Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Support max on single-level struct in aggregation context #4434

Merged
merged 7 commits into from
Jan 15, 2022
Merged

Support max on single-level struct in aggregation context #4434

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
56fd233
Support max on single-level struct in aggregation context
Dec 28, 2021
8c5e6de
Refactor
Dec 31, 2021
fbb40d8
Support min and max on single-level
Jan 6, 2022
a7cd407
Merge branch 'branch-22.02' into support-max-single-level-struct
Jan 6, 2022
332c23b
Merge branch 'branch-22.02' into support-max-single-level-struct
Jan 7, 2022
908097c
Update test case after Cudf fixed bug about null
Jan 13, 2022
4102954
Merge min max code
Jan 14, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
12 changes: 6 additions & 6 deletions docs/supported_ops.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15141,7 +15141,7 @@ are limited.
<td><b>NS</b></td>
<td><b>NS</b></td>
<td> </td>
<td><b>NS</b></td>
<td><em>PS<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types BINARY, CALENDAR, ARRAY, STRUCT, UDT</em></td>
<td><b>NS</b></td>
</tr>
<tr>
Expand All @@ -15162,7 +15162,7 @@ are limited.
<td><b>NS</b></td>
<td><b>NS</b></td>
<td> </td>
<td><b>NS</b></td>
<td><em>PS<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types BINARY, CALENDAR, ARRAY, STRUCT, UDT</em></td>
<td><b>NS</b></td>
</tr>
<tr>
Expand All @@ -15184,7 +15184,7 @@ are limited.
<td><b>NS</b></td>
<td><b>NS</b></td>
<td> </td>
<td><b>NS</b></td>
<td><em>PS<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types BINARY, CALENDAR, ARRAY, STRUCT, UDT</em></td>
<td><b>NS</b></td>
</tr>
<tr>
Expand All @@ -15205,7 +15205,7 @@ are limited.
<td><b>NS</b></td>
<td><b>NS</b></td>
<td> </td>
<td><b>NS</b></td>
<td><em>PS<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types BINARY, CALENDAR, ARRAY, STRUCT, UDT</em></td>
<td><b>NS</b></td>
</tr>
<tr>
Expand All @@ -15227,7 +15227,7 @@ are limited.
<td><b>NS</b></td>
<td><b>NS</b></td>
<td> </td>
<td><b>NS</b></td>
<td><em>PS<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types BINARY, CALENDAR, ARRAY, STRUCT, UDT</em></td>
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This says that it works for window operations too, but I see no tests for window operations.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done, removed the window support.

<td><b>NS</b></td>
</tr>
<tr>
Expand All @@ -15248,7 +15248,7 @@ are limited.
<td><b>NS</b></td>
<td><b>NS</b></td>
<td> </td>
<td><b>NS</b></td>
<td><em>PS<br/>UTC is only supported TZ for child TIMESTAMP;<br/>unsupported child types BINARY, CALENDAR, ARRAY, STRUCT, UDT</em></td>
<td><b>NS</b></td>
</tr>
<tr>
Expand Down
19 changes: 19 additions & 0 deletions integration_tests/src/main/python/hash_aggregate_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1695,3 +1695,22 @@ def test_groupby_std_variance_partial_replace_fallback(data_gen,
exist_classes=','.join(exist_clz),
non_exist_classes=','.join(non_exist_clz),
conf=local_conf)

@ignore_order
@pytest.mark.parametrize('data_gen', all_gen + [NullGen()], ids=idfn)
def test_max_single_level_struct(data_gen):
df_gen = [
('a', StructGen([
('aa', data_gen),
('ab', data_gen)])),
('b', IntegerGen())]
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we make it likely that b will have repeated data in it? That way the groupings are likely to have more than one thing in them to compare?

('b', RepeatSeqGen(IntegerGen(), length=20))]

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done, updated.

Copy link
Collaborator Author

@res-life res-life Jan 6, 2022 •
edited

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suported both Min and Max.
Removed the window support.
Updated test case.
Seems CUDF has a bug: rapidsai/cudf#8974 (comment). After confired, I'll file an issue.

assert_gpu_and_cpu_are_equal_sql(
lambda spark : gen_df(spark, df_gen, length=1024),
"hash_agg_table",
'select b, max(a) from hash_agg_table group by b',
_no_nans_float_conf)
assert_gpu_and_cpu_are_equal_sql(
lambda spark : gen_df(spark, df_gen, length=1024),
"hash_agg_table",
'select max(a) from hash_agg_table',
_no_nans_float_conf)
8 changes: 6 additions & 2 deletions sql-plugin/src/main/scala/com/nvidia/spark/rapids/GpuOverrides.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2236,9 +2236,13 @@ object GpuOverrides extends Logging {
expr[Max](
"Max aggregate operator",
ExprChecks.fullAgg(
TypeSig.commonCudfTypes + TypeSig.DECIMAL_128_FULL + TypeSig.NULL, TypeSig.orderable,
// Max supports single level struct, e.g.: max(struct(string, string))
(TypeSig.commonCudfTypes + TypeSig.DECIMAL_128_FULL + TypeSig.NULL + TypeSig.STRUCT)
.nested(TypeSig.commonCudfTypes + TypeSig.DECIMAL_128_FULL + TypeSig.NULL),
TypeSig.orderable,
Seq(ParamCheck("input",
(TypeSig.commonCudfTypes + TypeSig.DECIMAL_128_FULL + TypeSig.NULL)
(TypeSig.commonCudfTypes + TypeSig.DECIMAL_128_FULL + TypeSig.NULL + TypeSig.STRUCT)
.nested(TypeSig.commonCudfTypes + TypeSig.DECIMAL_128_FULL + TypeSig.NULL)
.withPsNote(TypeEnum.DOUBLE, nanAggPsNote)
.withPsNote(TypeEnum.FLOAT, nanAggPsNote),
TypeSig.orderable))
Expand Down