BUG: Aggregation on arrow array return same type.

Signed-off-by: Liang Yan <[email protected]>

Author: liang3zy22

pandas/core/dtypes/cast.py

@@ -463,7 +463,7 @@ def maybe_cast_pointwise_result(
     """
 
     if isinstance(dtype, ExtensionDtype):
-        if not isinstance(dtype, (CategoricalDtype, DatetimeTZDtype)):
+        if not isinstance(dtype, (CategoricalDtype, DatetimeTZDtype, ArrowDtype)):
             # TODO: avoid this special-casing
             # We have to special case categorical so as not to upcast
             # things like counts back to categorical
@@ -473,7 +473,17 @@ def maybe_cast_pointwise_result(
                 result = _maybe_cast_to_extension_array(cls, result, dtype=dtype)
             else:
                 result = _maybe_cast_to_extension_array(cls, result)
-
+        elif isinstance(dtype, ArrowDtype):
+            pyarrow_type = convert_dtypes(result, dtype_backend="pyarrow")
+            if isinstance(pyarrow_type, ExtensionDtype):
+                cls = pyarrow_type.construct_array_type()
+                result = _maybe_cast_to_extension_array(cls, result)
+            else:
+                cls = dtype.construct_array_type()
+                if same_dtype:
+                    result = _maybe_cast_to_extension_array(cls, result, dtype=dtype)
+                else:
+                    result = _maybe_cast_to_extension_array(cls, result)
     elif (numeric_only and dtype.kind in "iufcb") or not numeric_only:
         result = maybe_downcast_to_dtype(result, dtype)
 

pandas/tests/groupby/aggregate/test_aggregate.py

@@ -5,6 +5,7 @@
 import functools
 from functools import partial
 import re
+import typing
 
 import numpy as np
 import pytest
@@ -1630,4 +1631,25 @@ def test_groupby_agg_extension_timedelta_cumsum_with_named_aggregation():
     )
     gb = df.groupby("grps")
     result = gb.agg(td=("td", "cumsum"))
+
+
+@pytest.mark.skipif(
+    not typing.TYPE_CHECKING, reason="let pyarrow to be imported in dtypes.py"
+)
+def test_agg_arrow_type():
+    df = DataFrame.from_dict(
+        {
+            "category": ["A"] * 10 + ["B"] * 10,
+            "bool_numpy": [True] * 5 + [False] * 5 + [True] * 5 + [False] * 5,
+        }
+    )
+    df["bool_arrow"] = df["bool_numpy"].astype("bool[pyarrow]")
+    result = df.groupby("category").agg(lambda x: x.sum() / x.count())
+    expected = DataFrame(
+        {
+            "bool_numpy": [0.5, 0.5],
+            "bool_arrow": Series([0.5, 0.5]).astype("double[pyarrow]").values,
+        },
+        index=Index(["A", "B"], name="category"),
+    )
     tm.assert_frame_equal(result, expected)