Examples of org.apache.mahout.common.FileLineIterable

org.apache.mahout.common.FileLineIterable
Iterable representing the lines of a text file. It can produce an {@link Iterator} over those lines. Thisassumes the text file's lines are delimited in a manner consistent with how {@link java.io.BufferedReader}defines lines. This class will uncompress files that end in .zip or .gz accordingly, too.

    if (dfs.exists(outPath)) {
      dfs.delete(outPath, true);
    }
    
    Set<String> categories = new HashSet<String>();
    for (String line : new FileLineIterable(new File(catFile))) {
      categories.add(line.trim().toLowerCase());
    }
    
    DefaultStringifier<Set<String>> setStringifier = new DefaultStringifier<Set<String>>(conf, GenericsUtil
        .getClass(categories));

View Full Code Here

      dfs.delete(outPath, true);
    }
    
    Set<String> categories = new HashSet<String>();
    if (catFile.length() > 0) {
      for (String line : new FileLineIterable(new File(catFile))) {
        categories.add(line.trim().toLowerCase());
      }
    }
    
    DefaultStringifier<Set<String>> setStringifier = new DefaultStringifier<Set<String>>(conf, GenericsUtil

View Full Code Here

   * @throws IOException
   *             if there is an error
   */
  public static List<VectorWritable> readFile(String fileName) throws IOException {
    List<VectorWritable> results = new ArrayList<VectorWritable>();
    for (String line : new FileLineIterable(new File(fileName))) {
      results.add(new VectorWritable(AbstractVector.decodeVector(line)));
    }
    return results;
  }

View Full Code Here

        current.listFiles(new PrefixAdditionFilter(prefix + File.separator + current.getName(), writer,
            charset));
      } else {
        try {
          StringBuilder file = new StringBuilder();
          for (String aFit : new FileLineIterable(current, charset, false)) {
            file.append(aFit).append('\n');
          }
          writer.write(prefix + File.separator + current.getName(), file.toString());
          
        } catch (FileNotFoundException e) {

View Full Code Here

        usersToRecommendFor = null;
      } else {
        usersToRecommendFor = new FastIDSet();
        Path usersFilePath = new Path(usersFilePathString).makeQualified(fs);
        FSDataInputStream in = fs.open(usersFilePath);
        for (String line : new FileLineIterable(in)) {
          usersToRecommendFor.add(Long.parseLong(line));
        }
      }
    } catch (IOException ioe) {
      throw new IllegalStateException(ioe);

View Full Code Here

   * @throws IOException
   *             if there is an error
   */
  public static List<VectorWritable> readFile(String fileName) throws IOException {
    List<VectorWritable> results = new ArrayList<VectorWritable>();
    for (String line : new FileLineIterable(new File(fileName))) {
      results.add(new VectorWritable(AbstractVector.decodeVector(line)));
    }
    return results;
  }

View Full Code Here

        usersToRecommendFor = null;
      } else {
        usersToRecommendFor = new FastIDSet();
        Path usersFilePath = new Path(usersFilePathString).makeQualified(fs);
        FSDataInputStream in = fs.open(usersFilePath);
        for (String line : new FileLineIterable(in)) {
          usersToRecommendFor.add(Long.parseLong(line));
        }
      }
    } catch (IOException ioe) {
      throw new IllegalStateException(ioe);

View Full Code Here

    
    FPGrowth<String> fp = new FPGrowth<String>();
    Set<String> features = new HashSet<String>();
    
    fp.generateTopKFrequentPatterns(
        new StringRecordIterator(new FileLineIterable(new File(input), encoding, false), pattern),
        fp.generateFList(
            new StringRecordIterator(new FileLineIterable(new File(input), encoding, false), pattern),
            minSupport),
        minSupport,
        maxHeapSize,
        features,
        new StringOutputConverter(new SequenceFileOutputCollector<Text,TopKStringPatterns>(writer)),

View Full Code Here

        Path unqualifiedUsersFilePath = new Path(usersFilePathString);
        FileSystem fs = FileSystem.get(unqualifiedUsersFilePath.toUri(), jobConf);
        usersToRecommendFor = new FastIDSet();
        Path usersFilePath = unqualifiedUsersFilePath.makeQualified(fs);
        in = fs.open(usersFilePath);
        for (String line : new FileLineIterable(in)) {
          usersToRecommendFor.add(Long.parseLong(line));
        }     
      } catch (IOException ioe) {
        throw new IllegalStateException(ioe);
      } finally {

View Full Code Here

    File resultFile = new File(new File(System.getProperty("java.io.tmpdir")), "taste.bookcrossing.txt");
    resultFile.delete();
    PrintWriter writer = null;
    try {
      writer = new PrintWriter(new OutputStreamWriter(new FileOutputStream(resultFile), Charset.forName("UTF-8")));
      for (String line : new FileLineIterable(originalFile, true)) {
        // 0 ratings are basically "no rating", ignore them (thanks h.9000)
        if (line.endsWith("\"0\"")) {
          continue;
        }
        // Delete replace anything that isn't numeric, or a semicolon delimiter. Make comma the delimiter.

View Full Code Here

0 1 2 3 4

TOP

Related Classes of org.apache.mahout.common.FileLineIterable

org.apache.mahout.cf.taste.example.bookcrossing.BookCrossingDataModel

org.apache.mahout.cf.taste.example.grouplens.GroupLensDataModel

org.apache.mahout.cf.taste.example.netflix.NetflixDataModel

org.apache.mahout.cf.taste.example.netflix.NetflixFileDataModel

org.apache.mahout.cf.taste.example.netflix.TransposeToByUser

org.apache.mahout.cf.taste.hadoop.item.AggregateAndRecommendReducer

org.apache.mahout.cf.taste.hadoop.item.RecommenderJobTest

org.apache.mahout.cf.taste.hadoop.item.RecommenderMapper

org.apache.mahout.cf.taste.hadoop.item.UserVectorSplitterMapper

org.apache.mahout.cf.taste.hadoop.pseudo.RecommenderReducer

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.