Skip to content

fix: Preserve pandas column name attribute #2363

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.

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
wants to merge 20 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from 2 commits
Commits
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
2 changes: 1 addition & 1 deletion narwhals/_dask/dataframe.py
Copy link
Member Author

Choose a reason for hiding this comment

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

Dask select and with_columns have no issues since we use .assign method for both!

Original file line number Diff line number Diff line change
Expand Up @@ -110,7 +110,7 @@ def collect(
from narwhals._pandas_like.dataframe import PandasLikeDataFrame

return PandasLikeDataFrame(
result,
result.rename_axis(columns=self.native.columns.name),
Copy link
Member Author

Choose a reason for hiding this comment

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

This might be one step ahead of dask#11874

implementation=Implementation.PANDAS,
backend_version=parse_version(pd),
version=self._version,
Expand Down
5 changes: 3 additions & 2 deletions narwhals/_pandas_like/dataframe.py
Original file line number Diff line number Diff line change
Expand Up @@ -117,6 +117,7 @@ def __init__(
validate_backend_version(self._implementation, self._backend_version)
if validate_column_names:
check_column_names_are_unique(native_dataframe.columns)
self._native_columns_name = native_dataframe.columns.name

@classmethod
def from_arrow(cls, data: IntoArrowTable, /, *, context: _FullContext) -> Self:
Expand Down Expand Up @@ -248,7 +249,7 @@ def _with_version(self: Self, version: Version) -> Self:

def _with_native(self: Self, df: Any, *, validate_column_names: bool = True) -> Self:
return self.__class__(
df,
df.rename_axis(columns=self._native_columns_name),
implementation=self._implementation,
backend_version=self._backend_version,
version=self._version,
Expand Down Expand Up @@ -630,7 +631,7 @@ def collect(
import pandas as pd # ignore-banned-import

return PandasLikeDataFrame(
self.to_pandas(),
self.to_pandas().rename_axis(columns=self.native.columns.name),
implementation=Implementation.PANDAS,
backend_version=parse_version(pd),
version=self._version,
Expand Down
46 changes: 18 additions & 28 deletions narwhals/_pandas_like/namespace.py
Original file line number Diff line number Diff line change
Expand Up @@ -231,43 +231,33 @@ def concat(
) -> PandasLikeDataFrame:
dfs: list[Any] = [item._native_frame for item in items]
if how == "horizontal":
return PandasLikeDataFrame(
horizontal_concat(
dfs,
implementation=self._implementation,
backend_version=self._backend_version,
),
native_dataframe = horizontal_concat(
dfs,
implementation=self._implementation,
backend_version=self._backend_version,
version=self._version,
validate_column_names=True,
)
if how == "vertical":
return PandasLikeDataFrame(
vertical_concat(
dfs,
implementation=self._implementation,
backend_version=self._backend_version,
),
elif how == "vertical":
native_dataframe = vertical_concat(
dfs,
implementation=self._implementation,
backend_version=self._backend_version,
version=self._version,
validate_column_names=True,
)

if how == "diagonal":
return PandasLikeDataFrame(
diagonal_concat(
dfs,
implementation=self._implementation,
backend_version=self._backend_version,
),
elif how == "diagonal":
native_dataframe = diagonal_concat(
dfs,
implementation=self._implementation,
backend_version=self._backend_version,
version=self._version,
validate_column_names=True,
)
raise NotImplementedError
else:
raise NotImplementedError

return PandasLikeDataFrame(
native_dataframe,
implementation=self._implementation,
backend_version=self._backend_version,
version=self._version,
validate_column_names=True,
)

def when(self: Self, predicate: PandasLikeExpr) -> PandasWhen:
return PandasWhen.from_expr(predicate, context=self)
Expand Down
2 changes: 1 addition & 1 deletion narwhals/translate.py
Original file line number Diff line number Diff line change
Expand Up @@ -488,8 +488,8 @@ def _from_native_impl( # noqa: PLR0915
return DataFrame(
PandasLikeDataFrame(
native_object,
backend_version=parse_version(pd),
implementation=Implementation.PANDAS,
backend_version=parse_version(pd),
Copy link
Member Author

Choose a reason for hiding this comment

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

Mental sanity for symmetry with other pandas-like and order of the spec πŸ˜…

version=version,
validate_column_names=True,
),
Expand Down
28 changes: 28 additions & 0 deletions tests/preserve_pandas_like_columns_name_attr_test.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,28 @@
from __future__ import annotations

from typing import TYPE_CHECKING

import pytest

import narwhals.stable.v1 as nw

if TYPE_CHECKING:
from tests.utils import Constructor


def test_ops_preserve_column_index_name(constructor: Constructor) -> None:
if not any(x in str(constructor) for x in ("pandas", "modin", "cudf", "dask")):
pytest.skip(
reason="Dataframe columns is a list and do not have a `name` like a pandas Index does"
)

data = {"a": [1, 3, 2], "b": [4, 4, 6], "z": [7.0, 8.0, 9.0]}
df_native = constructor(data)
df_native.columns.name = "foo" # type: ignore[union-attr]

df = nw.from_native(df_native)

result = df.with_columns(b=nw.col("a") + 1, c=nw.col("a") * 2).select("c", "b")
Copy link
Member Author

Choose a reason for hiding this comment

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

I wanted to concatenate two methods here, mostly for the sake of it.
Might be worth reparametrizing it


assert result.to_native().columns.name == "foo" # type: ignore[union-attr]
assert result.lazy().collect(backend="pandas").to_native().columns.name == "foo"
Loading