Skip to content

CoW: Add warnings for interpolate #56289

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

Merged
merged 12 commits into from
Dec 8, 2023
Merged
Show file tree
Hide file tree
Changes from all 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
7 changes: 6 additions & 1 deletion pandas/core/internals/base.py
Original file line number Diff line number Diff line change
Expand Up @@ -284,7 +284,11 @@ def replace_list(

def interpolate(self, inplace: bool, **kwargs) -> Self:
return self.apply_with_block(
"interpolate", inplace=inplace, **kwargs, using_cow=using_copy_on_write()
"interpolate",
inplace=inplace,
**kwargs,
using_cow=using_copy_on_write(),
already_warned=_AlreadyWarned(),
)

def pad_or_backfill(self, inplace: bool, **kwargs) -> Self:
Expand All @@ -293,6 +297,7 @@ def pad_or_backfill(self, inplace: bool, **kwargs) -> Self:
inplace=inplace,
**kwargs,
using_cow=using_copy_on_write(),
already_warned=_AlreadyWarned(),
)

def shift(self, periods: int, fill_value) -> Self:
Expand Down
31 changes: 30 additions & 1 deletion pandas/core/internals/blocks.py
Original file line number Diff line number Diff line change
Expand Up @@ -1657,6 +1657,7 @@ def pad_or_backfill(
limit_area: Literal["inside", "outside"] | None = None,
downcast: Literal["infer"] | None = None,
using_cow: bool = False,
already_warned=None,
) -> list[Block]:
if not self._can_hold_na:
# If there are no NAs, then interpolate is a no-op
Expand All @@ -1677,6 +1678,19 @@ def pad_or_backfill(
limit_area=limit_area,
copy=copy,
)
if (
not copy
and warn_copy_on_write()
and already_warned is not None
and not already_warned.warned_already
):
if self.refs.has_reference():
warnings.warn(
COW_WARNING_GENERAL_MSG,
FutureWarning,
stacklevel=find_stack_level(),
)
already_warned.warned_already = True
if axis == 1:
new_values = new_values.T

Expand All @@ -1697,6 +1711,7 @@ def interpolate(
limit_area: Literal["inside", "outside"] | None = None,
downcast: Literal["infer"] | None = None,
using_cow: bool = False,
already_warned=None,
**kwargs,
) -> list[Block]:
inplace = validate_bool_kwarg(inplace, "inplace")
Expand Down Expand Up @@ -1735,6 +1750,20 @@ def interpolate(
)
data = extract_array(new_values, extract_numpy=True)

if (
not copy
and warn_copy_on_write()
and already_warned is not None
and not already_warned.warned_already
):
if self.refs.has_reference():
warnings.warn(
COW_WARNING_GENERAL_MSG,
FutureWarning,
stacklevel=find_stack_level(),
)
already_warned.warned_already = True

nb = self.make_block_same_class(data, refs=refs)
return nb._maybe_downcast([nb], downcast, using_cow, caller="interpolate")

Expand Down Expand Up @@ -2178,9 +2207,9 @@ def pad_or_backfill(
limit_area: Literal["inside", "outside"] | None = None,
downcast: Literal["infer"] | None = None,
using_cow: bool = False,
already_warned=None,
) -> list[Block]:
values = self.values
copy, refs = self._get_refs_and_copy(using_cow, inplace)

if values.ndim == 2 and axis == 1:
# NDArrayBackedExtensionArray.fillna assumes axis=0
Expand Down
13 changes: 9 additions & 4 deletions pandas/tests/copy_view/test_chained_assignment_deprecation.py
Original file line number Diff line number Diff line change
Expand Up @@ -42,11 +42,16 @@ def test_methods_iloc_warn(using_copy_on_write):
("ffill", ()),
],
)
def test_methods_iloc_getitem_item_cache(func, args, using_copy_on_write):
df = DataFrame({"a": [1, 2, 3], "b": 1})
def test_methods_iloc_getitem_item_cache(
func, args, using_copy_on_write, warn_copy_on_write
):
df = DataFrame({"a": [1.5, 2, 3], "b": 1.5})
ser = df.iloc[:, 0]
# TODO(CoW-warn) should warn about updating a view
getattr(ser, func)(*args, inplace=True)
# TODO(CoW-warn) should warn about updating a view for all methods
with tm.assert_cow_warning(
warn_copy_on_write and func not in ("replace", "fillna")
):
getattr(ser, func)(*args, inplace=True)

# parent that holds item_cache is dead, so don't increase ref count
ser = df.copy()["a"]
Expand Down
30 changes: 28 additions & 2 deletions pandas/tests/copy_view/test_interp_fillna.py
Original file line number Diff line number Diff line change
Expand Up @@ -91,12 +91,13 @@ def test_interpolate_inplace_no_reference_no_copy(using_copy_on_write, vals):
@pytest.mark.parametrize(
"vals", [[1, np.nan, 2], [Timestamp("2019-12-31"), NaT, Timestamp("2020-12-31")]]
)
def test_interpolate_inplace_with_refs(using_copy_on_write, vals):
def test_interpolate_inplace_with_refs(using_copy_on_write, vals, warn_copy_on_write):
df = DataFrame({"a": [1, np.nan, 2]})
df_orig = df.copy()
arr = get_array(df, "a")
view = df[:]
df.interpolate(method="linear", inplace=True)
with tm.assert_cow_warning(warn_copy_on_write):
df.interpolate(method="linear", inplace=True)

if using_copy_on_write:
# Check that copy was triggered in interpolate and that we don't
Expand All @@ -109,6 +110,31 @@ def test_interpolate_inplace_with_refs(using_copy_on_write, vals):
assert np.shares_memory(arr, get_array(df, "a"))


@pytest.mark.parametrize("func", ["ffill", "bfill"])
@pytest.mark.parametrize("dtype", ["float64", "Float64"])
def test_interp_fill_functions_inplace(
using_copy_on_write, func, warn_copy_on_write, dtype
):
# Check that these takes the same code paths as interpolate
df = DataFrame({"a": [1, np.nan, 2]}, dtype=dtype)
df_orig = df.copy()
arr = get_array(df, "a")
view = df[:]

with tm.assert_cow_warning(warn_copy_on_write and dtype == "float64"):
getattr(df, func)(inplace=True)

if using_copy_on_write:
# Check that copy was triggered in interpolate and that we don't
# have any references left
assert not np.shares_memory(arr, get_array(df, "a"))
tm.assert_frame_equal(df_orig, view)
assert df._mgr._has_no_reference(0)
assert view._mgr._has_no_reference(0)
else:
assert np.shares_memory(arr, get_array(df, "a")) is (dtype == "float64")


def test_interpolate_cleaned_fill_method(using_copy_on_write):
# Check that "method is set to None" case works correctly
df = DataFrame({"a": ["a", np.nan, "c"], "b": 1})
Expand Down
3 changes: 2 additions & 1 deletion pandas/tests/copy_view/test_methods.py
Original file line number Diff line number Diff line change
Expand Up @@ -1607,7 +1607,8 @@ def test_interpolate_creates_copy(using_copy_on_write, warn_copy_on_write):
view = df[:]
expected = df.copy()

df.ffill(inplace=True)
with tm.assert_cow_warning(warn_copy_on_write):
df.ffill(inplace=True)
with tm.assert_cow_warning(warn_copy_on_write):
df.iloc[0, 0] = 100.5

Expand Down