List of tableOf() Examples

Examples of tableOf()

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

      return left.parallelDo("mapjoin", mapJoinDoFn,
          tf.tableOf(left.getKeyType(), tf.pairs(left.getValueType(), right.getValueType())),
          optionsBuilder.build());
    } else { // in-memory pipeline
      return left.parallelDo(new InMemoryJoinFn<K, U, V>(iterable),
          tf.tableOf(left.getKeyType(), tf.pairs(left.getValueType(), right.getValueType())));
    }
  }


  static class InMemoryJoinFn<K, U, V> extends DoFn<Pair<K, U>, Pair<K, Pair<U, V>>> {

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 = new int[] { sampleSize };
    return groupedWeightedReservoirSample(groupedIn, ss, seed)
        .parallelDo(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(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()

  public static <S> PCollection<S> distinct(PCollection<S> input, int flushEvery) {
    Preconditions.checkArgument(flushEvery > 0);
    PType<S> pt = input.getPType();
    PTypeFamily ptf = pt.getFamily();
    return input
        .parallelDo("pre-distinct", new PreDistinctFn<S>(flushEvery, pt), ptf.tableOf(pt, ptf.nulls()))
        .groupByKey()
        .parallelDo("post-distinct", new PostDistinctFn<S>(), pt);
  }
  
  /**

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()
        .combineValues(Aggregators.SUM_LONGS());
  }


  /**
   * Returns the number of elements in the provided PCollection.

View Full Code Here

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

    PTable<Integer, Long> countTable = collect
        .parallelDo("Aggregate.count", new MapFn<S, Pair<Integer, Long>>() {
          public Pair<Integer, Long> map(S input) {
            return Pair.of(1, 1L);
          }
        }, 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), inter)
        .groupByKey(1).combineValues(new TopKCombineFn<K, V>(limit, maximize))
        .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

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

          public void cleanup(Emitter<Pair<Boolean, S>> emitter) {
            if (min != null) {
              emitter.emit(Pair.of(false, min));
            }
          }
        }, 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 min = null;
            for (S v : input.second()) {
              if (min == null || ((Comparable<S>) min).compareTo(v) > 0) {

View Full Code Here

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

            for (V value : values) {
              collected.add(valueType.getDetachedValue(value));
            }
            return collected;
          }
        }, tf.tableOf(collect.getKeyType(), tf.collections(collect.getValueType())));
  }
}

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.