TST: Add regression tests for fixed issues (#31161)

* TST: Add tests for fixed issues

* Platform compat test

* Use range instead

* Address comments

Author: mroeschke

pandas/tests/frame/test_constructors.py

@@ -2433,6 +2433,24 @@ def test_datetime_date_tuple_columns_from_dict(self):
         expected = DataFrame([0, 1, 2], columns=pd.Index(pd.Series([tup])))
         tm.assert_frame_equal(result, expected)
 
+    def test_construct_with_two_categoricalindex_series(self):
+        # GH 14600
+        s1 = pd.Series(
+            [39, 6, 4], index=pd.CategoricalIndex(["female", "male", "unknown"])
+        )
+        s2 = pd.Series(
+            [2, 152, 2, 242, 150],
+            index=pd.CategoricalIndex(["f", "female", "m", "male", "unknown"]),
+        )
+        result = pd.DataFrame([s1, s2])
+        expected = pd.DataFrame(
+            np.array(
+                [[np.nan, 39.0, np.nan, 6.0, 4.0], [2.0, 152.0, 2.0, 242.0, 150.0]]
+            ),
+            columns=["f", "female", "m", "male", "unknown"],
+        )
+        tm.assert_frame_equal(result, expected)
+
 
 class TestDataFrameConstructorWithDatetimeTZ:
     def test_from_dict(self):

pandas/tests/groupby/test_apply.py

@@ -792,3 +792,22 @@ def test_apply_multi_level_name(category):
     )
     tm.assert_frame_equal(result, expected)
     assert df.index.names == ["A", "B"]
+
+
+def test_groupby_apply_datetime_result_dtypes():
+    # GH 14849
+    data = pd.DataFrame.from_records(
+        [
+            (pd.Timestamp(2016, 1, 1), "red", "dark", 1, "8"),
+            (pd.Timestamp(2015, 1, 1), "green", "stormy", 2, "9"),
+            (pd.Timestamp(2014, 1, 1), "blue", "bright", 3, "10"),
+            (pd.Timestamp(2013, 1, 1), "blue", "calm", 4, "potato"),
+        ],
+        columns=["observation", "color", "mood", "intensity", "score"],
+    )
+    result = data.groupby("color").apply(lambda g: g.iloc[0]).dtypes
+    expected = Series(
+        [np.dtype("datetime64[ns]"), np.object, np.object, np.int64, np.object],
+        index=["observation", "color", "mood", "intensity", "score"],
+    )
+    tm.assert_series_equal(result, expected)

pandas/tests/groupby/test_groupby.py

@@ -1952,6 +1952,13 @@ def test_shift_bfill_ffill_tz(tz_naive_fixture, op, expected):
     tm.assert_frame_equal(result, expected)
 
 
+def test_ffill_missing_arguments():
+    # GH 14955
+    df = pd.DataFrame({"a": [1, 2], "b": [1, 1]})
+    with pytest.raises(ValueError, match="Must specify a fill"):
+        df.groupby("b").fillna()
+
+
 def test_groupby_only_none_group():
     # see GH21624
     # this was crashing with "ValueError: Length of passed values is 1, index implies 0"

pandas/tests/indexing/test_loc.py

@@ -1002,3 +1002,13 @@ def test_loc_axis_1_slice():
         ),
     )
     tm.assert_frame_equal(result, expected)
+
+
+def test_loc_set_dataframe_multiindex():
+    # GH 14592
+    expected = pd.DataFrame(
+        "a", index=range(2), columns=pd.MultiIndex.from_product([range(2), range(2)])
+    )
+    result = expected.copy()
+    result.loc[0, [(0, 1)]] = result.loc[0, [(0, 1)]]
+    tm.assert_frame_equal(result, expected)

pandas/tests/io/test_pickle.py

@@ -11,6 +11,7 @@
 3. Move the created pickle to "data/legacy_pickle/<version>" directory.
 """
 import bz2
+import datetime
 import glob
 import gzip
 import os
@@ -487,3 +488,17 @@ def open(self, *args):
         df.to_pickle(mockurl)
         result = pd.read_pickle(mockurl)
         tm.assert_frame_equal(df, result)
+
+
+class MyTz(datetime.tzinfo):
+    def __init__(self):
+        pass
+
+
+def test_read_pickle_with_subclass():
+    # GH 12163
+    expected = pd.Series(dtype=object), MyTz()
+    result = tm.round_trip_pickle(expected)
+
+    tm.assert_series_equal(result[0], expected[0])
+    assert isinstance(result[1], MyTz)

pandas/tests/series/test_constructors.py

@@ -1115,6 +1115,15 @@ def create_data(constructor):
         tm.assert_series_equal(result_datetime, expected)
         tm.assert_series_equal(result_Timestamp, expected)
 
+    def test_contructor_dict_tuple_indexer(self):
+        # GH 12948
+        data = {(1, 1, None): -1.0}
+        result = Series(data)
+        expected = Series(
+            -1.0, index=MultiIndex(levels=[[1], [1], [np.nan]], codes=[[0], [0], [-1]])
+        )
+        tm.assert_series_equal(result, expected)
+
     def test_constructor_mapping(self, non_mapping_dict_subclass):
         # GH 29788
         ndm = non_mapping_dict_subclass({3: "three"})

pandas/tests/test_multilevel.py

@@ -2147,6 +2147,40 @@ def test_sort_index_level_mixed(self):
             sorted_after.drop([("foo", "three")], axis=1),
         )
 
+    def test_sort_index_categorical_multiindex(self):
+        # GH 15058
+        df = DataFrame(
+            {
+                "a": range(6),
+                "l1": pd.Categorical(
+                    ["a", "a", "b", "b", "c", "c"],
+                    categories=["c", "a", "b"],
+                    ordered=True,
+                ),
+                "l2": [0, 1, 0, 1, 0, 1],
+            }
+        )
+        result = df.set_index(["l1", "l2"]).sort_index()
+        expected = DataFrame(
+            [4, 5, 0, 1, 2, 3],
+            columns=["a"],
+            index=MultiIndex(
+                levels=[
+                    pd.CategoricalIndex(
+                        ["c", "a", "b"],
+                        categories=["c", "a", "b"],
+                        ordered=True,
+                        name="l1",
+                        dtype="category",
+                    ),
+                    [0, 1],
+                ],
+                codes=[[0, 0, 1, 1, 2, 2], [0, 1, 0, 1, 0, 1]],
+                names=["l1", "l2"],
+            ),
+        )
+        tm.assert_frame_equal(result, expected)
+
     def test_is_lexsorted(self):
         levels = [[0, 1], [0, 1, 2]]