8000 fix: Fix issues with chunked arrow data by TrevorBergeron · Pull Request #1700 · googleapis/python-bigquery-dataframes · GitHub
[go: up one dir, main page]
Skip to content

fix: Fix issues with chunked arrow data #1700

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

Merged
TrevorBergeron merged 2 commits into from
May 8, 2025
Merged

fix: Fix issues with chunked arrow data #1700

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b5da10e
fix: Fix issues with chunked arrow data
TrevorBergeron May 6, 2025
5a33506
Merge branch 'main' into fix_chunked_struct
TrevorBergeron May 7, 2025
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
24 changes: 20 additions & 4 deletions bigframes/core/local_data.py
View file
Open in desktop
8000
Original file line number Diff line number Diff line change
Expand Up @@ -86,7 +86,7 @@ def from_pyarrow(self, table: pa.Table) -> ManagedArrowTable:
columns: list[pa.ChunkedArray] = []
fields: list[schemata.SchemaItem] = []
for name, arr in zip(table.column_names, table.columns):
new_arr, bf_type = _adapt_arrow_array(arr)
new_arr, bf_type = _adapt_chunked_array(arr)
columns.append(new_arr)
fields.append(schemata.SchemaItem(name, bf_type))

Expand Down Expand Up @@ -279,10 +279,26 @@ def _adapt_pandas_series(
raise e


def _adapt_arrow_array(
array: Union[pa.ChunkedArray, pa.Array]
) -> tuple[Union[pa.ChunkedArray, pa.Array], bigframes.dtypes.Dtype]:
def _adapt_chunked_array(
chunked_array: pa.ChunkedArray,
) -> tuple[pa.ChunkedArray, bigframes.dtypes.Dtype]:
if len(chunked_array.chunks) == 0:
return _adapt_arrow_array(chunked_array.combine_chunks())
dtype = None
arrays = []
for chunk in chunked_array.chunks:
array, arr_dtype = _adapt_arrow_array(chunk)
arrays.append(array)
dtype = dtype or arr_dtype
assert dtype is not None
return pa.chunked_array(arrays), dtype


def _adapt_arrow_array(array: pa.Array) -> tuple[pa.Array, bigframes.dtypes.Dtype]:
"""Normalize the array to managed storage types. Preverse shapes, only transforms values."""
if array.offset != 0: # Offset arrays don't have all operations implemented
return _adapt_arrow_array(pa.concat_arrays([array]))

if pa.types.is_struct(array.type):
assert isinstance(array, pa.StructArray)
assert isinstance(array.type, pa.StructType)
Expand Down
20 changes: 20 additions & 0 deletions tests/unit/test_local_data.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,3 +44,23 @@ def test_local_data_well_formed_round_trip():
local_entry = local_data.Manage 6C52 dArrowTable.from_pandas(pd_data)
result = pd.DataFrame(local_entry.itertuples(), columns=pd_data.columns)
pandas.testing.assert_frame_equal(pd_data_normalized, result, check_dtype=False)


def test_local_data_well_formed_round_trip_chunked():
pa_table = pa.Table.from_pandas(pd_data, preserve_index=False)
as_rechunked_pyarrow = pa.Table.from_batches(pa_table.to_batches(max_chunksize=2))
local_entry = local_data.ManagedArrowTable.from_pyarrow(as_rechunked_pyarrow)
result = pd.DataFrame(local_entry.itertuples(), columns=pd_data.columns)
pandas.testing.assert_frame_equal(pd_data_normalized, result, check_dtype=False)


def test_local_data_well_formed_round_trip_sliced():
pa_table = pa.Table.from_pandas(pd_data, preserve_index=False)
as_rechunked_pyarrow = pa.Table.from_batches(pa_table.slice(2, 4).to_batches())
local_entry = local_data.ManagedArrowTable.from_pyarrow(as_rechunked_pyarrow)
result = pd.DataFrame(local_entry.itertuples(), columns=pd_data.columns)
pandas.testing.assert_frame_equal(
pd_data_normalized[2:4].reset_index(drop=True),
result.reset_index(drop=True),
check_dtype=False,
)
0