List of tableOf() Examples

Examples of tableOf()

Examples of com.cloudera.crunch.type.PTypeFamily.tableOf()

              Pair<String, String> pair =
                  Pair.of(word.toLowerCase(), filename);
              emitter.emit(pair);
            }
          }
        }, tf.tableOf(tf.strings(), tf.strings()));
  }


  public static enum LogCounters {
    LOG_LINE_ERRORS
  }

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

        new MapFn<Pair<T, N>, Pair<Integer, Pair<T, N>>>() {
          @Override
          public Pair<Integer, Pair<T, N>> map(Pair<T, N> p) {
            return Pair.of(0, p);
          }
        }, ptf.tableOf(ptf.ints(), input.getPType()));
    int[] ss = { sampleSize };
    return groupedWeightedReservoirSample(groupedIn, ss, seed)
        .parallelDo("Extract sampled value from pair", new MapFn<Pair<Integer, T>, T>() {
          @Override
          public T map(Pair<Integer, T> p) {

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

      PTable<Integer, Pair<T, N>> input,
      int[] sampleSizes,
      Long seed) {
    PTypeFamily ptf = input.getTypeFamily();
    PType<T> ttype = (PType<T>) input.getPTableType().getValueType().getSubTypes().get(0);
    PTableType<Integer, Pair<Double, T>> ptt = ptf.tableOf(ptf.ints(),
        ptf.pairs(ptf.doubles(), ttype));
    
    return input.parallelDo("Initial reservoir sampling", new ReservoirSampleFn<T, N>(sampleSizes, seed, ttype), ptt)
        .groupByKey(1)
        .combineValues(new WRSCombineFn<T>(sampleSizes, ttype))

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

   * @param valueType The {@code PType} for the value of the SequenceFile entry
   * @return A new {@code SourceTable<K, V>} instance
   */
  public static <K, V> TableSource<K, V> sequenceFile(Path path, PType<K> keyType, PType<V> valueType) {
    PTypeFamily ptf = keyType.getFamily();
    return new SeqFileTableSource<K, V>(path, ptf.tableOf(keyType, valueType));
  }


  /**
   * Creates a {@code TableSource<K, V>} instance for the SequenceFile(s) at the given {@code Path}s.
   *

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

   * @param valueType The {@code PType} for the value of the SequenceFile entry
   * @return A new {@code SourceTable<K, V>} instance
   */
  public static <K, V> TableSource<K, V> sequenceFile(List<Path> paths, PType<K> keyType, PType<V> valueType) {
    PTypeFamily ptf = keyType.getFamily();
    return new SeqFileTableSource<K, V>(paths, ptf.tableOf(keyType, valueType));
  }


  /**
   * Creates a {@code Source<String>} instance for the text file(s) at the given path name.
   *

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

  
  private static <K, V1, V2> PGroupedTable<Pair<K, V1>, Pair<V1, V2>> prepare(
      PTable<K, Pair<V1, V2>> input, int numReducers) {
    PTypeFamily ptf = input.getTypeFamily();
    PType<Pair<V1, V2>> valueType = input.getValueType();
    PTableType<Pair<K, V1>, Pair<V1, V2>> inter = ptf.tableOf(
        ptf.pairs(input.getKeyType(), valueType.getSubTypes().get(0)),
        valueType);
    GroupingOptions.Builder gob = GroupingOptions.builder()
        .requireSortedKeys()
        .groupingComparatorClass(JoinUtils.getGroupingComparator(ptf))

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

    PTypeFamily tf = collect.getTypeFamily();
    return collect.parallelDo("Aggregate.count", new MapFn<S, Pair<S, Long>>() {
      public Pair<S, Long> map(S input) {
        return Pair.of(input, 1L);
      }
    }, tf.tableOf(collect.getPType(), tf.longs()))
        .groupByKey(numPartitions)
        .combineValues(Aggregators.SUM_LONGS());
  }
  
  /**

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

            return Pair.of(1, 1L);
          }
          public void cleanup(Emitter<Pair<Integer, Long>> e) {
            e.emit(Pair.of(1, 0L));
          }
        }, tf.tableOf(tf.ints(), tf.longs()))
        .groupByKey(GroupingOptions.builder().numReducers(1).build())
        .combineValues(Aggregators.SUM_LONGS());
    PCollection<Long> count = countTable.values();
    return new FirstElementPObject<Long>(count);
  }

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

   */
  public static <K, V> PTable<K, V> top(PTable<K, V> ptable, int limit, boolean maximize) {
    PTypeFamily ptf = ptable.getTypeFamily();
    PTableType<K, V> base = ptable.getPTableType();
    PType<Pair<K, V>> pairType = ptf.pairs(base.getKeyType(), base.getValueType());
    PTableType<Integer, Pair<K, V>> inter = ptf.tableOf(ptf.ints(), pairType);
    return ptable.parallelDo("top" + limit + "map", new TopKFn<K, V>(limit, maximize, pairType), inter)
        .groupByKey(1).combineValues(new TopKCombineFn<K, V>(limit, maximize, pairType))
        .parallelDo("top" + limit + "reduce", new DoFn<Pair<Integer, Pair<K, V>>, Pair<K, V>>() {
          public void process(Pair<Integer, Pair<K, V>> input, Emitter<Pair<K, V>> emitter) {
            emitter.emit(input.second());

View Full Code Here

Examples of org.apache.crunch.types.PTypeFamily.tableOf()

          public void cleanup(Emitter<Pair<Boolean, S>> emitter) {
            if (max != null) {
              emitter.emit(Pair.of(true, max));
            }
          }
        }, tf.tableOf(tf.booleans(), collect.getPType())).groupByKey(1)
        .combineValues(new CombineFn<Boolean, S>() {
          public void process(Pair<Boolean, Iterable<S>> input, Emitter<Pair<Boolean, S>> emitter) {
            S max = null;
            for (S v : input.second()) {
              if (max == null || ((Comparable<S>) max).compareTo(v) < 0) {

View Full Code Here

0 1 2 3 4 5

TOP

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.