add explicit tests and fix insert for empty object dtype

Author: jorisvandenbossche

pandas/core/indexes/base.py

@@ -6242,6 +6242,7 @@ def _find_common_type_compat(self, target) -> DtypeObj:
             # special case: if left or right is a zero-length RangeIndex or
             # Index[object], those can be created by the default empty constructors
             # -> for that case ignore this dtype and always return the other
+            # (https://github.com/pandas-dev/pandas/pull/60797)
             from pandas.core.indexes.range import RangeIndex
 
             if len(self) == 0 and (
@@ -6908,6 +6909,14 @@ def insert(self, loc: int, item) -> Index:
 
         arr = self._values
 
+        if using_string_dtype and len(self) == 0 and self.dtype == np.object_:
+            # special case: if we are an empty object-dtype Index, also
+            # take into account the inserted item for the resulting dtype
+            # (https://github.com/pandas-dev/pandas/pull/60797)
+            dtype = self._find_common_type_compat(item)
+            if dtype != self.dtype:
+                return self.astype(dtype).insert(loc, item)
+
         try:
             if isinstance(arr, ExtensionArray):
                 res_values = arr.insert(loc, item)

pandas/tests/frame/indexing/test_setitem.py

@@ -144,8 +144,9 @@ def test_setitem_different_dtype(self):
         )
         tm.assert_series_equal(result, expected)
 
-    def test_setitem_empty_columns(self):
-        # GH 13522
+    def test_setitem_overwrite_index(self):
+        # GH 13522 - assign the index as a column and then overwrite the values
+        # -> should not affect the index
         df = DataFrame(index=["A", "B", "C"])
         df["X"] = df.index
         df["X"] = ["x", "y", "z"]
@@ -154,6 +155,21 @@ def test_setitem_empty_columns(self):
         )
         tm.assert_frame_equal(df, exp)
 
+    def test_setitem_empty_columns(self):
+        # Starting from an empty DataFrame and setting a column should result
+        # in a default string dtype for the columns' Index
+        # https://github.com/pandas-dev/pandas/issues/60338
+
+        df = DataFrame()
+        df["foo"] = [1, 2, 3]
+        expected = DataFrame({"foo": [1, 2, 3]})
+        tm.assert_frame_equal(df, expected)
+
+        df = DataFrame(columns=Index([]))
+        df["foo"] = [1, 2, 3]
+        expected = DataFrame({"foo": [1, 2, 3]})
+        tm.assert_frame_equal(df, expected)
+
     def test_setitem_dt64_index_empty_columns(self):
         rng = date_range("1/1/2000 00:00:00", "1/1/2000 1:59:50", freq="10s")
         df = DataFrame(index=np.arange(len(rng)))

pandas/tests/frame/methods/test_reset_index.py

@@ -778,3 +778,34 @@ def test_reset_index_false_index_name():
     result_frame.reset_index()
     expected_frame = DataFrame(range(5, 10), RangeIndex(range(5), name=False))
     tm.assert_frame_equal(result_frame, expected_frame)
+
+
+@pytest.mark.parametrize("columns", [None, Index([])])
+def test_reset_index_with_empty_frame(columns):
+    # Currently empty DataFrame has RangeIndex or object dtype Index, but when
+    # resetting the index we still want to end up with the default string dtype
+    # https://github.com/pandas-dev/pandas/issues/60338
+
+    index = Index([], name="foo")
+    df = DataFrame(index=index, columns=columns)
+    result = df.reset_index()
+    expected = DataFrame(columns=["foo"])
+    tm.assert_frame_equal(result, expected)
+
+    index = Index([1, 2, 3], name="foo")
+    df = DataFrame(index=index, columns=columns)
+    result = df.reset_index()
+    expected = DataFrame({"foo": [1, 2, 3]})
+    tm.assert_frame_equal(result, expected)
+
+    index = MultiIndex.from_tuples([], names=["foo", "bar"])
+    df = DataFrame(index=index, columns=columns)
+    result = df.reset_index()
+    expected = DataFrame(columns=["foo", "bar"])
+    tm.assert_frame_equal(result, expected)
+
+    index = MultiIndex.from_tuples([(1, 2), (2, 3)], names=["foo", "bar"])
+    df = DataFrame(index=index, columns=columns)
+    result = df.reset_index()
+    expected = DataFrame({"foo": [1, 2], "bar": [2, 3]})
+    tm.assert_frame_equal(result, expected)

pandas/tests/indexes/base_class/test_reshape.py

@@ -34,7 +34,7 @@ def test_insert(self):
 
         # test empty
         null_index = Index([])
-        tm.assert_index_equal(Index(["a"], dtype=object), null_index.insert(0, "a"))
+        tm.assert_index_equal(Index(["a"]), null_index.insert(0, "a"))
 
     def test_insert_missing(self, nulls_fixture, using_infer_string):
         # GH#22295

pandas/tests/indexes/test_old_base.py

@@ -454,10 +454,12 @@ def test_insert_out_of_bounds(self, index, using_infer_string):
         else:
             msg = "slice indices must be integers or None or have an __index__ method"
 
-        if using_infer_string and (
-            index.dtype == "string" or index.dtype == "category"
-        ):
-            msg = "loc must be an integer between"
+        if using_infer_string:
+            if index.dtype == "string" or index.dtype == "category":
+                msg = "loc must be an integer between"
+            elif index.dtype == "object" and len(index) == 0:
+                msg = "loc must be an integer between"
+                err = TypeError
 
         with pytest.raises(err, match=msg):
             index.insert(0.5, "foo")