Examples with LongFirstAggregatorFactory org.apache.druid.query.aggregation.first.LongFirstAggregatorFactory used on opensource projects

Example 1 with LongFirstAggregatorFactory

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

the class CalciteQueryTest method testGroupByAggregatorDefaultValuesNonVectorized.

@Test
public void testGroupByAggregatorDefaultValuesNonVectorized() throws Exception {
    cannotVectorize();
    testQuery("SELECT\n" + " dim2,\n" + " ANY_VALUE(dim1, 1024) FILTER(WHERE dim1 = 'nonexistent'),\n" + " ANY_VALUE(l1) FILTER(WHERE dim1 = 'nonexistent'),\n" + " EARLIEST(dim1, 1024) FILTER(WHERE dim1 = 'nonexistent'),\n" + " EARLIEST(l1) FILTER(WHERE dim1 = 'nonexistent'),\n" + " LATEST(dim1, 1024) FILTER(WHERE dim1 = 'nonexistent'),\n" + " LATEST(l1) FILTER(WHERE dim1 = 'nonexistent'),\n" + " ARRAY_AGG(DISTINCT dim3) FILTER(WHERE dim1 = 'nonexistent'),\n" + " STRING_AGG(DISTINCT dim3, '|') FILTER(WHERE dim1 = 'nonexistent'),\n" + " BIT_AND(l1) FILTER(WHERE dim1 = 'nonexistent'),\n" + " BIT_OR(l1) FILTER(WHERE dim1 = 'nonexistent'),\n" + " BIT_XOR(l1) FILTER(WHERE dim1 = 'nonexistent')\n" + "FROM druid.numfoo WHERE dim2 = 'a' GROUP BY dim2", ImmutableList.of(GroupByQuery.builder().setDataSource(CalciteTests.DATASOURCE3).setInterval(querySegmentSpec(Filtration.eternity())).setDimFilter(selector("dim2", "a", null)).setGranularity(Granularities.ALL).setVirtualColumns(expressionVirtualColumn("v0", "'a'", ColumnType.STRING)).setDimensions(new DefaultDimensionSpec("v0", "_d0", ColumnType.STRING)).setAggregatorSpecs(aggregators(new FilteredAggregatorFactory(new StringAnyAggregatorFactory("a0", "dim1", 1024), selector("dim1", "nonexistent", null)), new FilteredAggregatorFactory(new LongAnyAggregatorFactory("a1", "l1"), selector("dim1", "nonexistent", null)), new FilteredAggregatorFactory(new StringFirstAggregatorFactory("a2", "dim1", null, 1024), selector("dim1", "nonexistent", null)), new FilteredAggregatorFactory(new LongFirstAggregatorFactory("a3", "l1", null), selector("dim1", "nonexistent", null)), new FilteredAggregatorFactory(new StringLastAggregatorFactory("a4", "dim1", null, 1024), selector("dim1", "nonexistent", null)), new FilteredAggregatorFactory(new LongLastAggregatorFactory("a5", "l1", null), selector("dim1", "nonexistent", null)), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a6", ImmutableSet.of("dim3"), "__acc", "ARRAY<STRING>[]", "ARRAY<STRING>[]", true, true, false, "array_set_add(\"__acc\", \"dim3\")", "array_set_add_all(\"__acc\", \"a6\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), selector("dim1", "nonexistent", null)), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a7", ImmutableSet.of("dim3"), "__acc", "[]", "[]", true, false, false, "array_set_add(\"__acc\", \"dim3\")", "array_set_add_all(\"__acc\", \"a7\")", null, "if(array_length(o) == 0, null, array_to_string(o, '|'))", ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), and(not(selector("dim3", null, null)), selector("dim1", "nonexistent", null))), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a8", ImmutableSet.of("l1"), "__acc", "0", "0", NullHandling.sqlCompatible(), false, false, "bitwiseAnd(\"__acc\", \"l1\")", "bitwiseAnd(\"__acc\", \"a8\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), and(not(selector("l1", null, null)), selector("dim1", "nonexistent", null))), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a9", ImmutableSet.of("l1"), "__acc", "0", "0", NullHandling.sqlCompatible(), false, false, "bitwiseOr(\"__acc\", \"l1\")", "bitwiseOr(\"__acc\", \"a9\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), and(not(selector("l1", null, null)), selector("dim1", "nonexistent", null))), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a10", ImmutableSet.of("l1"), "__acc", "0", "0", NullHandling.sqlCompatible(), false, false, "bitwiseXor(\"__acc\", \"l1\")", "bitwiseXor(\"__acc\", \"a10\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), and(not(selector("l1", null, null)), selector("dim1", "nonexistent", null))))).setContext(QUERY_CONTEXT_DEFAULT).build()), ImmutableList.of(useDefault ? new Object[] { "a", "", 0L, "", 0L, "", 0L, null, "", 0L, 0L, 0L } : new Object[] { "a", null, null, null, null, null, null, null, null, null, null, null }));
}

Example 2 with LongFirstAggregatorFactory

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

the class CalciteQueryTest method testOrderByEarliestLong.

@Test
public void testOrderByEarliestLong() throws Exception {
    // Cannot vectorize EARLIEST aggregator.
    skipVectorize();
    List<Object[]> expected;
    if (NullHandling.replaceWithDefault()) {
        expected = ImmutableList.of(new Object[] { "1", 0L }, new Object[] { "2", 0L }, new Object[] { "abc", 0L }, new Object[] { "def", 0L }, new Object[] { "", 7L }, new Object[] { "10.1", 325323L });
    } else {
        expected = ImmutableList.of(new Object[] { "1", null }, new Object[] { "abc", null }, new Object[] { "def", null }, new Object[] { "2", 0L }, new Object[] { "", 7L }, new Object[] { "10.1", 325323L });
    }
    testQuery("SELECT dim1, EARLIEST(l1) FROM druid.numfoo GROUP BY 1 ORDER BY 2 LIMIT 10", ImmutableList.of(new TopNQueryBuilder().dataSource(CalciteTests.DATASOURCE3).intervals(querySegmentSpec(Filtration.eternity())).granularity(Granularities.ALL).dimension(new DefaultDimensionSpec("dim1", "_d0")).aggregators(aggregators(new LongFirstAggregatorFactory("a0", "l1", null))).metric(new InvertedTopNMetricSpec(new NumericTopNMetricSpec("a0"))).threshold(10).context(QUERY_CONTEXT_DEFAULT).build()), expected);
}

Example 3 with LongFirstAggregatorFactory

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

the class CalciteQueryTest method testTimeseriesEmptyResultsAggregatorDefaultValuesNonVectorized.

@Test
public void testTimeseriesEmptyResultsAggregatorDefaultValuesNonVectorized() throws Exception {
    cannotVectorize();
    // timeseries with all granularity have a single group, so should return default results for given aggregators
    testQuery("SELECT\n" + " ANY_VALUE(dim1, 1024),\n" + " ANY_VALUE(l1),\n" + " EARLIEST(dim1, 1024),\n" + " EARLIEST(l1),\n" + " LATEST(dim1, 1024),\n" + " LATEST(l1),\n" + " ARRAY_AGG(DISTINCT dim3),\n" + " STRING_AGG(DISTINCT dim3, '|'),\n" + " BIT_AND(l1),\n" + " BIT_OR(l1),\n" + " BIT_XOR(l1)\n" + "FROM druid.numfoo WHERE dim2 = 0", ImmutableList.of(Druids.newTimeseriesQueryBuilder().dataSource(CalciteTests.DATASOURCE3).intervals(querySegmentSpec(Filtration.eternity())).filters(bound("dim2", "0", "0", false, false, null, StringComparators.NUMERIC)).granularity(Granularities.ALL).aggregators(aggregators(new StringAnyAggregatorFactory("a0", "dim1", 1024), new LongAnyAggregatorFactory("a1", "l1"), new StringFirstAggregatorFactory("a2", "dim1", null, 1024), new LongFirstAggregatorFactory("a3", "l1", null), new StringLastAggregatorFactory("a4", "dim1", null, 1024), new LongLastAggregatorFactory("a5", "l1", null), new ExpressionLambdaAggregatorFactory("a6", ImmutableSet.of("dim3"), "__acc", "ARRAY<STRING>[]", "ARRAY<STRING>[]", true, true, false, "array_set_add(\"__acc\", \"dim3\")", "array_set_add_all(\"__acc\", \"a6\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a7", ImmutableSet.of("dim3"), "__acc", "[]", "[]", true, false, false, "array_set_add(\"__acc\", \"dim3\")", "array_set_add_all(\"__acc\", \"a7\")", null, "if(array_length(o) == 0, null, array_to_string(o, '|'))", ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), not(selector("dim3", null, null))), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a8", ImmutableSet.of("l1"), "__acc", "0", "0", NullHandling.sqlCompatible(), false, false, "bitwiseAnd(\"__acc\", \"l1\")", "bitwiseAnd(\"__acc\", \"a8\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), not(selector("l1", null, null))), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a9", ImmutableSet.of("l1"), "__acc", "0", "0", NullHandling.sqlCompatible(), false, false, "bitwiseOr(\"__acc\", \"l1\")", "bitwiseOr(\"__acc\", \"a9\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), not(selector("l1", null, null))), new FilteredAggregatorFactory(new ExpressionLambdaAggregatorFactory("a10", ImmutableSet.of("l1"), "__acc", "0", "0", NullHandling.sqlCompatible(), false, false, "bitwiseXor(\"__acc\", \"l1\")", "bitwiseXor(\"__acc\", \"a10\")", null, null, ExpressionLambdaAggregatorFactory.DEFAULT_MAX_SIZE_BYTES, TestExprMacroTable.INSTANCE), not(selector("l1", null, null))))).context(QUERY_CONTEXT_DEFAULT).build()), ImmutableList.of(useDefault ? new Object[] { "", 0L, "", 0L, "", 0L, null, "", 0L, 0L, 0L } : new Object[] { null, null, null, null, null, null, null, null, null, null, null }));
}

Example 4 with LongFirstAggregatorFactory

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

the class GroupByQueryRunnerTest method testGroupByWithFirstLast.

@Test
public void testGroupByWithFirstLast() {
    // Cannot vectorize due to "first", "last" aggregators.
    cannotVectorize();
    GroupByQuery query = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setQuerySegmentSpec(QueryRunnerTestHelper.FULL_ON_INTERVAL_SPEC).setDimensions(new DefaultDimensionSpec("market", "market")).setAggregatorSpecs(new LongFirstAggregatorFactory("first", "index", null), new LongLastAggregatorFactory("last", "index", null)).setGranularity(QueryRunnerTestHelper.MONTH_GRAN).build();
    List<ResultRow> expectedResults = Arrays.asList(makeRow(query, "2011-01-01", "market", "spot", "first", 100L, "last", 155L), makeRow(query, "2011-01-01", "market", "total_market", "first", 1000L, "last", 1127L), makeRow(query, "2011-01-01", "market", "upfront", "first", 800L, "last", 943L), makeRow(query, "2011-02-01", "market", "spot", "first", 132L, "last", 114L), makeRow(query, "2011-02-01", "market", "total_market", "first", 1203L, "last", 1292L), makeRow(query, "2011-02-01", "market", "upfront", "first", 1667L, "last", 1101L), makeRow(query, "2011-03-01", "market", "spot", "first", 153L, "last", 125L), makeRow(query, "2011-03-01", "market", "total_market", "first", 1124L, "last", 1366L), makeRow(query, "2011-03-01", "market", "upfront", "first", 1166L, "last", 1063L), makeRow(query, "2011-04-01", "market", "spot", "first", 135L, "last", 120L), makeRow(query, "2011-04-01", "market", "total_market", "first", 1314L, "last", 1029L), makeRow(query, "2011-04-01", "market", "upfront", "first", 1447L, "last", 780L));
    Iterable<ResultRow> results = GroupByQueryRunnerTestHelper.runQuery(factory, runner, query);
    TestHelper.assertExpectedObjects(expectedResults, results, "first-last-aggs");
}

Example 5 with LongFirstAggregatorFactory

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

the class GroupByQueryRunnerTest method testSubqueryWithFirstLast.

@Test
public void testSubqueryWithFirstLast() {
    // Cannot vectorize due to "first", "last" aggregators.
    cannotVectorize();
    GroupByQuery subquery = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setQuerySegmentSpec(QueryRunnerTestHelper.FULL_ON_INTERVAL_SPEC).setDimensions(new DefaultDimensionSpec("market", "market")).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongFirstAggregatorFactory("innerfirst", "index", null), new LongLastAggregatorFactory("innerlast", "index", null)).setGranularity(QueryRunnerTestHelper.DAY_GRAN).overrideContext(ImmutableMap.of("finalize", true)).build();
    GroupByQuery query = makeQueryBuilder().setDataSource(subquery).setQuerySegmentSpec(QueryRunnerTestHelper.FULL_ON_INTERVAL_SPEC).setDimensions(Collections.emptyList()).setAggregatorSpecs(new LongFirstAggregatorFactory("first", "innerfirst", null), new LongLastAggregatorFactory("last", "innerlast", null)).setGranularity(QueryRunnerTestHelper.MONTH_GRAN).build();
    List<ResultRow> expectedResults = Arrays.asList(makeRow(query, "2011-01-01", "first", 100L, "last", 943L), makeRow(query, "2011-02-01", "first", 132L, "last", 1101L), makeRow(query, "2011-03-01", "first", 153L, "last", 1063L), makeRow(query, "2011-04-01", "first", 135L, "last", 780L));
    Iterable<ResultRow> results = GroupByQueryRunnerTestHelper.runQuery(factory, runner, query);
    TestHelper.assertExpectedObjects(expectedResults, results, "subquery-firstlast");
}