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

fix: unordered mode too many labels issue. #1148

Merged
merged 4 commits into from
Nov 13, 2024
Merged

fix: unordered mode too many labels issue. #1148

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
e712c65
fix: unordered mode too many labels issue.
Genesis929 Nov 12, 2024
882f705
update test
Genesis929 Nov 12, 2024
6f39d68
update test
Genesis929 Nov 12, 2024
ebc6d05
Update test
Genesis929 Nov 13, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions bigframes/session/executor.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -379,11 +379,12 @@ def _run_execute_query(
job_config.maximum_bytes_billed = (
bigframes.options.compute.maximum_bytes_billed
)
# Note: add_labels is global scope which may have unexpected effects
bq_io.add_labels(job_config, api_name=api_name)

if not self.strictly_ordered:
job_config.labels["bigframes-mode"] = "unordered"

# Note: add_labels is global scope which may have unexpected effects
bq_io.add_labels(job_config, api_name=api_name)
Copy link
Contributor

Choose a reason for hiding this comment

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

Highly recommend adding a test using the customer's use case

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Added

try:
query_job = self.bqclient.query(sql, job_config=job_config)
return (
Expand Down
27 changes: 11 additions & 16 deletions tests/system/small/test_series.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2766,22 +2766,17 @@ def test_series_case_when(scalars_dfs_maybe_ordered):
bf_series = scalars_df["int64_col"]
pd_series = scalars_pandas_df["int64_col"]

# TODO(tswast): pandas case_when appears to assume True when a value is
Copy link
Contributor

@shobsi shobsi Nov 12, 2024 •
edited
Loading

Choose a reason for hiding this comment

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

I would keep this comment, or rather add a dedicated test that clearly conveys the intent - test_series_case_when_with_large_number_of_operations, or even test_series_large_number_of_operations, with comment that describes the labels limit

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Updated to have all cases in the original test. Keep only one test can be overall faster I think.

# null. I suspect this should be considered a bug in pandas.
bf_result = bf_series.case_when(
[
((bf_series > 100).fillna(True), bf_series - 1),
((bf_series > 0).fillna(True), pd.NA),
((bf_series < -100).fillna(True), -1000),
]
).to_pandas()
pd_result = pd_series.case_when(
[
(pd_series > 100, pd_series - 1),
(pd_series > 0, pd.NA),
(pd_series < -100, -1000),
]
)
bf_conditions = [
((bf_series > (-100 + i * 5)).fillna(True), i) for i in range(149, 0, -1)
] + [((bf_series <= -100).fillna(True), 0)]

pd_conditions = [((pd_series > (-100 + i * 5)), i) for i in range(149, 0, -1)] + [
(pd_series <= -100, 0)
]

bf_result = bf_series.case_when(bf_conditions).to_pandas()
pd_result = pd_series.case_when(pd_conditions)

pd.testing.assert_series_equal(
bf_result,
pd_result.astype(pd.Int64Dtype()),
Expand Down