Skip to content

REF: implement Index._find_common_type_compat #39408

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
merged 1 commit into from
Jan 26, 2021
Merged

REF: implement Index._find_common_type_compat #39408

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
27 changes: 17 additions & 10 deletions pandas/core/indexes/base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3312,7 +3312,8 @@ def _get_indexer(
tolerance = self._convert_tolerance(tolerance, target)

if not is_dtype_equal(self.dtype, target.dtype):
dtype = find_common_type([self.dtype, target.dtype])
dtype = self._find_common_type_compat(target)

this = self.astype(dtype, copy=False)
target = target.astype(dtype, copy=False)
return this.get_indexer(
Expand Down Expand Up @@ -5075,15 +5076,7 @@ def get_indexer_non_unique(self, target):
if not is_dtype_equal(self.dtype, target.dtype):
# TODO: if object, could use infer_dtype to preempt costly
# conversion if still non-comparable?
dtype = find_common_type([self.dtype, target.dtype])
if (
dtype.kind in ["i", "u"]
and is_categorical_dtype(target.dtype)
and target.hasnans
):
# FIXME: find_common_type incorrect with Categorical GH#38240
# FIXME: some cases where float64 cast can be lossy?
dtype = np.dtype(np.float64)
dtype = self._find_common_type_compat(target)

this = self.astype(dtype, copy=False)
that = target.astype(dtype, copy=False)
Expand Down Expand Up @@ -5196,6 +5189,20 @@ def _maybe_promote(self, other: Index):

return self, other

def _find_common_type_compat(self, target: Index) -> DtypeObj:
"""
Implementation of find_common_type that adjusts for Index-specific
special cases.
"""
dtype = find_common_type([self.dtype, target.dtype])
if dtype.kind in ["i", "u"]:
# TODO: what about reversed with self being categorical?
if is_categorical_dtype(target.dtype) and target.hasnans:
# FIXME: find_common_type incorrect with Categorical GH#38240
# FIXME: some cases where float64 cast can be lossy?
dtype = np.dtype(np.float64)
return dtype

@final
def _should_compare(self, other: Index) -> bool:
"""
Expand Down
22 changes: 15 additions & 7 deletions pandas/tests/arrays/categorical/test_indexing.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -212,17 +212,25 @@ def test_categories_assignments_wrong_length_raises(self, new_categories):
# Combinations of missing/unique
@pytest.mark.parametrize("key_values", [[1, 2], [1, 5], [1, 1], [5, 5]])
@pytest.mark.parametrize("key_class", [Categorical, CategoricalIndex])
def test_get_indexer_non_unique(self, idx_values, key_values, key_class):
@pytest.mark.parametrize("dtype", [None, "category", "key"])
def test_get_indexer_non_unique(self, idx_values, key_values, key_class, dtype):
# GH 21448
key = key_class(key_values, categories=range(1, 5))

if dtype == "key":
dtype = key.dtype

# Test for flat index and CategoricalIndex with same/different cats:
for dtype in [None, "category", key.dtype]:
idx = Index(idx_values, dtype=dtype)
expected, exp_miss = idx.get_indexer_non_unique(key_values)
result, res_miss = idx.get_indexer_non_unique(key)
idx = Index(idx_values, dtype=dtype)
expected, exp_miss = idx.get_indexer_non_unique(key_values)
result, res_miss = idx.get_indexer_non_unique(key)

tm.assert_numpy_array_equal(expected, result)
tm.assert_numpy_array_equal(exp_miss, res_miss)

tm.assert_numpy_array_equal(expected, result)
tm.assert_numpy_array_equal(exp_miss, res_miss)
exp_unique = idx.unique().get_indexer(key_values)
res_unique = idx.unique().get_indexer(key)
tm.assert_numpy_array_equal(res_unique, exp_unique)

def test_where_unobserved_nan(self):
ser = Series(Categorical(["a", "b"]))
Expand Down