Search in sources :

Example 11 with FilteredAggregatorFactory

use of org.apache.druid.query.aggregation.FilteredAggregatorFactory in project druid by druid-io.

the class CalciteJoinQueryTest method testJoinWithLimitBeforeJoining.

@Test
public void testJoinWithLimitBeforeJoining() throws Exception {
    // Cannot vectorize JOIN operator.
    cannotVectorize();
    testQuery("SELECT t1.dim2, AVG(t1.m2) FROM (SELECT * FROM foo LIMIT 10) AS t1 INNER JOIN foo AS t2 ON t1.m1 = t2.m1 GROUP BY t1.dim2", ImmutableList.of(GroupByQuery.builder().setDataSource(join(new QueryDataSource(newScanQueryBuilder().dataSource(CalciteTests.DATASOURCE1).intervals(querySegmentSpec(Filtration.eternity())).columns("dim2", "m1", "m2").context(QUERY_CONTEXT_DEFAULT).limit(10).build()), new QueryDataSource(newScanQueryBuilder().dataSource(CalciteTests.DATASOURCE1).intervals(querySegmentSpec(Filtration.eternity())).columns(ImmutableList.of("m1")).resultFormat(ScanQuery.ResultFormat.RESULT_FORMAT_COMPACTED_LIST).context(QUERY_CONTEXT_DEFAULT).build()), "j0.", equalsCondition(DruidExpression.ofColumn(ColumnType.FLOAT, "m1"), DruidExpression.ofColumn(ColumnType.FLOAT, "j0.m1")), JoinType.INNER)).setInterval(querySegmentSpec(Filtration.eternity())).setDimensions(new DefaultDimensionSpec("dim2", "d0", ColumnType.STRING)).setGranularity(Granularities.ALL).setAggregatorSpecs(useDefault ? aggregators(new DoubleSumAggregatorFactory("a0:sum", "m2"), new CountAggregatorFactory("a0:count")) : aggregators(new DoubleSumAggregatorFactory("a0:sum", "m2"), new FilteredAggregatorFactory(new CountAggregatorFactory("a0:count"), not(selector("m2", null, null))))).setPostAggregatorSpecs(ImmutableList.of(new ArithmeticPostAggregator("a0", "quotient", ImmutableList.of(new FieldAccessPostAggregator(null, "a0:sum"), new FieldAccessPostAggregator(null, "a0:count"))))).setContext(QUERY_CONTEXT_DEFAULT).build()), NullHandling.sqlCompatible() ? ImmutableList.of(new Object[] { null, 4.0 }, new Object[] { "", 3.0 }, new Object[] { "a", 2.5 }, new Object[] { "abc", 5.0 }) : ImmutableList.of(new Object[] { "", 3.6666666666666665 }, new Object[] { "a", 2.5 }, new Object[] { "abc", 5.0 }));
}
Also used : FilteredAggregatorFactory(org.apache.druid.query.aggregation.FilteredAggregatorFactory) ArithmeticPostAggregator(org.apache.druid.query.aggregation.post.ArithmeticPostAggregator) FieldAccessPostAggregator(org.apache.druid.query.aggregation.post.FieldAccessPostAggregator) QueryDataSource(org.apache.druid.query.QueryDataSource) DoubleSumAggregatorFactory(org.apache.druid.query.aggregation.DoubleSumAggregatorFactory) CountAggregatorFactory(org.apache.druid.query.aggregation.CountAggregatorFactory) DefaultDimensionSpec(org.apache.druid.query.dimension.DefaultDimensionSpec) Test(org.junit.Test)

Example 12 with FilteredAggregatorFactory

use of org.apache.druid.query.aggregation.FilteredAggregatorFactory in project druid by druid-io.

the class CalciteCorrelatedQueryTest method testCorrelatedSubqueryWithCorrelatedQueryFilter.

@Test
@Parameters(source = QueryContextForJoinProvider.class)
public void testCorrelatedSubqueryWithCorrelatedQueryFilter(Map<String, Object> queryContext) throws Exception {
    cannotVectorize();
    queryContext = withLeftDirectAccessEnabled(queryContext);
    testQuery("select country, ANY_VALUE(\n" + "        select max(\"users\") from (\n" + "            select floor(__time to day), count(user) \"users\" from visits f where f.country = visits.country and f.city = 'A' group by 1\n" + "        )\n" + "     ) as \"dailyVisits\"\n" + "from visits \n" + " where city = 'B'" + " group by 1", queryContext, ImmutableList.of(GroupByQuery.builder().setDataSource(join(new TableDataSource(CalciteTests.USERVISITDATASOURCE), new QueryDataSource(GroupByQuery.builder().setDataSource(GroupByQuery.builder().setDataSource(CalciteTests.USERVISITDATASOURCE).setQuerySegmentSpec(querySegmentSpec(Intervals.ETERNITY)).setVirtualColumns(new ExpressionVirtualColumn("v0", "timestamp_floor(\"__time\",'P1D',null,'UTC')", ColumnType.LONG, TestExprMacroTable.INSTANCE)).setDimensions(new DefaultDimensionSpec("v0", "d0", ColumnType.LONG), new DefaultDimensionSpec("country", "d1")).setAggregatorSpecs(new FilteredAggregatorFactory(new CountAggregatorFactory("a0"), not(selector("user", null, null)))).setDimFilter(and(selector("city", "A", null), not(selector("country", null, null)))).setContext(withTimestampResultContext(queryContext, "d0", Granularities.DAY)).setGranularity(new AllGranularity()).build()).setQuerySegmentSpec(querySegmentSpec(Intervals.ETERNITY)).setDimensions(new DefaultDimensionSpec("d1", "_d0")).setAggregatorSpecs(new LongMaxAggregatorFactory("_a0", "a0")).setGranularity(new AllGranularity()).setContext(queryContext).build()), "j0.", equalsCondition(makeColumnExpression("country"), makeColumnExpression("j0._d0")), JoinType.LEFT, selector("city", "B", null))).setQuerySegmentSpec(querySegmentSpec(Intervals.ETERNITY)).setDimensions(new DefaultDimensionSpec("country", "d0")).setAggregatorSpecs(new LongAnyAggregatorFactory("a0", "j0._a0")).setGranularity(new AllGranularity()).setContext(queryContext).build()), ImmutableList.of(new Object[] { "canada", 2L }));
}
Also used : FilteredAggregatorFactory(org.apache.druid.query.aggregation.FilteredAggregatorFactory) ExpressionVirtualColumn(org.apache.druid.segment.virtual.ExpressionVirtualColumn) TableDataSource(org.apache.druid.query.TableDataSource) QueryDataSource(org.apache.druid.query.QueryDataSource) CountAggregatorFactory(org.apache.druid.query.aggregation.CountAggregatorFactory) LongAnyAggregatorFactory(org.apache.druid.query.aggregation.any.LongAnyAggregatorFactory) AllGranularity(org.apache.druid.java.util.common.granularity.AllGranularity) LongMaxAggregatorFactory(org.apache.druid.query.aggregation.LongMaxAggregatorFactory) DefaultDimensionSpec(org.apache.druid.query.dimension.DefaultDimensionSpec) Parameters(junitparams.Parameters) Test(org.junit.Test)

Example 13 with FilteredAggregatorFactory

use of org.apache.druid.query.aggregation.FilteredAggregatorFactory in project druid by druid-io.

the class CalciteCorrelatedQueryTest method testCorrelatedSubquery.

@Test
@Parameters(source = QueryContextForJoinProvider.class)
public void testCorrelatedSubquery(Map<String, Object> queryContext) throws Exception {
    cannotVectorize();
    queryContext = withLeftDirectAccessEnabled(queryContext);
    testQuery("select country, ANY_VALUE(\n" + "        select avg(\"users\") from (\n" + "            select floor(__time to day), count(distinct user) \"users\" from visits f where f.country = visits.country group by 1\n" + "        )\n" + "     ) as \"DAU\"\n" + "from visits \n" + "group by 1", queryContext, ImmutableList.of(GroupByQuery.builder().setDataSource(join(new TableDataSource(CalciteTests.USERVISITDATASOURCE), new QueryDataSource(GroupByQuery.builder().setDataSource(GroupByQuery.builder().setDataSource(CalciteTests.USERVISITDATASOURCE).setQuerySegmentSpec(querySegmentSpec(Intervals.ETERNITY)).setVirtualColumns(new ExpressionVirtualColumn("v0", "timestamp_floor(\"__time\",'P1D',null,'UTC')", ColumnType.LONG, TestExprMacroTable.INSTANCE)).setDimFilter(not(selector("country", null, null))).setDimensions(new DefaultDimensionSpec("v0", "d0", ColumnType.LONG), new DefaultDimensionSpec("country", "d1")).setAggregatorSpecs(new CardinalityAggregatorFactory("a0:a", null, Collections.singletonList(new DefaultDimensionSpec("user", "user")), false, true)).setPostAggregatorSpecs(Collections.singletonList(new HyperUniqueFinalizingPostAggregator("a0", "a0:a"))).setContext(withTimestampResultContext(queryContext, "d0", Granularities.DAY)).setGranularity(new AllGranularity()).build()).setQuerySegmentSpec(querySegmentSpec(Intervals.ETERNITY)).setDimensions(new DefaultDimensionSpec("d1", "_d0")).setAggregatorSpecs(new LongSumAggregatorFactory("_a0:sum", "a0"), useDefault ? new CountAggregatorFactory("_a0:count") : new FilteredAggregatorFactory(new CountAggregatorFactory("_a0:count"), not(selector("a0", null, null)))).setPostAggregatorSpecs(Collections.singletonList(new ArithmeticPostAggregator("_a0", "quotient", Arrays.asList(new FieldAccessPostAggregator(null, "_a0:sum"), new FieldAccessPostAggregator(null, "_a0:count"))))).setGranularity(new AllGranularity()).setContext(queryContext).build()), "j0.", equalsCondition(makeColumnExpression("country"), makeColumnExpression("j0._d0")), JoinType.LEFT)).setQuerySegmentSpec(querySegmentSpec(Intervals.ETERNITY)).setDimensions(new DefaultDimensionSpec("country", "d0")).setAggregatorSpecs(new LongAnyAggregatorFactory("a0", "j0._a0")).setGranularity(new AllGranularity()).setContext(queryContext).build()), ImmutableList.of(new Object[] { "India", 2L }, new Object[] { "USA", 1L }, new Object[] { "canada", 3L }));
}
Also used : FilteredAggregatorFactory(org.apache.druid.query.aggregation.FilteredAggregatorFactory) ArithmeticPostAggregator(org.apache.druid.query.aggregation.post.ArithmeticPostAggregator) FieldAccessPostAggregator(org.apache.druid.query.aggregation.post.FieldAccessPostAggregator) AllGranularity(org.apache.druid.java.util.common.granularity.AllGranularity) LongSumAggregatorFactory(org.apache.druid.query.aggregation.LongSumAggregatorFactory) HyperUniqueFinalizingPostAggregator(org.apache.druid.query.aggregation.hyperloglog.HyperUniqueFinalizingPostAggregator) DefaultDimensionSpec(org.apache.druid.query.dimension.DefaultDimensionSpec) ExpressionVirtualColumn(org.apache.druid.segment.virtual.ExpressionVirtualColumn) TableDataSource(org.apache.druid.query.TableDataSource) QueryDataSource(org.apache.druid.query.QueryDataSource) CountAggregatorFactory(org.apache.druid.query.aggregation.CountAggregatorFactory) LongAnyAggregatorFactory(org.apache.druid.query.aggregation.any.LongAnyAggregatorFactory) CardinalityAggregatorFactory(org.apache.druid.query.aggregation.cardinality.CardinalityAggregatorFactory) Parameters(junitparams.Parameters) Test(org.junit.Test)

Example 14 with FilteredAggregatorFactory

use of org.apache.druid.query.aggregation.FilteredAggregatorFactory in project druid by druid-io.

the class CalciteArraysQueryTest method testArrayAgg.

@Test
public void testArrayAgg() throws Exception {
    cannotVectorize();
    testQuery("SELECT ARRAY_AGG(dim1), ARRAY_AGG(DISTINCT dim1), ARRAY_AGG(DISTINCT dim1) FILTER(WHERE dim1 = 'shazbot') FROM foo WHERE dim1 is not null", ImmutableList.of(Druids.newTimeseriesQueryBuilder().dataSource(CalciteTests.DATASOURCE1).intervals(querySegmentSpec(Filtration.eternity())).granularity(Granularities.ALL).filters(not(selector("dim1", null, null))).aggregators(aggregators(new ExpressionLambdaAggregatorFactory("a0", ImmutableSet.of("dim1"), "__acc", "ARRAY<STRING>[]", "ARRAY<STRING>[]", true, true, false, "array_append(\"__acc\", \"dim1\")", "array_concat(\"__acc\", \"a0\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), new ExpressionLambdaAggregatorFactory("a1", ImmutableSet.of("dim1"), "__acc", "ARRAY<STRING>[]", "ARRAY<STRING>[]", true, true, false, "array_set_add(\"__acc\", \"dim1\")", "array_set_add_all(\"__acc\", \"a1\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a2", ImmutableSet.of("dim1"), "__acc", "ARRAY<STRING>[]", "ARRAY<STRING>[]", true, true, false, "array_set_add(\"__acc\", \"dim1\")", "array_set_add_all(\"__acc\", \"a2\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), selector("dim1", "shazbot", null)))).context(QUERY_CONTEXT_DEFAULT).build()), ImmutableList.of(useDefault ? new Object[] { "[\"10.1\",\"2\",\"1\",\"def\",\"abc\"]", "[\"1\",\"10.1\",\"2\",\"abc\",\"def\"]", null } : new Object[] { "[\"\",\"10.1\",\"2\",\"1\",\"def\",\"abc\"]", "[\"\",\"1\",\"10.1\",\"2\",\"abc\",\"def\"]", null }));
}
Also used : FilteredAggregatorFactory(org.apache.druid.query.aggregation.FilteredAggregatorFactory) ExpressionLambdaAggregatorFactory(org.apache.druid.query.aggregation.ExpressionLambdaAggregatorFactory) Test(org.junit.Test)

Example 15 with FilteredAggregatorFactory

use of org.apache.druid.query.aggregation.FilteredAggregatorFactory in project druid by druid-io.

the class TimeseriesQueryRunnerTest method testTimeSeriesWithFilteredAggInvertedNullValue.

@Test
public void testTimeSeriesWithFilteredAggInvertedNullValue() {
    TimeseriesQuery query = Druids.newTimeseriesQueryBuilder().dataSource(QueryRunnerTestHelper.DATA_SOURCE).granularity(QueryRunnerTestHelper.ALL_GRAN).intervals(QueryRunnerTestHelper.FIRST_TO_THIRD).aggregators(Lists.newArrayList(Iterables.concat(aggregatorFactoryList, Collections.singletonList(new FilteredAggregatorFactory(new CountAggregatorFactory("filteredAgg"), new NotDimFilter(new SelectorDimFilter(QueryRunnerTestHelper.MARKET_DIMENSION, null, null))))))).postAggregators(QueryRunnerTestHelper.ADD_ROWS_INDEX_CONSTANT).descending(descending).context(makeContext()).build();
    Iterable<Result<TimeseriesResultValue>> actualResults = runner.run(QueryPlus.wrap(query)).toList();
    List<Result<TimeseriesResultValue>> expectedResults = Collections.singletonList(new Result<>(DateTimes.of("2011-04-01"), new TimeseriesResultValue(ImmutableMap.of("filteredAgg", 26L, "addRowsIndexConstant", 12486.361190795898d, "index", 12459.361190795898d, "uniques", 9.019833517963864d, "rows", 26L))));
    assertExpectedResults(expectedResults, actualResults);
}
Also used : FilteredAggregatorFactory(org.apache.druid.query.aggregation.FilteredAggregatorFactory) NotDimFilter(org.apache.druid.query.filter.NotDimFilter) CountAggregatorFactory(org.apache.druid.query.aggregation.CountAggregatorFactory) SelectorDimFilter(org.apache.druid.query.filter.SelectorDimFilter) Result(org.apache.druid.query.Result) InitializedNullHandlingTest(org.apache.druid.testing.InitializedNullHandlingTest) Test(org.junit.Test)

Aggregations

FilteredAggregatorFactory (org.apache.druid.query.aggregation.FilteredAggregatorFactory)63 Test (org.junit.Test)52 CountAggregatorFactory (org.apache.druid.query.aggregation.CountAggregatorFactory)28 LongSumAggregatorFactory (org.apache.druid.query.aggregation.LongSumAggregatorFactory)23 DefaultDimensionSpec (org.apache.druid.query.dimension.DefaultDimensionSpec)23 SelectorDimFilter (org.apache.druid.query.filter.SelectorDimFilter)20 QueryDataSource (org.apache.druid.query.QueryDataSource)14 InitializedNullHandlingTest (org.apache.druid.testing.InitializedNullHandlingTest)14 ExpressionLambdaAggregatorFactory (org.apache.druid.query.aggregation.ExpressionLambdaAggregatorFactory)13 NotDimFilter (org.apache.druid.query.filter.NotDimFilter)11 AggregatorFactory (org.apache.druid.query.aggregation.AggregatorFactory)10 BaseCalciteQueryTest (org.apache.druid.sql.calcite.BaseCalciteQueryTest)10 ArithmeticPostAggregator (org.apache.druid.query.aggregation.post.ArithmeticPostAggregator)8 FieldAccessPostAggregator (org.apache.druid.query.aggregation.post.FieldAccessPostAggregator)8 BoundDimFilter (org.apache.druid.query.filter.BoundDimFilter)8 MultipleIntervalSegmentSpec (org.apache.druid.query.spec.MultipleIntervalSegmentSpec)8 Result (org.apache.druid.query.Result)7 LongMaxAggregatorFactory (org.apache.druid.query.aggregation.LongMaxAggregatorFactory)7 TableDataSource (org.apache.druid.query.TableDataSource)6 DimFilter (org.apache.druid.query.filter.DimFilter)6