Examples with MultipleIntervalSegmentSpec org.apache.druid.query.spec.MultipleIntervalSegmentSpec used on opensource projects

Example 66 with MultipleIntervalSegmentSpec

use of org.apache.druid.query.spec.MultipleIntervalSegmentSpec in project druid by druid-io.

the class FixedBucketsHistogramQuantileSqlAggregatorTest method testQuantileOnComplexColumn.

@Test
public void testQuantileOnComplexColumn() throws Exception {
    cannotVectorize();
    final List<Object[]> expectedResults = ImmutableList.of(new Object[] { 1.0299999713897705, 3.5, 6.293333530426025, 6.470000267028809, 6.494999885559082, 5.497499942779541, 6.499499797821045 });
    testQuery("SELECT\n" + "APPROX_QUANTILE_FIXED_BUCKETS(fbhist_m1, 0.01, 20, 0.0, 10.0),\n" + "APPROX_QUANTILE_FIXED_BUCKETS(fbhist_m1, 0.5, 20, 0.0, 10.0),\n" + "APPROX_QUANTILE_FIXED_BUCKETS(fbhist_m1, 0.98, 30, 0.0, 10.0),\n" + "APPROX_QUANTILE_FIXED_BUCKETS(fbhist_m1, 0.99, 20, 0.0, 10.0),\n" + "APPROX_QUANTILE_FIXED_BUCKETS(fbhist_m1, 0.99, 20, 0.0, 10.0) FILTER(WHERE dim1 = 'abc'),\n" + "APPROX_QUANTILE_FIXED_BUCKETS(fbhist_m1, 0.999, 20, 0.0, 10.0) FILTER(WHERE dim1 <> 'abc'),\n" + "APPROX_QUANTILE_FIXED_BUCKETS(fbhist_m1, 0.999, 20, 0.0, 10.0) FILTER(WHERE dim1 = 'abc')\n" + "FROM foo", ImmutableList.of(Druids.newTimeseriesQueryBuilder().dataSource(CalciteTests.DATASOURCE1).intervals(new MultipleIntervalSegmentSpec(ImmutableList.of(Filtration.eternity()))).granularity(Granularities.ALL).aggregators(ImmutableList.of(new FixedBucketsHistogramAggregatorFactory("a0:agg", "fbhist_m1", 20, 0.0, 10.0, FixedBucketsHistogram.OutlierHandlingMode.IGNORE, false), new FixedBucketsHistogramAggregatorFactory("a2:agg", "fbhist_m1", 30, 0.0, 10.0, FixedBucketsHistogram.OutlierHandlingMode.IGNORE, false), new FilteredAggregatorFactory(new FixedBucketsHistogramAggregatorFactory("a4:agg", "fbhist_m1", 20, 0.0, 10.0, FixedBucketsHistogram.OutlierHandlingMode.IGNORE, false), new SelectorDimFilter("dim1", "abc", null)), new FilteredAggregatorFactory(new FixedBucketsHistogramAggregatorFactory("a5:agg", "fbhist_m1", 20, 0.0, 10.0, FixedBucketsHistogram.OutlierHandlingMode.IGNORE, false), new NotDimFilter(new SelectorDimFilter("dim1", "abc", null))))).postAggregators(new QuantilePostAggregator("a0", "a0:agg", 0.01f), new QuantilePostAggregator("a1", "a0:agg", 0.50f), new QuantilePostAggregator("a2", "a2:agg", 0.98f), new QuantilePostAggregator("a3", "a0:agg", 0.99f), new QuantilePostAggregator("a4", "a4:agg", 0.99f), new QuantilePostAggregator("a5", "a5:agg", 0.999f), new QuantilePostAggregator("a6", "a4:agg", 0.999f)).context(QUERY_CONTEXT_DEFAULT).build()), expectedResults);
}

Example 67 with MultipleIntervalSegmentSpec

use of org.apache.druid.query.spec.MultipleIntervalSegmentSpec in project druid by druid-io.

the class QuantileSqlAggregatorTest method testQuantileOnCastedString.

@Test
public void testQuantileOnCastedString() throws Exception {
    cannotVectorize();
    final List<Object[]> expectedResults;
    if (NullHandling.replaceWithDefault()) {
        expectedResults = ImmutableList.of(new Object[] { "", 0.0d }, new Object[] { "a", 0.0d }, new Object[] { "b", 0.0d }, new Object[] { "c", 10.100000381469727d }, new Object[] { "d", 2.0d });
    } else {
        expectedResults = ImmutableList.of(new Object[] { null, Double.NaN }, new Object[] { "", 1.0d }, new Object[] { "a", Double.NaN }, new Object[] { "b", 10.100000381469727d }, new Object[] { "c", 10.100000381469727d }, new Object[] { "d", 2.0d });
    }
    testQuery("SELECT dim3, APPROX_QUANTILE(CAST(dim1 as DOUBLE), 0.5) from foo group by dim3", ImmutableList.of(GroupByQuery.builder().setDataSource(CalciteTests.DATASOURCE1).setInterval(new MultipleIntervalSegmentSpec(ImmutableList.of(Filtration.eternity()))).setGranularity(Granularities.ALL).setVirtualColumns(new ExpressionVirtualColumn("v0", "CAST(\"dim1\", 'DOUBLE')", ColumnType.FLOAT, ExprMacroTable.nil())).setDimensions(new DefaultDimensionSpec("dim3", "d0")).setAggregatorSpecs(new ApproximateHistogramAggregatorFactory("a0:agg", "v0", 50, 7, Float.NEGATIVE_INFINITY, Float.POSITIVE_INFINITY, false)).setPostAggregatorSpecs(ImmutableList.of(new QuantilePostAggregator("a0", "a0:agg", 0.5f))).setContext(QUERY_CONTEXT_DEFAULT).build()), expectedResults);
}

Example 68 with MultipleIntervalSegmentSpec

use of org.apache.druid.query.spec.MultipleIntervalSegmentSpec in project druid by druid-io.

the class ScanBenchmark method basicB.

private static Druids.ScanQueryBuilder basicB(final GeneratorSchemaInfo basicSchema) {
    final QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
    List<String> dimHyperUniqueFilterVals = new ArrayList<>();
    int numResults = (int) (100000 * 0.1);
    int step = 100000 / numResults;
    for (int i = 0; i < 100001 && dimHyperUniqueFilterVals.size() < numResults; i += step) {
        dimHyperUniqueFilterVals.add(String.valueOf(i));
    }
    DimFilter filter = new InDimFilter("dimHyperUnique", dimHyperUniqueFilterVals, null);
    return Druids.newScanQueryBuilder().filters(filter).dataSource("blah").intervals(intervalSpec).order(ordering);
}

Example 69 with MultipleIntervalSegmentSpec

use of org.apache.druid.query.spec.MultipleIntervalSegmentSpec in project druid by druid-io.

the class TimeseriesBenchmark method setupQueries.

private void setupQueries() {
    // queries for the basic schema
    Map<String, TimeseriesQuery> basicQueries = new LinkedHashMap<>();
    GeneratorSchemaInfo basicSchema = GeneratorBasicSchemas.SCHEMA_MAP.get("basic");
    {
        // basic.A
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(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"));
        TimeseriesQuery queryA = Druids.newTimeseriesQueryBuilder().dataSource("blah").granularity(Granularities.ALL).intervals(intervalSpec).aggregators(queryAggs).descending(descending).build();
        basicQueries.put("A", queryA);
    }
    {
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        LongSumAggregatorFactory lsaf = new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential");
        BoundDimFilter timeFilter = new BoundDimFilter(ColumnHolder.TIME_COLUMN_NAME, "200000", "300000", false, false, null, null, StringComparators.NUMERIC);
        queryAggs.add(new FilteredAggregatorFactory(lsaf, timeFilter));
        TimeseriesQuery timeFilterQuery = Druids.newTimeseriesQueryBuilder().dataSource("blah").granularity(Granularities.ALL).intervals(intervalSpec).aggregators(queryAggs).descending(descending).build();
        basicQueries.put("timeFilterNumeric", timeFilterQuery);
    }
    {
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        LongSumAggregatorFactory lsaf = new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential");
        BoundDimFilter timeFilter = new BoundDimFilter(ColumnHolder.TIME_COLUMN_NAME, "200000", "300000", false, false, null, null, StringComparators.ALPHANUMERIC);
        queryAggs.add(new FilteredAggregatorFactory(lsaf, timeFilter));
        TimeseriesQuery timeFilterQuery = Druids.newTimeseriesQueryBuilder().dataSource("blah").granularity(Granularities.ALL).intervals(intervalSpec).aggregators(queryAggs).descending(descending).build();
        basicQueries.put("timeFilterAlphanumeric", timeFilterQuery);
    }
    {
        QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.utc(200000, 300000)));
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        LongSumAggregatorFactory lsaf = new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential");
        queryAggs.add(lsaf);
        TimeseriesQuery timeFilterQuery = Druids.newTimeseriesQueryBuilder().dataSource("blah").granularity(Granularities.ALL).intervals(intervalSpec).aggregators(queryAggs).descending(descending).build();
        basicQueries.put("timeFilterByInterval", timeFilterQuery);
    }
    SCHEMA_QUERY_MAP.put("basic", basicQueries);
}

Example 70 with MultipleIntervalSegmentSpec

use of org.apache.druid.query.spec.MultipleIntervalSegmentSpec in project druid by druid-io.

the class TimeCompareBenchmark method setupQueries.

private void setupQueries() {
    // queries for the basic schema
    GeneratorSchemaInfo basicSchema = GeneratorBasicSchemas.SCHEMA_MAP.get("basic");
    QuerySegmentSpec intervalSpec = new MultipleIntervalSegmentSpec(Collections.singletonList(basicSchema.getDataInterval()));
    long startMillis = basicSchema.getDataInterval().getStartMillis();
    long endMillis = basicSchema.getDataInterval().getEndMillis();
    long half = (endMillis - startMillis) / 2;
    Interval recent = Intervals.utc(half, endMillis);
    Interval previous = Intervals.utc(startMillis, half);
    log.info("Recent interval: " + recent);
    log.info("Previous interval: " + previous);
    {
        // basic.topNTimeCompare
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new FilteredAggregatorFactory(// jsAgg1,
        new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"), new IntervalDimFilter(ColumnHolder.TIME_COLUMN_NAME, Collections.singletonList(recent), null)));
        queryAggs.add(new FilteredAggregatorFactory(new LongSumAggregatorFactory("_cmp_sumLongSequential", "sumLongSequential"), new IntervalDimFilter(ColumnHolder.TIME_COLUMN_NAME, Collections.singletonList(previous), null)));
        TopNQueryBuilder queryBuilderA = new TopNQueryBuilder().dataSource("blah").granularity(Granularities.ALL).dimension("dimUniform").metric("sumLongSequential").intervals(intervalSpec).aggregators(queryAggs).threshold(threshold);
        topNQuery = queryBuilderA.build();
        topNFactory = new TopNQueryRunnerFactory(new StupidPool<>("TopNBenchmark-compute-bufferPool", new OffheapBufferGenerator("compute", 250000000), 0, Integer.MAX_VALUE), new TopNQueryQueryToolChest(new TopNQueryConfig()), QueryBenchmarkUtil.NOOP_QUERYWATCHER);
    }
    {
        // basic.timeseriesTimeCompare
        List<AggregatorFactory> queryAggs = new ArrayList<>();
        queryAggs.add(new FilteredAggregatorFactory(new LongSumAggregatorFactory("sumLongSequential", "sumLongSequential"), new IntervalDimFilter(ColumnHolder.TIME_COLUMN_NAME, Collections.singletonList(recent), null)));
        queryAggs.add(new FilteredAggregatorFactory(new LongSumAggregatorFactory("_cmp_sumLongSequential", "sumLongSequential"), new IntervalDimFilter(ColumnHolder.TIME_COLUMN_NAME, Collections.singletonList(previous), null)));
        Druids.TimeseriesQueryBuilder timeseriesQueryBuilder = Druids.newTimeseriesQueryBuilder().dataSource("blah").granularity(Granularities.ALL).intervals(intervalSpec).aggregators(queryAggs).descending(false);
        timeseriesQuery = timeseriesQueryBuilder.build();
        timeseriesFactory = new TimeseriesQueryRunnerFactory(new TimeseriesQueryQueryToolChest(), new TimeseriesQueryEngine(), QueryBenchmarkUtil.NOOP_QUERYWATCHER);
    }
}