Package org.nebulaframework.benchmark.scimark2

Examples of org.nebulaframework.benchmark.scimark2.commandline

71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
    .withOption(substringOpt).withOption(countOpt).withOption(helpOpt).create();
   
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
     
      if (cmdLine.hasOption(helpOpt)) {
       
        printHelp(group);
        return;
      }
     
      if (cmdLine.hasOption(seqOpt)) {
        Path path = new Path(cmdLine.getValue(seqOpt).toString());
        JobClient client = new JobClient();
        JobConf conf = new JobConf(Job.class);
        client.setConf(conf);
        FileSystem fs = FileSystem.get(path.toUri(), conf);
        SequenceFile.Reader reader = new SequenceFile.Reader(fs, path, conf);
       
        Writer writer;
        if (cmdLine.hasOption(outputOpt)) {
          writer = new FileWriter(cmdLine.getValue(outputOpt).toString());
        } else {
          writer = new OutputStreamWriter(System.out);
        }
        writer.append("Input Path: ").append(String.valueOf(path)).append('\n');
       
        int sub = Integer.MAX_VALUE;
        if (cmdLine.hasOption(substringOpt)) {
          sub = Integer.parseInt(cmdLine.getValue(substringOpt).toString());
        }
        boolean countOnly = cmdLine.hasOption(countOpt);
        Writable key = (Writable) reader.getKeyClass().newInstance();
        Writable value = (Writable) reader.getValueClass().newInstance();
        writer.append("Key class: ").append(String.valueOf(reader.getKeyClass())).append(" Value Class: ")
        .append(String.valueOf(value.getClass())).append('\n');
        writer.flush();
        long count = 0;
        if (countOnly == false) {
          while (reader.next(key, value)) {
            writer.append("Key: ").append(String.valueOf(key));
            String str = value.toString();
            writer.append(": Value: ").append(str.length() > sub ? str.substring(0, sub) : str);
            writer.write('\n');
            writer.flush();
            count++;
          }
          writer.append("Count: ").append(String.valueOf(count)).append('\n');
        } else {
          while (reader.next(key, value)) {
            count++;
          }
          writer.append("Count: ").append(String.valueOf(count)).append('\n');
        }
        writer.flush();
        if (cmdLine.hasOption(outputOpt)) {
          writer.close();
        }
      }
     
    } catch (OptionException e) {
View Full Code Here
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
        .create();
   
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
     
      if (cmdLine.hasOption(helpOpt)) {
        CommandLineUtil.printHelp(group);
        return;
      }
     
      String input = cmdLine.getValue(inputOpt).toString();
      String output = cmdLine.getValue(outputOpt).toString();
     
      job.runJob(input, output, new BayesParameters(1));
    } catch (OptionException e) {
      log.error(e.getMessage());
      CommandLineUtil.printHelp(group);
View Full Code Here
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
      printKeyOpt).create();
   
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
     
      if (cmdLine.hasOption(helpOpt)) {
       
        printHelp(group);
        return;
      }
     
      if (cmdLine.hasOption(seqOpt)) {
        Path path = new Path(cmdLine.getValue(seqOpt).toString());
        System.out.println("Input Path: " + path);
        JobClient client = new JobClient();
        JobConf conf = new JobConf(Job.class);
        client.setConf(conf);
       
        FileSystem fs = FileSystem.get(path.toUri(), conf);
       
        String dictionaryType = "text";
        if (cmdLine.hasOption(dictTypeOpt)) {
          dictionaryType = cmdLine.getValue(dictTypeOpt).toString();
        }
       
        String[] dictionary = null;
        if (cmdLine.hasOption(dictOpt)) {
          if (dictionaryType.equals("text")) {
            dictionary = VectorHelper.loadTermDictionary(new File(cmdLine.getValue(dictOpt).toString()));
          } else if (dictionaryType.equals("sequencefile")) {
            dictionary = VectorHelper.loadTermDictionary(conf, fs, cmdLine.getValue(dictOpt).toString());
          } else {
            throw new OptionException(dictTypeOpt);
          }
        }
        boolean useJSON = cmdLine.hasOption(centroidJSonOpt);
       
        SequenceFile.Reader reader = new SequenceFile.Reader(fs, path, conf);
        SequenceFileVectorIterable vectorIterable = new SequenceFileVectorIterable(reader, cmdLine
            .hasOption(vectorAsKeyOpt));
        Writer writer;
        if (cmdLine.hasOption(outputOpt)) {
          writer = new FileWriter(cmdLine.getValue(outputOpt).toString());
        } else {
          writer = new OutputStreamWriter(System.out);
        }
        boolean printKey = cmdLine.hasOption(printKeyOpt);
        SeqFileIterator iterator = (SeqFileIterator) vectorIterable.iterator();
        int i = 0;
        while (iterator.hasNext()) {
          Vector vector = iterator.next();
          if (printKey) {
            writer.write(iterator.key().toString());
            writer.write("\t");
          }
          String fmtStr = useJSON ? vector.asFormatString() : VectorHelper.vectorToString(vector, dictionary);
          writer.write(fmtStr);
          writer.write('\n');
          i++;
        }
        writer.flush();
        if (cmdLine.hasOption(outputOpt)) {
          writer.close();
        }
        System.err.println("Dumped " + i + " Vectors");
      }
     
View Full Code Here
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
      measureClassOpt).withOption(convergenceDeltaOpt).withOption(maxIterationsOpt)
        .withOption(vectorClassOpt).withOption(t1Opt).withOption(t2Opt).withOption(helpOpt).create();
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
     
      if (cmdLine.hasOption(helpOpt)) {
        CommandLineUtil.printHelp(group);
        return;
      }
      String input = cmdLine.getValue(inputOpt, "testdata").toString();
      String output = cmdLine.getValue(outputOpt, "output").toString();
      String measureClass = cmdLine.getValue(measureClassOpt,
        "org.apache.mahout.common.distance.EuclideanDistanceMeasure").toString();
      double t1 = Double.parseDouble(cmdLine.getValue(t1Opt, "80").toString());
      double t2 = Double.parseDouble(cmdLine.getValue(t2Opt, "55").toString());
      double convergenceDelta = Double.parseDouble(cmdLine.getValue(convergenceDeltaOpt, "0.5").toString());
      int maxIterations = Integer.parseInt(cmdLine.getValue(maxIterationsOpt, 10).toString());
      // String className = cmdLine.getValue(vectorClassOpt,
      // "org.apache.mahout.math.RandomAccessSparseVector").toString();
      // Class<? extends Vector> vectorClass = Class.forName(className).asSubclass(Vector.class);
     
      runJob(input, output, measureClass, t1, t2, convergenceDelta, maxIterations);
View Full Code Here
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
        .create();
   
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
      if (cmdLine.hasOption(helpOpt)) {
        CommandLineUtil.printHelp(group);
        return;
      }
     
      String input = cmdLine.getValue(inputOpt, "testdata").toString();
      String output = cmdLine.getValue(outputOpt, "output").toString();
      runJob(input, output);
    } catch (OptionException e) {
      InputDriver.LOG.error("Exception parsing command line: ", e);
      CommandLineUtil.printHelp(group);
    }
View Full Code Here
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
      vectorOpt).withOption(helpOpt).create();

    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
      if (cmdLine.hasOption(helpOpt)) {
        CommandLineUtil.printHelp(group);
        return;
      }
     
      String input = cmdLine.getValue(inputOpt, "testdata").toString();
      String output = cmdLine.getValue(outputOpt, "output").toString();
      String vectorClassName = cmdLine.getValue(vectorOpt,
         "org.apache.mahout.math.RandomAccessSparseVector").toString();
      runJob(input, output, vectorClassName);
    } catch (OptionException e) {
      InputDriver.LOG.error("Exception parsing command line: ", e);
      CommandLineUtil.printHelp(group);
View Full Code Here
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
        .withOption(maxIterOpt).withOption(threshold2Opt).create();
   
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
      if (cmdLine.hasOption(helpOpt)) {
        CommandLineUtil.printHelp(group);
        return;
      }
     
      String input = cmdLine.getValue(inputOpt, "testdata").toString();
      String output = cmdLine.getValue(outputOpt, "output").toString();
      String measureClassName = cmdLine.getValue(modelOpt,
        "org.apache.mahout.common.distance.EuclideanDistanceMeasure").toString();
      double t1 = Double.parseDouble(cmdLine.getValue(threshold1Opt, "47.6").toString());
      double t2 = Double.parseDouble(cmdLine.getValue(threshold2Opt, "1").toString());
      double convergenceDelta = Double.parseDouble(cmdLine.getValue(convergenceDeltaOpt, "0.5").toString());
      int maxIterations = Integer.parseInt(cmdLine.getValue(maxIterOpt, "10").toString());
      runJob(input, output, measureClassName, t1, t2, convergenceDelta, maxIterations);
    } catch (OptionException e) {
      log.error("Exception parsing command line: ", e);
      CommandLineUtil.printHelp(group);
    }
View Full Code Here
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
        .withOption(t1Opt).withOption(t2Opt).withOption(helpOpt).create();
   
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
     
      if (cmdLine.hasOption(helpOpt)) {
        CommandLineUtil.printHelp(group);
        return;
      }
     
      String input = cmdLine.getValue(inputOpt, "testdata").toString();
      String output = cmdLine.getValue(outputOpt, "output").toString();
      String measureClass = cmdLine.getValue(measureClassOpt,
        "org.apache.mahout.common.distance.EuclideanDistanceMeasure").toString();
     
      // String className = cmdLine.getValue(vectorClassOpt,
      // "org.apache.mahout.math.RandomAccessSparseVector").toString();
      // Class<? extends Vector> vectorClass = Class.forName(className).asSubclass(Vector.class);
      double t1 = Double.parseDouble(cmdLine.getValue(t1Opt, "80").toString());
      double t2 = Double.parseDouble(cmdLine.getValue(t2Opt, "55").toString());
     
      runJob(input, output, measureClass, t1, t2);
    } catch (OptionException e) {
      Job.log.error("Exception", e);
      CommandLineUtil.printHelp(group);
View Full Code Here
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
        .create();
   
    try {
      Parser parser = new Parser();
      parser.setGroup(group);
      CommandLine cmdLine = parser.parse(args);
      if (cmdLine.hasOption(helpOpt)) {
        CommandLineUtil.printHelp(group);
        return;
      }
     
      String input = cmdLine.getValue(inputOpt, "testdata").toString();
      String output = cmdLine.getValue(outputOpt, "output").toString();
      runJob(input, output);
    } catch (OptionException e) {
      OutputDriver.LOG.error("Exception parsing command line: ", e);
      CommandLineUtil.printHelp(group);
    }
View Full Code Here
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
    Group group = gbuilder.withName("Options").withOption(dumpFileOpt).withOption(outputDirOpt).withOption(
      chunkSizeOpt).withOption(numChunksOpt).withOption(s3IdOpt).withOption(s3SecretOpt).create();
   
    Parser parser = new Parser();
    parser.setGroup(group);
    CommandLine cmdLine;
    try {
      cmdLine = parser.parse(args);
    } catch (OptionException e) {
      log.error("Error while parsing options", e);
      CommandLineUtil.printHelp(group);
      return;
    }
   
    Configuration conf = new Configuration();
    String dumpFilePath = (String) cmdLine.getValue(dumpFileOpt);
    String outputDirPath = (String) cmdLine.getValue(outputDirOpt);
   
    if (cmdLine.hasOption(s3IdOpt)) {
      String id = (String) cmdLine.getValue(s3IdOpt);
      conf.set("fs.s3n.awsAccessKeyId", id);
      conf.set("fs.s3.awsAccessKeyId", id);
    }
    if (cmdLine.hasOption(s3SecretOpt)) {
      String secret = (String) cmdLine.getValue(s3SecretOpt);
      conf.set("fs.s3n.awsSecretAccessKey", secret);
      conf.set("fs.s3.awsSecretAccessKey", secret);
    }
    // do not compute crc file when using local FS
    conf.set("fs.file.impl", "org.apache.hadoop.fs.RawLocalFileSystem");
    FileSystem fs = FileSystem.get(URI.create(outputDirPath), conf);
   
    int chunkSize = 1024 * 1024 * Integer.parseInt((String) cmdLine.getValue(chunkSizeOpt));
   
    int numChunks = Integer.MAX_VALUE;
    if (cmdLine.hasOption(numChunksOpt)) {
      numChunks = Integer.parseInt((String) cmdLine.getValue(numChunksOpt));
    }
   
    String header = "<mediawiki xmlns=\"http://www.mediawiki.org/xml/export-0.3/\" "
                    + "xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\" "
                    + "xsi:schemaLocation=\"http://www.mediawiki.org/xml/export-0.3/ "
View Full Code Here
TOP

Related Classes of org.nebulaframework.benchmark.scimark2.commandline

  • com.digitalpebble.behemoth.mahout.SparseVectorsFromBehemoth
  • com.digitalpebble.behemoth.tika.TikaDriver
  • org.apache.commons.cli2.commandline.DefaultingCommandLine
  • org.apache.hadoop.mapred.pipes.Submitter
  • org.apache.hadoop.streaming.StreamJob
  • org.apache.mahout.benchmark.VectorBenchmarks
  • org.apache.mahout.cf.taste.example.TasteOptionParser
  • org.apache.mahout.classifier.bayes.mapreduce.common.JobExecutor
  • org.apache.mahout.classifier.bayes.PrepareTwentyNewsgroups
  • org.apache.mahout.classifier.bayes.SplitBayesInput
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.