Examples with ResultRow org.apache.druid.query.groupby.ResultRow used on opensource projects

Example 61 with ResultRow

use of org.apache.druid.query.groupby.ResultRow in project druid by druid-io.

the class ArrayOfDoublesSketchAggregationTest method buildingSketchesAtQueryTimeWithNullsTest.

// Three buckets with null values
@Test
public void buildingSketchesAtQueryTimeWithNullsTest() throws Exception {
    Sequence<ResultRow> seq = helper.createIndexAndRunQueryOnSegment(new File(this.getClass().getClassLoader().getResource("tuple/array_of_doubles_build_data_three_values_and_nulls.tsv").getFile()), String.join("\n", "{", "  \"type\": \"string\",", "  \"parseSpec\": {", "    \"format\": \"tsv\",", "    \"timestampSpec\": {\"column\": \"timestamp\", \"format\": \"yyyyMMddHH\"},", "    \"dimensionsSpec\": {", "      \"dimensions\": [\"product\", \"key\"],", "      \"dimensionExclusions\": [],", "      \"spatialDimensions\": []", "    },", "    \"columns\": [\"timestamp\", \"product\", \"key\", \"value1\", \"value2\", \"value3\"]", "  }", "}"), String.join("\n", "[", "  {\"type\": \"doubleSum\", \"name\": \"value1\", \"fieldName\": \"value1\"},", "  {\"type\": \"doubleSum\", \"name\": \"value2\", \"fieldName\": \"value2\"},", "  {\"type\": \"doubleSum\", \"name\": \"value3\", \"fieldName\": \"value3\"}", "]"), // minTimestamp
    0, Granularities.NONE, // maxRowCount
    10, String.join("\n", "{", "  \"queryType\": \"groupBy\",", "  \"dataSource\": \"test_datasource\",", "  \"granularity\": \"ALL\",", "  \"dimensions\": [],", "  \"virtualColumns\": [{\"type\": \"expression\",\"name\": \"nonulls3\",\"expression\": \"nvl(value3, 0.0)\",\"outputType\": \"DOUBLE\"}],", "  \"aggregations\": [", "   {\"type\": \"arrayOfDoublesSketch\", \"name\": \"sketch\", \"fieldName\": \"key\", \"metricColumns\": [ \"value1\", \"value2\", \"value3\" ], \"nominalEntries\": 1024},", "   {\"type\": \"arrayOfDoublesSketch\", \"name\": \"sketchNoNulls\", \"fieldName\": \"key\", \"metricColumns\": [ \"value1\", \"value2\", \"nonulls3\" ], \"nominalEntries\": 1024}", "  ],", "  \"postAggregations\": [", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"estimate\", \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}},", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"estimateNoNulls\", \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketchNoNulls\"}},", "    {\"type\": \"arrayOfDoublesSketchToQuantilesSketch\", \"name\": \"quantiles-sketch\", \"column\": 2, \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}},", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"union\", \"field\": {", "      \"type\": \"arrayOfDoublesSketchSetOp\",", "      \"name\": \"union\",", "      \"operation\": \"UNION\",", "      \"nominalEntries\": 1024,", "      \"numberOfValues\": 3,", "      \"fields\": [{\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}, {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}]", "    }},", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"intersection\", \"field\": {", "      \"type\": \"arrayOfDoublesSketchSetOp\",", "      \"name\": \"intersection\",", "      \"operation\": \"INTERSECT\",", "      \"nominalEntries\": 1024,", "      \"numberOfValues\": 3,", "      \"fields\": [{\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}, {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}]", "    }},", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"anotb\", \"field\": {", "      \"type\": \"arrayOfDoublesSketchSetOp\",", "      \"name\": \"anotb\",", "      \"operation\": \"NOT\",", "      \"nominalEntries\": 1024,", "      \"numberOfValues\": 3,", "      \"fields\": [{\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}, {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}]", "    }},", "    {", "      \"type\": \"arrayOfDoublesSketchToMeans\",", "      \"name\": \"means\",", "      \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}", "    },", "    {\"type\": \"arrayOfDoublesSketchToQuantilesSketch\", \"name\": \"quantiles-sketch-with-nulls\", \"column\": 3, \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}},", "    {\"type\": \"arrayOfDoublesSketchToQuantilesSketch\", \"name\": \"quantiles-sketch-with-no-nulls\", \"column\": 3, \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketchNoNulls\"}}", "  ],", "  \"intervals\": [\"2015-01-01T00:00:00.000Z/2015-01-31T00:00:00.000Z\"]", "}"));
    List<ResultRow> results = seq.toList();
    Assert.assertEquals(1, results.size());
    ResultRow row = results.get(0);
    Assert.assertEquals("sketch", NullHandling.replaceWithDefault() ? 40.0 : 30.0, (double) row.get(0), 0);
    Assert.assertEquals("sketchNoNulls", 40.0, (double) row.get(1), 0);
    Assert.assertEquals("estimate", NullHandling.replaceWithDefault() ? 40.0 : 30.0, (double) row.get(2), 0);
    Assert.assertEquals("estimateNoNulls", 40.0, (double) row.get(3), 0);
    Assert.assertEquals("union", NullHandling.replaceWithDefault() ? 40.0 : 30.0, (double) row.get(5), 0);
    Assert.assertEquals("intersection", NullHandling.replaceWithDefault() ? 40.0 : 30.0, (double) row.get(6), 0);
    Assert.assertEquals("anotb", 0, (double) row.get(7), 0);
    // means
    Object meansObj = row.get(8);
    Assert.assertTrue(meansObj instanceof double[]);
    double[] means = (double[]) meansObj;
    Assert.assertEquals(3, means.length);
    Assert.assertEquals(1.0, means[0], 0);
    Assert.assertEquals(2.0, means[1], 0);
    Assert.assertEquals(NullHandling.replaceWithDefault() ? 2.25 : 3.0, means[2], 0.1);
    // quantiles-sketch
    Object obj = row.get(4);
    Assert.assertTrue(obj instanceof DoublesSketch);
    DoublesSketch ds = (DoublesSketch) obj;
    Assert.assertEquals(NullHandling.replaceWithDefault() ? 40 : 30, ds.getN());
    Assert.assertEquals(2.0, ds.getMinValue(), 0);
    Assert.assertEquals(2.0, ds.getMaxValue(), 0);
    // quantiles-sketch-with-nulls
    Object objSketch2 = row.get(9);
    Assert.assertTrue(objSketch2 instanceof DoublesSketch);
    DoublesSketch ds2 = (DoublesSketch) objSketch2;
    Assert.assertEquals(NullHandling.replaceWithDefault() ? 40 : 30, ds2.getN());
    Assert.assertEquals(NullHandling.replaceWithDefault() ? 0.0 : 3.0, ds2.getMinValue(), 0);
    Assert.assertEquals(3.0, ds2.getMaxValue(), 0);
    // quantiles-sketch-no-nulls
    Object objSketch3 = row.get(10);
    Assert.assertTrue(objSketch3 instanceof DoublesSketch);
    DoublesSketch ds3 = (DoublesSketch) objSketch3;
    Assert.assertEquals(40, ds3.getN());
    Assert.assertEquals(0.0, ds3.getMinValue(), 0);
    Assert.assertEquals(3.0, ds3.getMaxValue(), 0);
}

Example 62 with ResultRow

use of org.apache.druid.query.groupby.ResultRow in project druid by druid-io.

the class ArrayOfDoublesSketchAggregationTest method buildingSketchesAtQueryTimeTwoBucketsTest.

// Two buckets with statistically significant difference.
// See GenerateTestData class for details.
@Test
public void buildingSketchesAtQueryTimeTwoBucketsTest() throws Exception {
    Sequence<ResultRow> seq = helper.createIndexAndRunQueryOnSegment(new File(this.getClass().getClassLoader().getResource("tuple/bucket_test_data.tsv").getFile()), String.join("\n", "{", "  \"type\": \"string\",", "  \"parseSpec\": {", "    \"format\": \"tsv\",", "    \"timestampSpec\": {\"column\": \"timestamp\", \"format\": \"yyyyMMdd\"},", "    \"dimensionsSpec\": {", "      \"dimensions\": [\"label\", \"userid\"]", "    },", "    \"columns\": [\"timestamp\", \"label\", \"userid\", \"parameter\"]", "  }", "}"), String.join("\n", "[", "  {\"type\": \"doubleSum\", \"name\": \"parameter\", \"fieldName\": \"parameter\"}", "]"), // minTimestamp
    0, Granularities.NONE, // maxRowCount
    2000, String.join("\n", "{", "  \"queryType\": \"groupBy\",", "  \"dataSource\": \"test_datasource\",", "  \"granularity\": \"ALL\",", "  \"dimensions\": [],", "  \"aggregations\": [", "    {", "      \"type\": \"filtered\",", "      \"filter\": {\"type\": \"selector\", \"dimension\": \"label\", \"value\": \"test\"},", "      \"aggregator\": {\"type\": \"arrayOfDoublesSketch\", \"name\": \"sketch-test\", \"fieldName\": \"userid\", \"metricColumns\": [\"parameter\"]}", "    },", "    {", "      \"type\": \"filtered\",", "      \"filter\": {\"type\": \"selector\", \"dimension\": \"label\", \"value\": \"control\"},", "      \"aggregator\": {\"type\": \"arrayOfDoublesSketch\", \"name\": \"sketch-control\", \"fieldName\": \"userid\", \"metricColumns\": [\"parameter\"]}", "    }", "  ],", "  \"postAggregations\": [", "    {\"type\": \"arrayOfDoublesSketchTTest\",", "      \"name\": \"p-value\", \"fields\": [", "        {\"type\": \"fieldAccess\", \"fieldName\": \"sketch-test\"},", "        {\"type\": \"fieldAccess\", \"fieldName\": \"sketch-control\"}", "      ]", "    }", "  ],", "  \"intervals\": [\"2017-01-01T00:00:00.000Z/2017-01-31T00:00:00.000Z\"]", "}"));
    List<ResultRow> results = seq.toList();
    Assert.assertEquals(1, results.size());
    ResultRow row = results.get(0);
    // p-value
    Object obj = row.get(2);
    Assert.assertTrue(obj instanceof double[]);
    double[] array = (double[]) obj;
    Assert.assertEquals(1, array.length);
    double pValue = array[0];
    // Test and control buckets were constructed to have different means, so we
    // expect very low p value
    Assert.assertEquals(0, pValue, 0.001);
}

Example 63 with ResultRow

use of org.apache.druid.query.groupby.ResultRow in project druid by druid-io.

the class ArrayOfDoublesSketchAggregationTest method buildingSketchesAtIngestionTimeTwoValuesAndNumericalKey.

@Test
public void buildingSketchesAtIngestionTimeTwoValuesAndNumericalKey() throws Exception {
    Sequence<ResultRow> seq = helper.createIndexAndRunQueryOnSegment(new File(this.getClass().getClassLoader().getResource("tuple/array_of_doubles_build_data_two_values_and_key_as_number.tsv").getFile()), String.join("\n", "{", "  \"type\": \"string\",", "  \"parseSpec\": {", "    \"format\": \"tsv\",", "    \"timestampSpec\": {\"column\": \"timestamp\", \"format\": \"yyyyMMddHH\"},", "    \"dimensionsSpec\": {", "      \"dimensions\": [\"product\", {\"type\": \"long\", \"name\": \"key_num\"}],", "      \"dimensionExclusions\": [],", "      \"spatialDimensions\": []", "    },", "    \"columns\": [\"timestamp\", \"product\", \"key\", \"key_num\", \"value1\", \"value2\"]", "  }", "}"), String.join("\n", "[", "  {\"type\": \"arrayOfDoublesSketch\", \"name\": \"sketch\", \"fieldName\": \"key_num\", \"metricColumns\": [ \"value1\", \"value2\" ], \"nominalEntries\": 1024}", "]"), // minTimestamp
    0, Granularities.NONE, // maxRowCount
    10, String.join("\n", "{", "  \"queryType\": \"groupBy\",", "  \"dataSource\": \"test_datasource\",", "  \"granularity\": \"ALL\",", "  \"dimensions\": [],", "  \"aggregations\": [", "    {\"type\": \"arrayOfDoublesSketch\", \"name\": \"sketch\", \"fieldName\": \"sketch\", \"nominalEntries\": 1024, \"numberOfValues\": 2}", "  ],", "  \"postAggregations\": [", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"estimate\", \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}},", "    {\"type\": \"arrayOfDoublesSketchToQuantilesSketch\", \"name\": \"quantiles-sketch\", \"column\": 2, \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}},", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"union\", \"field\": {", "      \"type\": \"arrayOfDoublesSketchSetOp\",", "      \"name\": \"union\",", "      \"operation\": \"UNION\",", "      \"nominalEntries\": 1024,", "      \"numberOfValues\": 2,", "      \"fields\": [{\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}, {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}]", "    }},", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"intersection\", \"field\": {", "      \"type\": \"arrayOfDoublesSketchSetOp\",", "      \"name\": \"intersection\",", "      \"operation\": \"INTERSECT\",", "      \"nominalEntries\": 1024,", "      \"numberOfValues\": 2,", "      \"fields\": [{\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}, {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}]", "    }},", "    {\"type\": \"arrayOfDoublesSketchToEstimate\", \"name\": \"anotb\", \"field\": {", "      \"type\": \"arrayOfDoublesSketchSetOp\",", "      \"name\": \"anotb\",", "      \"operation\": \"NOT\",", "      \"nominalEntries\": 1024,", "      \"numberOfValues\": 2,", "      \"fields\": [{\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}, {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}]", "    }},", "    {", "      \"type\": \"arrayOfDoublesSketchToMeans\",", "      \"name\": \"means\",", "      \"field\": {\"type\": \"fieldAccess\", \"fieldName\": \"sketch\"}", "    }", "  ],", "  \"intervals\": [\"2015-01-01T00:00:00.000Z/2015-01-31T00:00:00.000Z\"]", "}"));
    List<ResultRow> results = seq.toList();
    Assert.assertEquals(1, results.size());
    ResultRow row = results.get(0);
    Assert.assertEquals("sketch", 40.0, (double) row.get(0), 0);
    Assert.assertEquals("estimate", 40.0, (double) row.get(1), 0);
    Assert.assertEquals("union", 40.0, (double) row.get(3), 0);
    Assert.assertEquals("intersection", 40.0, (double) row.get(4), 0);
    Assert.assertEquals("anotb", 0, (double) row.get(5), 0);
    // means
    Object meansObj = row.get(6);
    Assert.assertTrue(meansObj instanceof double[]);
    double[] means = (double[]) meansObj;
    Assert.assertEquals(2, means.length);
    Assert.assertEquals(1.0, means[0], 0);
    Assert.assertEquals(2.0, means[1], 0);
    // quantiles-sketch
    Object obj = row.get(2);
    Assert.assertTrue(obj instanceof DoublesSketch);
    DoublesSketch ds = (DoublesSketch) obj;
    Assert.assertEquals(40, ds.getN());
    Assert.assertEquals(2.0, ds.getMinValue(), 0);
    Assert.assertEquals(2.0, ds.getMaxValue(), 0);
}

Example 64 with ResultRow

use of org.apache.druid.query.groupby.ResultRow in project druid by druid-io.

the class FixedBucketsHistogramGroupByQueryTest method constructorFeeder.

@Parameterized.Parameters(name = "{0}")
public static Iterable<Object[]> constructorFeeder() {
    final GroupByQueryConfig v1Config = new GroupByQueryConfig() {

        @Override
        public String getDefaultStrategy() {
            return GroupByStrategySelector.STRATEGY_V1;
        }

        @Override
        public String toString() {
            return "v1";
        }
    };
    final GroupByQueryConfig v1SingleThreadedConfig = new GroupByQueryConfig() {

        @Override
        public boolean isSingleThreaded() {
            return true;
        }

        @Override
        public String getDefaultStrategy() {
            return GroupByStrategySelector.STRATEGY_V1;
        }

        @Override
        public String toString() {
            return "v1SingleThreaded";
        }
    };
    final GroupByQueryConfig v2Config = new GroupByQueryConfig() {

        @Override
        public String getDefaultStrategy() {
            return GroupByStrategySelector.STRATEGY_V2;
        }

        @Override
        public String toString() {
            return "v2";
        }
    };
    v1Config.setMaxIntermediateRows(10000);
    v1SingleThreadedConfig.setMaxIntermediateRows(10000);
    final List<Object[]> constructors = new ArrayList<>();
    final List<GroupByQueryConfig> configs = ImmutableList.of(v1Config, v1SingleThreadedConfig, v2Config);
    for (GroupByQueryConfig config : configs) {
        final Pair<GroupByQueryRunnerFactory, Closer> factoryAndCloser = GroupByQueryRunnerTest.makeQueryRunnerFactory(config);
        final GroupByQueryRunnerFactory factory = factoryAndCloser.lhs;
        RESOURCE_CLOSER.register(factoryAndCloser.rhs);
        for (QueryRunner<ResultRow> runner : QueryRunnerTestHelper.makeQueryRunners(factory)) {
            final String testName = StringUtils.format("config=%s, runner=%s", config.toString(), runner.toString());
            constructors.add(new Object[] { testName, factory, runner });
        }
    }
    return constructors;
}

Example 65 with ResultRow

use of org.apache.druid.query.groupby.ResultRow in project druid by druid-io.

the class GroupByBenchmark method querySingleQueryableIndex.

@Benchmark
@BenchmarkMode(Mode.AverageTime)
@OutputTimeUnit(TimeUnit.MICROSECONDS)
public void querySingleQueryableIndex(Blackhole blackhole, QueryableIndexState state) {
    QueryRunner<ResultRow> runner = QueryBenchmarkUtil.makeQueryRunner(factory, SegmentId.dummy("qIndex"), new QueryableIndexSegment(state.queryableIndexes.get(0), SegmentId.dummy("qIndex")));
    final Sequence<ResultRow> results = GroupByBenchmark.runQuery(factory, runner, query);
    final ResultRow lastRow = results.accumulate(null, (accumulated, in) -> in);
    blackhole.consume(lastRow);
}