Skip to content

fix: handle aggregate operations on empty selections #2510

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

Open
shuoweil wants to merge 4 commits into
base: main
Choose a base branch
from
+28 −3
Open

fix: handle aggregate operations on empty selections #2510

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
ee4e279
fix: inject SELECT 1 fallback for empty agg projections
shuoweil Mar 11, 2026
cbf27fe
test: add verification for empty selection aggregate projections
shuoweil Mar 11, 2026
05805aa
Merge branch 'main' into shuowei-fix-compiler-syntax-guards
shuoweil Mar 12, 2026
e4e5347
Merge branch 'main' into shuowei-fix-compiler-syntax-guards
shuoweil Mar 12, 2026
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
8 changes: 8 additions & 0 deletions tests/system/small/test_dataframe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6283,3 +6283,11 @@ def test_agg_with_dict_containing_non_existing_col_raise_key_error(scalars_dfs):

with pytest.raises(KeyError):
bf_df.agg(agg_funcs)


def test_dataframe_count_empty_selection_succeeds(session):
# Tests that aggregate ops on empty selections don't trigger invalid empty SELECT syntax
df = session.read_gbq("SELECT 1 AS int_col")
empty_df = df[[]]
Copy link
Contributor

@chelsea-lin chelsea-lin Mar 12, 2026

Choose a reason for hiding this comment

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

Alternatively: empty_df = bpd.DataFrame()

count_series = empty_df.count().to_pandas()
assert len(count_series) == 0
14 changes: 11 additions & 3 deletions third_party/bigframes_vendored/ibis/backends/sql/compilers/base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1394,9 +1394,17 @@ def _generate_groups(groups):
return map(sge.convert, range(1, len(groups) + 1))

def visit_Aggregate(self, op, *, parent, groups, metrics):
sel = sg.select(
*self._cleanup_names(groups), *self._cleanup_names(metrics), copy=False
).from_(parent, copy=False)
exprs = []
if groups:
exprs.extend(self._cleanup_names(groups))
if metrics:
exprs.extend(self._cleanup_names(metrics))

if not exprs:
# Empty aggregated projections are invalid in BigQuery
exprs = [sge.Literal.number(1)]

sel = sg.select(*exprs, copy=False).from_(parent, copy=False)

if groups:
sel = sel.group_by(*self._generate_groups(groups.values()), copy=False)
Expand Down
9 changes: 9 additions & 0 deletions third_party/bigframes_vendored/ibis/backends/sql/compilers/bigquery/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -540,6 +540,15 @@ def visit_TimestampFromUNIX(self, op, *, arg, unit):

def visit_Cast(self, op, *, arg, to):
from_ = op.arg.dtype
if to.is_null():
return sge.Null()
if arg is NULL or (
isinstance(arg, sge.Cast)
and getattr(arg, "to", None) is not None
and str(arg.to).upper() == "NULL"
):
if to.is_struct() or to.is_array():
return sge.Cast(this=NULL, to=self.type_mapper.from_ibis(to))
if from_.is_timestamp() and to.is_integer():
return self.f.unix_micros(arg)
elif from_.is_integer() and to.is_timestamp():
Expand Down
Loading