Examples with PTypeFamily org.apache.crunch.types.PTypeFamily used on opensource projects

Example 11 with PTypeFamily

use of org.apache.crunch.types.PTypeFamily in project crunch by cloudera.

the class Aggregate method top.

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());
        }
    }, base);
}

Example 12 with PTypeFamily

use of org.apache.crunch.types.PTypeFamily in project crunch by cloudera.

the class Join method join.

public static <K, U, V> PTable<K, Pair<U, V>> join(PTable<K, U> left, PTable<K, V> right, JoinFn<K, U, V> joinFn) {
    PTypeFamily ptf = left.getTypeFamily();
    PGroupedTable<Pair<K, Integer>, Pair<U, V>> grouped = preJoin(left, right);
    PTableType<K, Pair<U, V>> ret = ptf.tableOf(left.getKeyType(), ptf.pairs(left.getValueType(), right.getValueType()));
    return grouped.parallelDo(joinFn.getJoinType() + grouped.getName(), joinFn, ret);
}

Example 13 with PTypeFamily

use of org.apache.crunch.types.PTypeFamily in project crunch by cloudera.

the class Join method preJoin.

private static <K, U, V> PGroupedTable<Pair<K, Integer>, Pair<U, V>> preJoin(PTable<K, U> left, PTable<K, V> right) {
    PTypeFamily ptf = left.getTypeFamily();
    PTableType<Pair<K, Integer>, Pair<U, V>> ptt = ptf.tableOf(ptf.pairs(left.getKeyType(), ptf.ints()), ptf.pairs(left.getValueType(), right.getValueType()));
    PTable<Pair<K, Integer>, Pair<U, V>> tag1 = left.parallelDo("joinTagLeft", new MapFn<Pair<K, U>, Pair<Pair<K, Integer>, Pair<U, V>>>() {

        @Override
        public Pair<Pair<K, Integer>, Pair<U, V>> map(Pair<K, U> input) {
            return Pair.of(Pair.of(input.first(), 0), Pair.of(input.second(), (V) null));
        }
    }, ptt);
    PTable<Pair<K, Integer>, Pair<U, V>> tag2 = right.parallelDo("joinTagRight", new MapFn<Pair<K, V>, Pair<Pair<K, Integer>, Pair<U, V>>>() {

        @Override
        public Pair<Pair<K, Integer>, Pair<U, V>> map(Pair<K, V> input) {
            return Pair.of(Pair.of(input.first(), 1), Pair.of((U) null, input.second()));
        }
    }, ptt);
    GroupingOptions.Builder optionsBuilder = GroupingOptions.builder();
    optionsBuilder.partitionerClass(JoinUtils.getPartitionerClass(ptf));
    return (tag1.union(tag2)).groupByKey(optionsBuilder.build());
}

Example 14 with PTypeFamily

use of org.apache.crunch.types.PTypeFamily in project crunch by cloudera.

the class PageRankTest method testWritablesBSON.

@Test
public void testWritablesBSON() throws Exception {
    PTypeFamily tf = WritableTypeFamily.getInstance();
    PType<PageRankData> prType = PTypes.smile(PageRankData.class, tf);
    run(new MRPipeline(PageRankTest.class), prType, tf);
}

Example 15 with PTypeFamily

use of org.apache.crunch.types.PTypeFamily in project crunch by cloudera.

the class PageRankTest method testWritablesJSON.

@Test
public void testWritablesJSON() throws Exception {
    PTypeFamily tf = WritableTypeFamily.getInstance();
    PType<PageRankData> prType = PTypes.jsonString(PageRankData.class, tf);
    run(new MRPipeline(PageRankTest.class), prType, tf);
}