You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@druid.apache.org by "clintropolis (via GitHub)" <gi...@apache.org> on 2024/07/09 05:26:09 UTC
Re: [PR] Improve the fallback strategy when the broker is unable to materialize the subquery's results as frames for estimating the bytes (druid)
clintropolis commented on code in PR #16679:
URL: https://github.com/apache/druid/pull/16679#discussion_r1669740971
##########
extensions-core/datasketches/src/test/java/org/apache/druid/query/aggregation/datasketches/quantiles/sql/DoublesSketchSqlAggregatorTest.java:
##########
@@ -263,6 +269,143 @@ public void testQuantileOnComplexColumn()
);
}
+ @Test
+ public void testSubqueryWithNestedGroupBy()
+ {
+ final List<Object[]> expectedResults = ImmutableList.of(
+ new Object[]{946684800000L, "", 1L, "val1"},
+ new Object[]{946684800000L, "1", 1L, "val1"},
+ new Object[]{946684800000L, "10.1", 1L, "val1"},
+ new Object[]{946684800000L, "2", 1L, "val1"},
+ new Object[]{946684800000L, "abc", 1L, "val1"},
+ new Object[]{946684800000L, "def", 1L, "val1"}
+ );
+
+ testQuery(
+ "SELECT\n"
+ + " MILLIS_TO_TIMESTAMP(946684800000) AS __time,\n"
+ + " alias.\"user\",\n"
+ + " alias.days,\n"
+ + " (CASE WHEN alias.days < quantiles.first_quartile THEN 'val2' \n"
+ + " WHEN alias.days >= quantiles.first_quartile AND alias.days < quantiles.third_quartile THEN 'val3' \n"
+ + " WHEN alias.days >= quantiles.third_quartile THEN 'val1' END) AS val4\n"
+ + "FROM (\n"
+ + " SELECT\n"
+ + " APPROX_QUANTILE_DS(alias.days, 0.25) AS first_quartile,\n"
+ + " APPROX_QUANTILE_DS(alias.days, 0.75) AS third_quartile\n"
+ + " FROM (\n"
+ + " SELECT\n"
+ + " dim1 \"user\",\n"
+ + " COUNT(DISTINCT __time) AS days\n"
+ + " FROM \"foo\"\n"
+ + " GROUP BY 1\n"
+ + " ) AS alias\n"
+ + ") AS quantiles, (\n"
+ + " SELECT\n"
+ + " dim1 \"user\",\n"
+ + " COUNT(DISTINCT __time) AS days\n"
+ + " FROM \"foo\"\n"
+ + " GROUP BY 1\n"
+ + ") AS alias\n",
+ ImmutableMap.<String, Object>builder()
+ .putAll(QUERY_CONTEXT_DEFAULT)
+ .put(QueryContexts.MAX_SUBQUERY_BYTES_KEY, "100000")
+ // Disallows the fallback to row based limiting
+ .put(QueryContexts.MAX_SUBQUERY_ROWS_KEY, "10")
+ .build(),
+ ImmutableList.of(
+ newScanQueryBuilder()
+ .dataSource(
+ JoinDataSource.create(
+ new QueryDataSource(
+ GroupByQuery.builder()
+ .setDataSource(
+ new QueryDataSource(
+ GroupByQuery.builder()
+ .setDataSource("foo")
+ .setQuerySegmentSpec(querySegmentSpec(Intervals.ETERNITY))
+ .setGranularity(Granularities.ALL)
+ .addDimension(new DefaultDimensionSpec(
+ "dim1",
+ "d0",
+ ColumnType.STRING
+ ))
+ .addAggregator(new CardinalityAggregatorFactory(
Review Comment:
super nit: doesn't it feel kind of strange to be mixing non-datasketches approx distinct count with datasketches extensions tests?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@druid.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@druid.apache.org
For additional commands, e-mail: commits-help@druid.apache.org