Skip to content

Fix xr.where(..., keep_attrs=True) bug #6461

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
max-sixty merged 4 commits into from
Apr 12, 2022
Merged

Fix xr.where(..., keep_attrs=True) bug #6461

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
5868aed
check for attrs on x
slevang Apr 9, 2022
b3b9f57
ensure we only get attrs of x
slevang Apr 10, 2022
7059d8e
Merge branch 'main' into fix_where_attrs
slevang Apr 10, 2022
38ef3b7
add whats-new
slevang Apr 10, 2022
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: 2 additions & 0 deletions doc/whats-new.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,6 +73,8 @@ Bug fixes
- In the API for backends, support dimensions that express their preferred chunk sizes
as a tuple of integers. (:issue:`6333`, :pull:`6334`)
By `Stan West <https://github.com/stanwest>`_.
- Fix bug in :py:func:`where` when passing non-xarray objects with ``keep_attrs=True``. (:issue:`6444`, :pull:`6461`)
By `Sam Levang <https://github.com/slevang>`_.

Documentation
~~~~~~~~~~~~~
Expand Down
3 changes: 1 addition & 2 deletions xarray/core/computation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1825,11 +1825,10 @@ def where(cond, x, y, keep_attrs=None):
"""
if keep_attrs is None:
keep_attrs = _get_keep_attrs(default=False)

if keep_attrs is True:
# keep the attributes of x, the second parameter, by default to
# be consistent with the `where` method of `DataArray` and `Dataset`
keep_attrs = lambda attrs, context: attrs[1]
keep_attrs = lambda attrs, context: getattr(x, "attrs", {})
Copy link
Contributor

@dcherian dcherian Oct 25, 2022

Choose a reason for hiding this comment

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

What is attrs here for a scalar? The issue is that you're using x.attrs which could be a dataset while presumably attrs has Variable attributes when apply_ufunc is iterating through variables in a dataset.. We want attrs[1] or {} I think with some handling for errors.

PS: Sorry for the not very throrough review earlier

Copy link
Collaborator

@keewis keewis Oct 25, 2022

Choose a reason for hiding this comment

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

I think something like this should work (with getitem defined in xarray.core.utils):

_default = object()
def getitem(sequence, item, default=_default):
    try:
        return sequence[item]
    except IndexError:
        if default is _default:
            raise
        return default

keep_attrs = lambda attrs, context: getitem(attrs, 1, {})


# alignment for three arguments is complicated, so don't support it yet
return apply_ufunc(
Expand Down
4 changes: 4 additions & 0 deletions xarray/tests/test_computation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1928,6 +1928,10 @@ def test_where_attrs() -> None:
expected = xr.DataArray([1, 0], dims="x", attrs={"attr": "x"})
assert_identical(expected, actual)

# ensure keep_attrs can handle scalar values
actual = xr.where(cond, 1, 0, keep_attrs=True)
assert actual.attrs == {}


@pytest.mark.parametrize("use_dask", [True, False])
@pytest.mark.parametrize("use_datetime", [True, False])
Expand Down