Merge branch '1.3.x' into auto-backport-of-pr-43150-on-1.3.x

Author: debnathshoham

doc/source/whatsnew/v1.3.2.rst

@@ -26,6 +26,7 @@ Fixed regressions
 - Fixed regression in :func:`concat` where ``copy=False`` was not honored in ``axis=1`` Series concatenation (:issue:`42501`)
 - Regression in :meth:`Series.nlargest` and :meth:`Series.nsmallest` with nullable integer or float dtype (:issue:`42816`)
 - Fixed regression in :meth:`Series.quantile` with :class:`Int64Dtype` (:issue:`42626`)
+- Fixed regression in :meth:`Series.groupby` and :meth:`DataFrame.groupby` where supplying the ``by`` argument with a Series named with a tuple would incorrectly raise (:issue:`42731`)
 
 .. ---------------------------------------------------------------------------
 

doc/source/whatsnew/v1.3.3.rst

@@ -18,9 +18,11 @@ Fixed regressions
 - Performance regression in :meth:`core.window.ewm.ExponentialMovingWindow.mean` (:issue:`42333`)
 - Fixed regression in :meth:`.GroupBy.agg` incorrectly raising in some cases (:issue:`42390`)
 - Fixed regression in :meth:`.GroupBy.quantile` which was failing with ``pandas.NA`` (:issue:`42849`)
+- Fixed regression in :meth:`.GroupBy.apply` where ``nan`` values were dropped even with ``dropna=False`` (:issue:`43205`)
 - Fixed regression in :meth:`merge` where ``on`` columns with ``ExtensionDtype`` or ``bool`` data types were cast to ``object`` in ``right`` and ``outer`` merge (:issue:`40073`)
 - Fixed regression in :meth:`RangeIndex.where` and :meth:`RangeIndex.putmask` raising ``AssertionError`` when result did not represent a :class:`RangeIndex` (:issue:`43240`)
 - Fixed regression in :meth:`read_parquet` where the ``fastparquet`` engine would not work properly with fastparquet 0.7.0 (:issue:`43075`)
+- Fixed regression in :func:`is_list_like` where objects with ``__iter__`` set to ``None`` would be identified as iterable (:issue:`43373`)
 
 .. ---------------------------------------------------------------------------
 

pandas/_libs/lib.pyx

@@ -1105,7 +1105,7 @@ def is_list_like(obj: object, allow_sets: bool = True) -> bool:
 cdef inline bint c_is_list_like(object obj, bint allow_sets) except -1:
     return (
         # equiv: `isinstance(obj, abc.Iterable)`
-        hasattr(obj, "__iter__") and not isinstance(obj, type)
+        getattr(obj, "__iter__", None) is not None and not isinstance(obj, type)
         # we do not count strings/unicode/bytes as list-like
         and not isinstance(obj, (str, bytes))
         # exclude zero-dimensional numpy arrays, effectively scalars

pandas/core/groupby/groupby.py

@@ -1013,7 +1013,11 @@ def reset_identity(values):
 
         if not not_indexed_same:
             result = concat(values, axis=self.axis)
-            ax = self.filter(lambda x: True).axes[self.axis]
+            ax = (
+                self.filter(lambda x: True).axes[self.axis]
+                if self.dropna
+                else self._selected_obj._get_axis(self.axis)
+            )
 
             # this is a very unfortunate situation
             # we can't use reindex to restore the original order

pandas/core/groupby/grouper.py

@@ -833,9 +833,11 @@ def is_in_obj(gpr) -> bool:
             return False
         try:
             return gpr is obj[gpr.name]
-        except (KeyError, IndexError):
+        except (KeyError, IndexError, InvalidIndexError):
             # IndexError reached in e.g. test_skip_group_keys when we pass
             #  lambda here
+            # InvalidIndexError raised on key-types inappropriate for index,
+            #  e.g. DatetimeIndex.get_loc(tuple())
             return False
 
     for gpr, level in zip(keys, levels):

pandas/tests/dtypes/test_inference.py

@@ -150,6 +150,18 @@ def foo():
         foo()
 
 
+def test_is_list_like_iter_is_none():
+    # GH 43373
+    # is_list_like was yielding false positives with __iter__ == None
+    class NotListLike:
+        def __getitem__(self, item):
+            return self
+
+        __iter__ = None
+
+    assert not inference.is_list_like(NotListLike())
+
+
 def test_is_sequence():
     is_seq = inference.is_sequence
     assert is_seq((1, 2))

pandas/tests/groupby/test_apply.py

@@ -1102,25 +1102,19 @@ def test_apply_by_cols_equals_apply_by_rows_transposed():
     tm.assert_frame_equal(by_cols, df)
 
 
-def test_apply_dropna_with_indexed_same():
+@pytest.mark.parametrize("dropna", [True, False])
+def test_apply_dropna_with_indexed_same(dropna):
     # GH 38227
-
+    # GH#43205
     df = DataFrame(
         {
             "col": [1, 2, 3, 4, 5],
             "group": ["a", np.nan, np.nan, "b", "b"],
         },
         index=list("xxyxz"),
     )
-    result = df.groupby("group").apply(lambda x: x)
-    expected = DataFrame(
-        {
-            "col": [1, 4, 5],
-            "group": ["a", "b", "b"],
-        },
-        index=list("xxz"),
-    )
-
+    result = df.groupby("group", dropna=dropna).apply(lambda x: x)
+    expected = df.dropna() if dropna else df.iloc[[0, 3, 1, 2, 4]]
     tm.assert_frame_equal(result, expected)
 
 

pandas/tests/groupby/test_grouping.py

@@ -396,6 +396,23 @@ def test_groupby_dict_mapping(self):
         tm.assert_series_equal(result, result2)
         tm.assert_series_equal(result, expected2)
 
+    @pytest.mark.parametrize(
+        "index",
+        [
+            [0, 1, 2, 3],
+            ["a", "b", "c", "d"],
+            [Timestamp(2021, 7, 28 + i) for i in range(4)],
+        ],
+    )
+    def test_groupby_series_named_with_tuple(self, frame_or_series, index):
+        # GH 42731
+        obj = frame_or_series([1, 2, 3, 4], index=index)
+        groups = Series([1, 0, 1, 0], index=index, name=("a", "a"))
+        result = obj.groupby(groups).last()
+        expected = frame_or_series([4, 3])
+        expected.index.name = ("a", "a")
+        tm.assert_equal(result, expected)
+
     def test_groupby_grouper_f_sanity_checked(self):
         dates = date_range("01-Jan-2013", periods=12, freq="MS")
         ts = Series(np.random.randn(12), index=dates)