DEPR: Enforce DataFrame(list_with_categorical) deprecation (#49592)

* DEPR: Enforce DataFrame(list_with_categorical) deprecation

* Update doc/source/whatsnew/v2.0.0.rst

Co-authored-by: Matthew Roeschke <10647082+mroeschke@users.noreply.github.com>

Co-authored-by: Matthew Roeschke <10647082+mroeschke@users.noreply.github.com>

Author: jbrockmendel

doc/source/whatsnew/v2.0.0.rst

@@ -482,6 +482,7 @@ Removal of prior version deprecations/changes
 - Changed behavior of :class:`DataFrame` constructor given floating-point ``data`` and an integer ``dtype``, when the data cannot be cast losslessly, the floating point dtype is retained, matching :class:`Series` behavior (:issue:`41170`)
 - Changed behavior of :class:`Index` constructor when given a ``np.ndarray`` with object-dtype containing numeric entries; this now retains object dtype rather than inferring a numeric dtype, consistent with :class:`Series` behavior (:issue:`42870`)
 - Changed behavior of :meth:`Index.__and__`, :meth:`Index.__or__` and :meth:`Index.__xor__` to behave as logical operations (matching :class:`Series` behavior) instead of aliases for set operations (:issue:`37374`)
+- Changed behavior of :class:`DataFrame` constructor when passed a list whose first element is a :class:`Categorical`, this now treats the elements as rows casting to ``object`` dtype, consistent with behavior for other types (:issue:`38845`)
 - Changed behavior of :class:`DataFrame` constructor when passed a ``dtype`` (other than int) that the data cannot be cast to; it now raises instead of silently ignoring the dtype (:issue:`41733`)
 - Changed the behavior of :class:`Series` constructor, it will no longer infer a datetime64 or timedelta64 dtype from string entries (:issue:`41731`)
 - Changed behavior of :class:`Timestamp` constructor with a ``np.datetime64`` object and a ``tz`` passed to interpret the input as a wall-time as opposed to a UTC time (:issue:`42288`)

pandas/core/internals/construction.py

@@ -10,7 +10,6 @@
     Hashable,
     Sequence,
 )
-import warnings
 
 import numpy as np
 from numpy import ma
@@ -22,7 +21,6 @@
     Manager,
     npt,
 )
-from pandas.util._exceptions import find_stack_level
 
 from pandas.core.dtypes.cast import (
     construct_1d_arraylike_from_scalar,
@@ -51,10 +49,7 @@
     algorithms,
     common as com,
 )
-from pandas.core.arrays import (
-    Categorical,
-    ExtensionArray,
-)
+from pandas.core.arrays import ExtensionArray
 from pandas.core.construction import (
     ensure_wrapped_if_datetimelike,
     extract_array,
@@ -476,9 +471,6 @@ def nested_data_to_arrays(
     if index is None:
         if isinstance(data[0], ABCSeries):
             index = _get_names_from_index(data)
-        elif isinstance(data[0], Categorical):
-            # GH#38845 hit in test_constructor_categorical
-            index = default_index(len(data[0]))
         else:
             index = default_index(len(data))
 
@@ -795,26 +787,6 @@ def to_arrays(
                 return arrays, columns
         return [], ensure_index([])
 
-    elif isinstance(data[0], Categorical):
-        # GH#38845 deprecate special case
-        warnings.warn(
-            "The behavior of DataFrame([categorical, ...]) is deprecated and "
-            "in a future version will be changed to match the behavior of "
-            "DataFrame([any_listlike, ...]). "
-            "To retain the old behavior, pass as a dictionary "
-            "DataFrame({col: categorical, ..})",
-            FutureWarning,
-            stacklevel=find_stack_level(),
-        )
-        if columns is None:
-            columns = default_index(len(data))
-        elif len(columns) > len(data):
-            raise ValueError("len(columns) > len(data)")
-        elif len(columns) < len(data):
-            # doing this here is akin to a pre-emptive reindex
-            data = data[: len(columns)]
-        return data, columns
-
     elif isinstance(data, np.ndarray) and data.dtype.names is not None:
         # e.g. recarray
         columns = Index(list(data.dtype.names))

pandas/tests/frame/test_constructors.py

@@ -2220,47 +2220,34 @@ def test_constructor_categorical(self):
         tm.assert_series_equal(df[0], expected)
 
     def test_construct_from_1item_list_of_categorical(self):
+        # pre-2.0 this behaved as DataFrame({0: cat}), in 2.0 we remove
+        #  Categorical special case
         # ndim != 1
-        msg = "will be changed to match the behavior"
-        with tm.assert_produces_warning(FutureWarning, match=msg):
-            df = DataFrame([Categorical(list("abc"))])
-        expected = DataFrame({0: Series(list("abc"), dtype="category")})
+        cat = Categorical(list("abc"))
+        df = DataFrame([cat])
+        expected = DataFrame([cat.astype(object)])
         tm.assert_frame_equal(df, expected)
 
     def test_construct_from_list_of_categoricals(self):
-        msg = "will be changed to match the behavior"
-        with tm.assert_produces_warning(FutureWarning, match=msg):
-            df = DataFrame([Categorical(list("abc")), Categorical(list("abd"))])
-        expected = DataFrame(
-            {
-                0: Series(list("abc"), dtype="category"),
-                1: Series(list("abd"), dtype="category"),
-            },
-            columns=[0, 1],
-        )
+        # pre-2.0 this behaved as DataFrame({0: cat}), in 2.0 we remove
+        #  Categorical special case
+
+        df = DataFrame([Categorical(list("abc")), Categorical(list("abd"))])
+        expected = DataFrame([["a", "b", "c"], ["a", "b", "d"]])
         tm.assert_frame_equal(df, expected)
 
     def test_from_nested_listlike_mixed_types(self):
+        # pre-2.0 this behaved as DataFrame({0: cat}), in 2.0 we remove
+        #  Categorical special case
         # mixed
-        msg = "will be changed to match the behavior"
-        with tm.assert_produces_warning(FutureWarning, match=msg):
-            df = DataFrame([Categorical(list("abc")), list("def")])
-        expected = DataFrame(
-            {0: Series(list("abc"), dtype="category"), 1: list("def")}, columns=[0, 1]
-        )
+        df = DataFrame([Categorical(list("abc")), list("def")])
+        expected = DataFrame([["a", "b", "c"], ["d", "e", "f"]])
         tm.assert_frame_equal(df, expected)
 
     def test_construct_from_listlikes_mismatched_lengths(self):
-        # invalid (shape)
-        msg = "|".join(
-            [
-                r"Length of values \(6\) does not match length of index \(3\)",
-            ]
-        )
-        msg2 = "will be changed to match the behavior"
-        with pytest.raises(ValueError, match=msg):
-            with tm.assert_produces_warning(FutureWarning, match=msg2):
-                DataFrame([Categorical(list("abc")), Categorical(list("abdefg"))])
+        df = DataFrame([Categorical(list("abc")), Categorical(list("abdefg"))])
+        expected = DataFrame([list("abc"), list("abdefg")])
+        tm.assert_frame_equal(df, expected)
 
     def test_constructor_categorical_series(self):