriver ext
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, "true"));
- Map<String, String> parsedArgs = parseArguments(args);
+ Map<String, List<String>> parsedArgs = parseArguments(args);
Path input = getInputPath();
Path output = getOutputPath();
- int chunkSize = Integer.parseInt(parsedArgs.get("--chunkSize"));
- String separator = parsedArgs.get("--separator");
+ int chunkSize = Integer.parseInt(getOption("chunkSize"));
+ String separator = getOption("separator");
Configuration conf = getConf();
if (conf == null) {
setConf(new Configuration());
conf = getConf();
}
- boolean useCounts = hasOption("--useCounts");
+ boolean useCounts = hasOption("useCounts");
AtomicInteger currentPhase = new AtomicInteger();
int[] msgDim = new int[1];
//TODO: mod this to not do so many passes over the data. Dictionary creation could probably be a chain mapper
@@ -170,8 +170,8 @@ public final class MailToPrefsDriver ext
conf.set(EmailUtility.MSG_ID_DIMENSION, String.valueOf(msgDim[0]));
conf.set(EmailUtility.FROM_PREFIX, "fromIds-dictionary-");
conf.set(EmailUtility.MSG_IDS_PREFIX, "msgIds-dictionary-");
- conf.set(EmailUtility.FROM_INDEX, parsedArgs.get("--from"));
- conf.set(EmailUtility.REFS_INDEX, parsedArgs.get("--refs"));
+ conf.set(EmailUtility.FROM_INDEX, getOption("from"));
+ conf.set(EmailUtility.REFS_INDEX, getOption("refs"));
conf.set(EmailUtility.SEPARATOR, separator);
conf.set(MailToRecReducer.USE_COUNTS_PREFERENCE, String.valueOf(useCounts));
int j = 0;
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=1243557&r1=1243556&r2=1243557&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 Mon Feb 13 15:14:42 2012
@@ -29,6 +29,7 @@ import org.apache.mahout.common.HadoopUt
import org.apache.mahout.common.commandline.DefaultOptionCreator;
import org.apache.mahout.math.VectorWritable;
+import java.util.List;
import java.util.Map;
/**
@@ -52,7 +53,7 @@ public class PrepEmailVectorsDriver exte
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"));
- Map<String,String> parsedArgs = parseArguments(args);
+ Map<String,List<String>> parsedArgs = parseArguments(args);
if (parsedArgs == null) {
return -1;
}
@@ -64,8 +65,8 @@ public class PrepEmailVectorsDriver exte
}
Job convertJob = prepareJob(input, output, SequenceFileInputFormat.class, PrepEmailMapper.class,
Text.class, VectorWritable.class, PrepEmailReducer.class, Text.class, VectorWritable.class, SequenceFileOutputFormat.class);
- convertJob.getConfiguration().set(ITEMS_PER_CLASS, parsedArgs.get("--maxItemsPerLabel"));
- convertJob.getConfiguration().set(USE_LIST_NAME, String.valueOf(parsedArgs.containsKey("--useListName")));
+ convertJob.getConfiguration().set(ITEMS_PER_CLASS, getOption("maxItemsPerLabel"));
+ convertJob.getConfiguration().set(USE_LIST_NAME, String.valueOf(hasOption("useListName")));
boolean succeeded = convertJob.waitForCompletion(true);
return succeeded ? 0 : -1;
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/canopy/Job.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/canopy/Job.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/canopy/Job.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/canopy/Job.java Mon Feb 13 15:14:42 2012
@@ -17,6 +17,7 @@
package org.apache.mahout.clustering.syntheticcontrol.canopy;
+import java.util.List;
import java.util.Map;
import org.apache.hadoop.conf.Configuration;
@@ -101,7 +102,7 @@ public final class Job extends AbstractJ
addOption(DefaultOptionCreator.t2Option().create());
addOption(DefaultOptionCreator.overwriteOption().create());
- Map<String, String> argMap = parseArguments(args);
+ Map<String, List<String>> argMap = parseArguments(args);
if (argMap == null) {
return -1;
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/dirichlet/Job.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/dirichlet/Job.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/dirichlet/Job.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/dirichlet/Job.java Mon Feb 13 15:14:42 2012
@@ -17,6 +17,7 @@
package org.apache.mahout.clustering.syntheticcontrol.dirichlet;
+import java.util.List;
import java.util.Map;
import org.apache.commons.cli2.builder.ArgumentBuilder;
@@ -87,7 +88,7 @@ public final class Job extends AbstractJ
addOption(DefaultOptionCreator.emitMostLikelyOption().create());
addOption(DefaultOptionCreator.thresholdOption().create());
- Map<String, String> argMap = parseArguments(args);
+ Map<String, List<String>> argMap = parseArguments(args);
if (argMap == null) {
return -1;
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/fuzzykmeans/Job.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/fuzzykmeans/Job.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/fuzzykmeans/Job.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/fuzzykmeans/Job.java Mon Feb 13 15:14:42 2012
@@ -18,6 +18,7 @@
package org.apache.mahout.clustering.syntheticcontrol.fuzzykmeans;
import java.io.IOException;
+import java.util.List;
import java.util.Map;
import org.apache.commons.cli2.builder.ArgumentBuilder;
@@ -79,7 +80,7 @@ public final class Job extends AbstractJ
addOption(M_OPTION, M_OPTION,
"coefficient normalization factor, must be greater than 1", true);
- Map<String, String> argMap = parseArguments(args);
+ Map<String, List<String>> argMap = parseArguments(args);
if (argMap == null) {
return -1;
}
Modified: mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/kmeans/Job.java
URL: http://svn.apache.org/viewvc/mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/kmeans/Job.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/kmeans/Job.java (original)
+++ mahout/trunk/examples/src/main/java/org/apache/mahout/clustering/syntheticcontrol/kmeans/Job.java Mon Feb 13 15:14:42 2012
@@ -18,6 +18,7 @@
package org.apache.mahout.clustering.syntheticcontrol.kmeans;
import java.io.IOException;
+import java.util.List;
import java.util.Map;
import org.apache.hadoop.conf.Configuration;
@@ -74,7 +75,7 @@ public final class Job extends AbstractJ
addOption(DefaultOptionCreator.maxIterationsOption().create());
addOption(DefaultOptionCreator.overwriteOption().create());
- Map<String, String> argMap = parseArguments(args);
+ Map<String, List<String>> argMap = parseArguments(args);
if (argMap == null) {
return -1;
}
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=1243557&r1=1243556&r2=1243557&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 Mon Feb 13 15:14:42 2012
@@ -17,6 +17,7 @@
package org.apache.mahout.clustering.syntheticcontrol.meanshift;
+import java.util.List;
import java.util.Map;
import org.apache.hadoop.conf.Configuration;
@@ -73,7 +74,7 @@ public final class Job extends AbstractJ
addOption(DefaultOptionCreator.t2Option().create());
addOption(DefaultOptionCreator.clusteringOption().create());
- Map<String,String> argMap = parseArguments(args);
+ Map<String,List<String>> argMap = parseArguments(args);
if (argMap == null) {
return -1;
}
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=1243557&r1=1243556&r2=1243557&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 Mon Feb 13 15:14:42 2012
@@ -139,13 +139,13 @@ public final class ConfusionMatrixDumper
addOption(DefaultOptionCreator.overwriteOption().create());
addFlag("html", null, "Create complete HTML page");
addFlag("text", null, "Dump simple text");
- Map<String,String> parsedArgs = parseArguments(args);
+ Map<String,List<String>> parsedArgs = parseArguments(args);
if (parsedArgs == null) {
return -1;
}
Path inputPath = getInputPath();
- String outputFile = parsedArgs.containsKey("--output") ? parsedArgs.get("--output") : null;
+ String outputFile = hasOption("output") ? getOption("output") : null;
boolean text = parsedArgs.containsKey("--text");
boolean wrapHtml = parsedArgs.containsKey("--html");
PrintStream out = getPrintStream(outputFile);
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/utils/MatrixDumper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/utils/MatrixDumper.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/utils/MatrixDumper.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/utils/MatrixDumper.java Mon Feb 13 15:14:42 2012
@@ -22,6 +22,7 @@ import java.io.FileOutputStream;
import java.io.IOException;
import java.io.OutputStream;
import java.io.PrintStream;
+import java.util.List;
import java.util.Map;
import org.apache.hadoop.conf.Configuration;
@@ -55,11 +56,11 @@ public final class MatrixDumper extends
addInputOption();
addOption("output", "o", "Output path", null); // AbstractJob output feature requires param
- Map<String, String> parsedArgs = parseArguments(args);
+ Map<String, List<String>> parsedArgs = parseArguments(args);
if (parsedArgs == null) {
return -1;
}
- String outputFile = parsedArgs.containsKey("--output") ? parsedArgs.get("--output") : null;
+ String outputFile = hasOption("output") ? getOption("output") : null;
exportCSV(getInputPath(), outputFile, false);
return 0;
}
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/utils/SequenceFileDumper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/utils/SequenceFileDumper.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/utils/SequenceFileDumper.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/utils/SequenceFileDumper.java Mon Feb 13 15:14:42 2012
@@ -50,33 +50,27 @@ public final class SequenceFileDumper ex
@Override
public int run(String[] args) throws Exception {
- addOption("seqFile", "s", "The Sequence File to read in", false);
- addOption("seqDirectory", "d", "A directory containing sequence files to read", false);
- addOption(DefaultOptionCreator.outputOption().create());
+ addInputOption();
+ addOutputOption();
addOption("substring", "b", "The number of chars to print out per value", false);
addOption(buildOption("count", "c", "Report the count only", false, false, null));
addOption("numItems", "n", "Output at most <n> key value pairs", false);
addOption(buildOption("facets", "fa", "Output the counts per key. Note, if there are a lot of unique keys, this can take up a fair amount of memory", false, false, null));
addOption(buildOption("quiet", "q", "Print only file contents.", false, false, null));
- if (parseArguments(args) == null) {
+ if (parseArguments(args, false, true) == null) {
return -1;
}
Path[] pathArr= null;
Configuration conf = new Configuration();
-
- if (getOption("seqFile") != null) {
+ Path input = getInputPath();
+ FileSystem fs = input.getFileSystem(conf);
+ if (fs.getFileStatus(input).isDir()) {
+ pathArr = FileUtil.stat2Paths(fs.listStatus(input, new OutputFilesFilter()));
+ } else {
pathArr = new Path[1];
- pathArr[0] = new Path(getOption("seqFile"));
- } else if (getOption("seqDirectory") != null) {
- Path dirPath = new Path(getOption("seqDirectory"));
- FileSystem fs = dirPath.getFileSystem(conf);
- pathArr = FileUtil.stat2Paths(fs.listStatus(dirPath, new OutputFilesFilter()));
- }
- if (pathArr == null) {
- System.out.println("Must specify --seqFile (-s) or --seqDirectory (-d)!");
- return -1;
+ pathArr[0] = input;
}
@@ -91,8 +85,9 @@ public final class SequenceFileDumper ex
}
try {
for (Path path : pathArr) {
- if (!hasOption("quiet"))
+ if (!hasOption("quiet")){
writer.append("Input Path: ").append(String.valueOf(path)).append('\n');
+ }
int sub = Integer.MAX_VALUE;
if (hasOption("substring")) {
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/utils/clustering/ClusterDumper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/utils/clustering/ClusterDumper.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/utils/clustering/ClusterDumper.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/utils/clustering/ClusterDumper.java Mon Feb 13 15:14:42 2012
@@ -81,7 +81,6 @@ public final class ClusterDumper extends
private long maxPointsPerCluster = Long.MAX_VALUE;
private String termDictionary;
private String dictionaryFormat;
- private String outputFile;
private int subString = Integer.MAX_VALUE;
private int numTopFeatures = 10;
private Map<Integer, List<WeightedVectorWritable>> clusterIdToPoints;
@@ -104,8 +103,8 @@ public final class ClusterDumper extends
@Override
public int run(String[] args) throws Exception {
- addOption(SEQ_FILE_DIR_OPTION, "s", "The directory containing Sequence Files for the Clusters", true);
- addOption(OUTPUT_OPTION, "o", "Optional output directory. Default is to output to the console.");
+ addInputOption();
+ addOutputOption();
addOption(OUTPUT_FORMAT_OPT, "of", "The optional output format to write the results as. Options: TEXT, CSV or GRAPH_ML", "TEXT");
addOption(SUBSTRING_OPTION, "b", "The number of chars of the asFormatString() to print");
addOption(NUM_WORDS_OPTION, "n", "The number of top terms to print");
@@ -121,11 +120,11 @@ public final class ClusterDumper extends
return -1;
}
- seqFileDir = new Path(getOption(SEQ_FILE_DIR_OPTION));
+ seqFileDir = getInputPath();
if (hasOption(POINTS_DIR_OPTION)) {
pointsDir = new Path(getOption(POINTS_DIR_OPTION));
}
- outputFile = getOption(OUTPUT_OPTION);
+ outputFile = getOutputFile();
if (hasOption(SUBSTRING_OPTION)) {
int sub = Integer.parseInt(getOption(SUBSTRING_OPTION));
if (sub >= 0) {
@@ -174,12 +173,12 @@ public final class ClusterDumper extends
writer = new OutputStreamWriter(System.out);
} else {
shouldClose = true;
- if (outputFile.startsWith("s3n://")) {
- Path p = new Path(this.outputFile);
+ if (outputFile.getName().startsWith("s3n://")) {
+ Path p = outputPath;
FileSystem fs = FileSystem.get(p.toUri(), conf);
writer = new OutputStreamWriter(fs.create(p), Charsets.UTF_8);
} else {
- writer = Files.newWriter(new File(this.outputFile), Charsets.UTF_8);
+ writer = Files.newWriter(this.outputFile, Charsets.UTF_8);
}
}
ClusterWriter clusterWriter = createClusterWriter(writer, dictionary);
@@ -248,13 +247,6 @@ public final class ClusterDumper extends
}
}
- public String getOutputFile() {
- return outputFile;
- }
-
- public void setOutputFile(String outputFile) {
- this.outputFile = outputFile;
- }
public int getSubString() {
return subString;
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/RowIdJob.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/RowIdJob.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/RowIdJob.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/RowIdJob.java Mon Feb 13 15:14:42 2012
@@ -34,6 +34,7 @@ import org.apache.mahout.math.VectorWrit
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
+import java.util.List;
import java.util.Map;
public class RowIdJob extends AbstractJob {
@@ -45,7 +46,7 @@ public class RowIdJob extends AbstractJo
addInputOption();
addOutputOption();
- Map<String,String> parsedArgs = parseArguments(args);
+ Map<String,List<String>> parsedArgs = parseArguments(args);
if (parsedArgs == null) {
return -1;
}
Modified: mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/VectorDumper.java
URL: http://svn.apache.org/viewvc/mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/VectorDumper.java?rev=1243557&r1=1243556&r2=1243557&view=diff
==============================================================================
--- mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/VectorDumper.java (original)
+++ mahout/trunk/integration/src/main/java/org/apache/mahout/utils/vectors/VectorDumper.java Mon Feb 13 15:14:42 2012
@@ -92,7 +92,7 @@ public final class VectorDumper extends
addOption(buildOption("filter", "fi", "Only dump out those vectors whose name matches the filter." +
" Multiple items may be specified by repeating the argument.", true, 1, 100, false, null));
- if (parseArguments(args) == null) {
+ if (parseArguments(args, false, true) == null) {
return -1;
}