Examples with MapBasedRow org.apache.druid.data.input.MapBasedRow used on opensource projects

Example 46 with MapBasedRow

use of org.apache.druid.data.input.MapBasedRow in project druid by druid-io.

the class SketchAggregationTest method testThetaCardinalityOnSimpleColumn.

@Test
public void testThetaCardinalityOnSimpleColumn() throws Exception {
    final GroupByQuery groupByQuery = readQueryFromClasspath("simple_test_data_group_by_query.json", helper.getObjectMapper(), vectorize);
    final Sequence<ResultRow> seq = helper.createIndexAndRunQueryOnSegment(new File(SketchAggregationTest.class.getClassLoader().getResource("simple_test_data.tsv").getFile()), readFileFromClasspathAsString("simple_test_data_record_parser2.json"), "[" + "  {" + "    \"type\": \"count\"," + "    \"name\": \"count\"" + "  }" + "]", 0, Granularities.NONE, 1000, groupByQuery);
    List<ResultRow> results = seq.toList();
    Assert.assertEquals(5, results.size());
    Assert.assertEquals(ImmutableList.of(new MapBasedRow(DateTimes.of("2014-10-19T00:00:00.000Z"), ImmutableMap.<String, Object>builder().put("product", "product_3").put("sketch_count", 38.0).put("sketchEstimatePostAgg", 38.0).put("sketchUnionPostAggEstimate", 38.0).put("sketchIntersectionPostAggEstimate", 38.0).put("sketchAnotBPostAggEstimate", 0.0).put("non_existing_col_validation", 0.0).build()), new MapBasedRow(DateTimes.of("2014-10-19T00:00:00.000Z"), ImmutableMap.<String, Object>builder().put("product", "product_1").put("sketch_count", 42.0).put("sketchEstimatePostAgg", 42.0).put("sketchUnionPostAggEstimate", 42.0).put("sketchIntersectionPostAggEstimate", 42.0).put("sketchAnotBPostAggEstimate", 0.0).put("non_existing_col_validation", 0.0).build()), new MapBasedRow(DateTimes.of("2014-10-19T00:00:00.000Z"), ImmutableMap.<String, Object>builder().put("product", "product_2").put("sketch_count", 42.0).put("sketchEstimatePostAgg", 42.0).put("sketchUnionPostAggEstimate", 42.0).put("sketchIntersectionPostAggEstimate", 42.0).put("sketchAnotBPostAggEstimate", 0.0).put("non_existing_col_validation", 0.0).build()), new MapBasedRow(DateTimes.of("2014-10-19T00:00:00.000Z"), ImmutableMap.<String, Object>builder().put("product", "product_4").put("sketch_count", 42.0).put("sketchEstimatePostAgg", 42.0).put("sketchUnionPostAggEstimate", 42.0).put("sketchIntersectionPostAggEstimate", 42.0).put("sketchAnotBPostAggEstimate", 0.0).put("non_existing_col_validation", 0.0).build()), new MapBasedRow(DateTimes.of("2014-10-19T00:00:00.000Z"), ImmutableMap.<String, Object>builder().put("product", "product_5").put("sketch_count", 42.0).put("sketchEstimatePostAgg", 42.0).put("sketchUnionPostAggEstimate", 42.0).put("sketchIntersectionPostAggEstimate", 42.0).put("sketchAnotBPostAggEstimate", 0.0).put("non_existing_col_validation", 0.0).build())).stream().map(row -> ResultRow.fromLegacyRow(row, groupByQuery)).collect(Collectors.toList()), results);
}

Example 47 with MapBasedRow

use of org.apache.druid.data.input.MapBasedRow in project druid by druid-io.

the class BufferHashGrouperUsingSketchMergeAggregatorFactoryTest method testGrowingBufferGrouper.

@Test
public void testGrowingBufferGrouper() {
    final TestColumnSelectorFactory columnSelectorFactory = GrouperTestUtil.newColumnSelectorFactory();
    final Grouper<Integer> grouper = makeGrouper(columnSelectorFactory, 100000, 2);
    try {
        final int expectedMaxSize = 5;
        SketchHolder sketchHolder = SketchHolder.of(Sketches.updateSketchBuilder().setNominalEntries(16).build());
        UpdateSketch updateSketch = (UpdateSketch) sketchHolder.getSketch();
        updateSketch.update(1);
        columnSelectorFactory.setRow(new MapBasedRow(0, ImmutableMap.of("sketch", sketchHolder)));
        for (int i = 0; i < expectedMaxSize; i++) {
            Assert.assertTrue(String.valueOf(i), grouper.aggregate(i).isOk());
        }
        updateSketch.update(3);
        columnSelectorFactory.setRow(new MapBasedRow(0, ImmutableMap.of("sketch", sketchHolder)));
        for (int i = 0; i < expectedMaxSize; i++) {
            Assert.assertTrue(String.valueOf(i), grouper.aggregate(i).isOk());
        }
        Object[] holders = Lists.newArrayList(grouper.iterator(true)).get(0).getValues();
        Assert.assertEquals(2.0d, ((SketchHolder) holders[0]).getEstimate(), 0);
    } finally {
        grouper.close();
    }
}

Example 48 with MapBasedRow

use of org.apache.druid.data.input.MapBasedRow in project druid by druid-io.

the class OldApiSketchAggregationTest method testSimpleDataIngestAndQuery.

@Test
public void testSimpleDataIngestAndQuery() throws Exception {
    final String groupByQueryString = readFileFromClasspathAsString("oldapi/old_simple_test_data_group_by_query.json");
    final GroupByQuery groupByQuery = (GroupByQuery) helper.getObjectMapper().readValue(groupByQueryString, Query.class);
    final Sequence seq = helper.createIndexAndRunQueryOnSegment(new File(this.getClass().getClassLoader().getResource("simple_test_data.tsv").getFile()), readFileFromClasspathAsString("simple_test_data_record_parser.json"), readFileFromClasspathAsString("oldapi/old_simple_test_data_aggregators.json"), 0, Granularities.NONE, 1000, groupByQueryString);
    List results = seq.toList();
    Assert.assertEquals(1, results.size());
    Assert.assertEquals(ResultRow.fromLegacyRow(new MapBasedRow(DateTimes.of("2014-10-19T00:00:00.000Z"), ImmutableMap.<String, Object>builder().put("sketch_count", 50.0).put("sketchEstimatePostAgg", 50.0).put("sketchUnionPostAggEstimate", 50.0).put("sketchIntersectionPostAggEstimate", 50.0).put("sketchAnotBPostAggEstimate", 0.0).put("non_existing_col_validation", 0.0).build()), groupByQuery), results.get(0));
}

Example 49 with MapBasedRow

use of org.apache.druid.data.input.MapBasedRow in project druid by druid-io.

the class BloomFilterGroupByQueryTest method testNestedQuery.

@Test
public void testNestedQuery() throws Exception {
    if (!isV2) {
        return;
    }
    String query = "{" + "\"queryType\": \"groupBy\"," + "\"dataSource\": {" + "\"type\": \"query\"," + "\"query\": {" + "\"queryType\":\"groupBy\"," + "\"dataSource\": \"test_datasource\"," + "\"intervals\": [ \"1970/2050\" ]," + "\"granularity\":\"ALL\"," + "\"dimensions\":[]," + "\"aggregations\": [{ \"type\":\"longSum\", \"name\":\"innerSum\", \"fieldName\":\"count\"}]" + "}" + "}," + "\"granularity\": \"ALL\"," + "\"dimensions\": []," + "\"aggregations\": [" + "  { \"type\": \"bloom\", \"name\": \"bloom\", \"field\": \"innerSum\" }" + "]," + "\"intervals\": [ \"1970/2050\" ]" + "}";
    MapBasedRow row = ingestAndQuery(query);
    BloomKFilter filter = BloomKFilter.deserialize((ByteBuffer) row.getRaw("bloom"));
    Assert.assertTrue(filter.testLong(13L));
    Assert.assertFalse(filter.testLong(5L));
}

Example 50 with MapBasedRow

use of org.apache.druid.data.input.MapBasedRow in project druid by druid-io.

the class BloomFilterGroupByQueryTest method testQueryFakeDimension.

@Test
public void testQueryFakeDimension() throws Exception {
    String query = "{" + "\"queryType\": \"groupBy\"," + "\"dataSource\": \"test_datasource\"," + "\"granularity\": \"ALL\"," + "\"dimensions\": []," + "\"filter\":{ \"type\":\"selector\", \"dimension\":\"market\", \"value\":\"upfront\"}," + "\"aggregations\": [" + "  { \"type\": \"bloom\", \"name\": \"blooming_quality\", \"field\": \"nope\" }" + "]," + "\"intervals\": [ \"1970/2050\" ]" + "}";
    MapBasedRow row = ingestAndQuery(query);
    // a nil column results in a totally empty bloom filter
    BloomKFilter filter = new BloomKFilter(1500);
    Object val = row.getRaw("blooming_quality");
    String serialized = BloomFilterAggregatorTest.filterToString(BloomKFilter.deserialize((ByteBuffer) val));
    String empty = BloomFilterAggregatorTest.filterToString(filter);
    Assert.assertEquals(empty, serialized);
}