Examples with LongSumAggregatorFactory io.druid.query.aggregation.LongSumAggregatorFactory used on opensource projects

Example 81 with LongSumAggregatorFactory

use of io.druid.query.aggregation.LongSumAggregatorFactory in project druid by druid-io.

the class TopNTypeInterfaceBenchmark method setupQueries.

private void setupQueries() {
    // queries for the basic schema
    Map<String, TopNQueryBuilder> basicQueries = new LinkedHashMap<>();
    BenchmarkSchemaInfo basicSchema = BenchmarkSchemas.SCHEMA_MAP.get("basic");
    {
        // basic.A
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Arrays.asList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"));
        queryAggs.add(new LongMaxAggregatorFactory("maxLongUniform", "maxLongUniform"));
        queryAggs.add(new DoubleSumAggregatorFactory("sumFloatNormal", "sumFloatNormal"));
        queryAggs.add(new DoubleMinAggregatorFactory("minFloatZipf", "minFloatZipf"));
        queryAggs.add(new HyperUniquesAggregatorFactory("hyperUniquesMet", "hyper"));
        // Use an IdentityExtractionFn to force usage of DimExtractionTopNAlgorithm
        TopNQueryBuilder queryBuilderString = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension(new ExtractionDimensionSpec("dimSequential", "dimSequential", IdentityExtractionFn.getInstance())).metric("sumFloatNormal").intervals(intervalSpec).aggregators(queryAggs);
        // DimExtractionTopNAlgorithm is always used for numeric columns
        TopNQueryBuilder queryBuilderLong = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("metLongUniform").metric("sumFloatNormal").intervals(intervalSpec).aggregators(queryAggs);
        TopNQueryBuilder queryBuilderFloat = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("metFloatNormal").metric("sumFloatNormal").intervals(intervalSpec).aggregators(queryAggs);
        basicQueries.put("string", queryBuilderString);
        basicQueries.put("long", queryBuilderLong);
        basicQueries.put("float", queryBuilderFloat);
    }
    {
        // basic.numericSort
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Arrays.asList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"));
        TopNQueryBuilder queryBuilderA = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("dimUniform").metric(new DimensionTopNMetricSpec(null, StringComparators.NUMERIC)).intervals(intervalSpec).aggregators(queryAggs);
        basicQueries.put("numericSort", queryBuilderA);
    }
    {
        // basic.alphanumericSort
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Arrays.asList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"));
        TopNQueryBuilder queryBuilderA = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("dimUniform").metric(new DimensionTopNMetricSpec(null, StringComparators.ALPHANUMERIC)).intervals(intervalSpec).aggregators(queryAggs);
        basicQueries.put("alphanumericSort", queryBuilderA);
    }
    SCHEMA_QUERY_MAP.put("basic", basicQueries);
}

Example 82 with LongSumAggregatorFactory

use of io.druid.query.aggregation.LongSumAggregatorFactory in project druid by druid-io.

the class IngestSegmentFirehoseFactoryTest method constructorFeeder.

@Parameterized.Parameters(name = "{1}")
public static Collection<Object[]> constructorFeeder() throws IOException {
    final IndexSpec indexSpec = new IndexSpec();
    final HeapMemoryTaskStorage ts = new HeapMemoryTaskStorage(new TaskStorageConfig(null) {
    });
    final IncrementalIndexSchema schema = new IncrementalIndexSchema.Builder().withQueryGranularity(Granularities.NONE).withMinTimestamp(JodaUtils.MIN_INSTANT).withDimensionsSpec(ROW_PARSER).withMetrics(new AggregatorFactory[] { new LongSumAggregatorFactory(METRIC_LONG_NAME, DIM_LONG_NAME), new DoubleSumAggregatorFactory(METRIC_FLOAT_NAME, DIM_FLOAT_NAME) }).build();
    final OnheapIncrementalIndex index = new OnheapIncrementalIndex(schema, true, MAX_ROWS * MAX_SHARD_NUMBER);
    for (Integer i = 0; i < MAX_ROWS; ++i) {
        index.add(ROW_PARSER.parse(buildRow(i.longValue())));
    }
    if (!persistDir.mkdirs() && !persistDir.exists()) {
        throw new IOException(String.format("Could not create directory at [%s]", persistDir.getAbsolutePath()));
    }
    INDEX_MERGER.persist(index, persistDir, indexSpec);
    final TaskLockbox tl = new TaskLockbox(ts);
    final IndexerSQLMetadataStorageCoordinator mdc = new IndexerSQLMetadataStorageCoordinator(null, null, null) {

        private final Set<DataSegment> published = Sets.newHashSet();

        private final Set<DataSegment> nuked = Sets.newHashSet();

        @Override
        public List<DataSegment> getUsedSegmentsForInterval(String dataSource, Interval interval) throws IOException {
            return ImmutableList.copyOf(segmentSet);
        }

        @Override
        public List<DataSegment> getUsedSegmentsForIntervals(String dataSource, List<Interval> interval) throws IOException {
            return ImmutableList.copyOf(segmentSet);
        }

        @Override
        public List<DataSegment> getUnusedSegmentsForInterval(String dataSource, Interval interval) {
            return ImmutableList.of();
        }

        @Override
        public Set<DataSegment> announceHistoricalSegments(Set<DataSegment> segments) {
            Set<DataSegment> added = Sets.newHashSet();
            for (final DataSegment segment : segments) {
                if (published.add(segment)) {
                    added.add(segment);
                }
            }
            return ImmutableSet.copyOf(added);
        }

        @Override
        public void deleteSegments(Set<DataSegment> segments) {
            nuked.addAll(segments);
        }
    };
    final LocalTaskActionClientFactory tac = new LocalTaskActionClientFactory(ts, new TaskActionToolbox(tl, mdc, newMockEmitter(), EasyMock.createMock(SupervisorManager.class)));
    SegmentHandoffNotifierFactory notifierFactory = EasyMock.createNiceMock(SegmentHandoffNotifierFactory.class);
    EasyMock.replay(notifierFactory);
    final TaskToolboxFactory taskToolboxFactory = new TaskToolboxFactory(new TaskConfig(tmpDir.getAbsolutePath(), null, null, 50000, null, false, null, null), tac, newMockEmitter(), new DataSegmentPusher() {

        @Deprecated
        @Override
        public String getPathForHadoop(String dataSource) {
            return getPathForHadoop();
        }

        @Override
        public String getPathForHadoop() {
            throw new UnsupportedOperationException();
        }

        @Override
        public DataSegment push(File file, DataSegment segment) throws IOException {
            return segment;
        }
    }, new DataSegmentKiller() {

        @Override
        public void kill(DataSegment segments) throws SegmentLoadingException {
        }

        @Override
        public void killAll() throws IOException {
            throw new UnsupportedOperationException("not implemented");
        }
    }, new DataSegmentMover() {

        @Override
        public DataSegment move(DataSegment dataSegment, Map<String, Object> targetLoadSpec) throws SegmentLoadingException {
            return dataSegment;
        }
    }, new DataSegmentArchiver() {

        @Override
        public DataSegment archive(DataSegment segment) throws SegmentLoadingException {
            return segment;
        }

        @Override
        public DataSegment restore(DataSegment segment) throws SegmentLoadingException {
            return segment;
        }
    }, // segment announcer
    null, notifierFactory, // query runner factory conglomerate corporation unionized collective
    null, // query executor service
    null, // monitor scheduler
    null, new SegmentLoaderFactory(new SegmentLoaderLocalCacheManager(null, new SegmentLoaderConfig() {

        @Override
        public List<StorageLocationConfig> getLocations() {
            return Lists.newArrayList();
        }
    }, MAPPER)), MAPPER, INDEX_MERGER, INDEX_IO, null, null, INDEX_MERGER_V9);
    Collection<Object[]> values = new LinkedList<>();
    for (InputRowParser parser : Arrays.<InputRowParser>asList(ROW_PARSER, new MapInputRowParser(new JSONParseSpec(new TimestampSpec(TIME_COLUMN, "auto", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.<String>of()), ImmutableList.of(DIM_FLOAT_NAME, DIM_LONG_NAME), ImmutableList.<SpatialDimensionSchema>of()), null, null)))) {
        for (List<String> dim_names : Arrays.<List<String>>asList(null, ImmutableList.of(DIM_NAME))) {
            for (List<String> metric_names : Arrays.<List<String>>asList(null, ImmutableList.of(METRIC_LONG_NAME, METRIC_FLOAT_NAME))) {
                values.add(new Object[] { new IngestSegmentFirehoseFactory(DATA_SOURCE_NAME, FOREVER, new SelectorDimFilter(DIM_NAME, DIM_VALUE, null), dim_names, metric_names, Guice.createInjector(new Module() {

                    @Override
                    public void configure(Binder binder) {
                        binder.bind(TaskToolboxFactory.class).toInstance(taskToolboxFactory);
                    }
                }), INDEX_IO), String.format("DimNames[%s]MetricNames[%s]ParserDimNames[%s]", dim_names == null ? "null" : "dims", metric_names == null ? "null" : "metrics", parser == ROW_PARSER ? "dims" : "null"), parser });
            }
        }
    }
    return values;
}

Example 83 with LongSumAggregatorFactory

use of io.druid.query.aggregation.LongSumAggregatorFactory in project druid by druid-io.

the class IngestSegmentFirehoseFactoryTimelineTest method persist.

private static Map<String, Object> persist(File tmpDir, InputRow... rows) {
    final File persistDir = new File(tmpDir, UUID.randomUUID().toString());
    final IncrementalIndexSchema schema = new IncrementalIndexSchema.Builder().withQueryGranularity(Granularities.NONE).withMinTimestamp(JodaUtils.MIN_INSTANT).withDimensionsSpec(ROW_PARSER).withMetrics(new AggregatorFactory[] { new LongSumAggregatorFactory(METRICS[0], METRICS[0]) }).build();
    final OnheapIncrementalIndex index = new OnheapIncrementalIndex(schema, true, rows.length);
    for (InputRow row : rows) {
        try {
            index.add(row);
        } catch (IndexSizeExceededException e) {
            throw Throwables.propagate(e);
        }
    }
    try {
        INDEX_MERGER.persist(index, persistDir, new IndexSpec());
    } catch (IOException e) {
        throw Throwables.propagate(e);
    }
    return ImmutableMap.<String, Object>of("type", "local", "path", persistDir.getAbsolutePath());
}

Example 84 with LongSumAggregatorFactory

use of io.druid.query.aggregation.LongSumAggregatorFactory in project druid by druid-io.

the class GroupByQueryRunnerTest method testPostAggMergedHavingSpec.

@Test
public void testPostAggMergedHavingSpec() {
    List<Row> expectedResults = Arrays.asList(GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "mezzanine", "rows", 6L, "index", 4420L, QueryRunnerTestHelper.addRowsIndexConstantMetric, (double) (6L + 4420L + 1L)), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "premium", "rows", 6L, "index", 4416L, QueryRunnerTestHelper.addRowsIndexConstantMetric, (double) (6L + 4416L + 1L)));
    GroupByQuery.Builder builder = GroupByQuery.builder().setDataSource(QueryRunnerTestHelper.dataSource).setInterval("2011-04-02/2011-04-04").setDimensions(Lists.<DimensionSpec>newArrayList(new DefaultDimensionSpec("quality", "alias"))).setAggregatorSpecs(Arrays.asList(QueryRunnerTestHelper.rowsCount, new LongSumAggregatorFactory("index", "index"))).setPostAggregatorSpecs(ImmutableList.<PostAggregator>of(QueryRunnerTestHelper.addRowsIndexConstant)).setGranularity(new PeriodGranularity(new Period("P1M"), null, null)).setHavingSpec(new OrHavingSpec(ImmutableList.<HavingSpec>of(new GreaterThanHavingSpec(QueryRunnerTestHelper.addRowsIndexConstantMetric, 1000L))));
    final GroupByQuery fullQuery = builder.build();
    QueryRunner mergedRunner = factory.getToolchest().mergeResults(new QueryRunner<Row>() {

        @Override
        public Sequence<Row> run(Query<Row> query, Map<String, Object> responseContext) {
            // simulate two daily segments
            final Query query1 = query.withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Lists.newArrayList(new Interval("2011-04-02/2011-04-03"))));
            final Query query2 = query.withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Lists.newArrayList(new Interval("2011-04-03/2011-04-04"))));
            return new MergeSequence(query.getResultOrdering(), Sequences.simple(Arrays.asList(runner.run(query1, responseContext), runner.run(query2, responseContext))));
        }
    });
    Map<String, Object> context = Maps.newHashMap();
    TestHelper.assertExpectedObjects(expectedResults, mergedRunner.run(fullQuery, context), "merged");
}

Example 85 with LongSumAggregatorFactory

use of io.druid.query.aggregation.LongSumAggregatorFactory in project druid by druid-io.

the class GroupByQueryRunnerTest method testSubqueryWithMultipleIntervalsInOuterQueryAndChunkPeriod.

@Test
public void testSubqueryWithMultipleIntervalsInOuterQueryAndChunkPeriod() {
    GroupByQuery subquery = GroupByQuery.builder().setDataSource(QueryRunnerTestHelper.dataSource).setQuerySegmentSpec(QueryRunnerTestHelper.firstToThird).setDimensions(Lists.<DimensionSpec>newArrayList(new DefaultDimensionSpec("quality", "alias"))).setDimFilter(new JavaScriptDimFilter("quality", "function(dim){ return true; }", null, JavaScriptConfig.getEnabledInstance())).setAggregatorSpecs(Arrays.asList(QueryRunnerTestHelper.rowsCount, new LongSumAggregatorFactory("idx", "index"), new LongSumAggregatorFactory("indexMaxPlusTen", "indexMaxPlusTen"))).setGranularity(QueryRunnerTestHelper.dayGran).setContext(ImmutableMap.<String, Object>of("chunkPeriod", "P1D")).build();
    GroupByQuery query = GroupByQuery.builder().setDataSource(subquery).setQuerySegmentSpec(new MultipleIntervalSegmentSpec(ImmutableList.of(new Interval("2011-04-01T00:00:00.000Z/2011-04-01T23:58:00.000Z"), new Interval("2011-04-02T00:00:00.000Z/2011-04-03T00:00:00.000Z")))).setDimensions(Lists.<DimensionSpec>newArrayList(new DefaultDimensionSpec("alias", "alias"))).setAggregatorSpecs(Arrays.<AggregatorFactory>asList(new LongSumAggregatorFactory("rows", "rows"), new LongSumAggregatorFactory("idx", "idx"))).setGranularity(QueryRunnerTestHelper.dayGran).build();
    List<Row> expectedResults = Arrays.asList(GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "automotive", "rows", 1L, "idx", 135L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "business", "rows", 1L, "idx", 118L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "entertainment", "rows", 1L, "idx", 158L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "health", "rows", 1L, "idx", 120L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "mezzanine", "rows", 3L, "idx", 2870L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "news", "rows", 1L, "idx", 121L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "premium", "rows", 3L, "idx", 2900L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "technology", "rows", 1L, "idx", 78L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-01", "alias", "travel", "rows", 1L, "idx", 119L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "automotive", "rows", 1L, "idx", 147L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "business", "rows", 1L, "idx", 112L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "entertainment", "rows", 1L, "idx", 166L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "health", "rows", 1L, "idx", 113L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "mezzanine", "rows", 3L, "idx", 2447L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "news", "rows", 1L, "idx", 114L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "premium", "rows", 3L, "idx", 2505L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "technology", "rows", 1L, "idx", 97L), GroupByQueryRunnerTestHelper.createExpectedRow("2011-04-02", "alias", "travel", "rows", 1L, "idx", 126L));
    // Subqueries are handled by the ToolChest
    Iterable<Row> results = GroupByQueryRunnerTestHelper.runQuery(factory, runner, query);
    TestHelper.assertExpectedObjects(expectedResults, results, "");
}