ENH: concat of nullable int + bool preserves int dtype (#34985)

Author: jorisvandenbossche

doc/source/whatsnew/v1.1.0.rst

@@ -322,7 +322,7 @@ Other enhancements
 - :meth:`DataFrame.hist`, :meth:`Series.hist`, :meth:`core.groupby.DataFrameGroupBy.hist`, and :meth:`core.groupby.SeriesGroupBy.hist` have gained the ``legend`` argument. Set to True to show a legend in the histogram. (:issue:`6279`)
 - :func:`concat` and :meth:`~DataFrame.append` now preserve extension dtypes, for example
   combining a nullable integer column with a numpy integer column will no longer
-  result in object dtype but preserve the integer dtype (:issue:`33607`, :issue:`34339`).
+  result in object dtype but preserve the integer dtype (:issue:`33607`, :issue:`34339`, :issue:`34095`).
 - :meth:`~pandas.io.gbq.read_gbq` now allows to disable progress bar (:issue:`33360`).
 - :meth:`~pandas.io.gbq.read_gbq` now supports the ``max_results`` kwarg from ``pandas-gbq`` (:issue:`34639`).
 - :meth:`DataFrame.cov` and :meth:`Series.cov` now support a new parameter ddof to support delta degrees of freedom as in the corresponding numpy methods (:issue:`34611`).

pandas/core/arrays/integer.py

@@ -92,10 +92,13 @@ def construct_array_type(cls) -> Type["IntegerArray"]:
         return IntegerArray
 
     def _get_common_dtype(self, dtypes: List[DtypeObj]) -> Optional[DtypeObj]:
-        # for now only handle other integer types
+        # we only handle nullable EA dtypes and numeric numpy dtypes
         if not all(
-            isinstance(t, _IntegerDtype)
-            or (isinstance(t, np.dtype) and np.issubdtype(t, np.integer))
+            isinstance(t, BaseMaskedDtype)
+            or (
+                isinstance(t, np.dtype)
+                and (np.issubdtype(t, np.number) or np.issubdtype(t, np.bool_))
+            )
             for t in dtypes
         ):
             return None

pandas/tests/arrays/integer/test_concat.py

@@ -1,3 +1,4 @@
+import numpy as np
 import pytest
 
 import pandas as pd
@@ -15,12 +16,52 @@
         (["Int32", "UInt32"], "Int64"),
         # this still gives object (awaiting float extension dtype)
         (["Int64", "UInt64"], "object"),
+        (["Int64", "boolean"], "Int64"),
+        (["UInt8", "boolean"], "UInt8"),
     ],
 )
 def test_concat_series(to_concat_dtypes, result_dtype):
 
-    result = pd.concat([pd.Series([1, 2, pd.NA], dtype=t) for t in to_concat_dtypes])
-    expected = pd.concat([pd.Series([1, 2, pd.NA], dtype=object)] * 2).astype(
+    result = pd.concat([pd.Series([0, 1, pd.NA], dtype=t) for t in to_concat_dtypes])
+    expected = pd.concat([pd.Series([0, 1, pd.NA], dtype=object)] * 2).astype(
         result_dtype
     )
     tm.assert_series_equal(result, expected)
+
+    # order doesn't matter for result
+    result = pd.concat(
+        [pd.Series([0, 1, pd.NA], dtype=t) for t in to_concat_dtypes[::-1]]
+    )
+    expected = pd.concat([pd.Series([0, 1, pd.NA], dtype=object)] * 2).astype(
+        result_dtype
+    )
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "to_concat_dtypes, result_dtype",
+    [
+        (["Int64", "int64"], "Int64"),
+        (["UInt64", "uint64"], "UInt64"),
+        (["Int8", "int8"], "Int8"),
+        (["Int8", "int16"], "Int16"),
+        (["UInt8", "int8"], "Int16"),
+        (["Int32", "uint32"], "Int64"),
+        # this still gives object (awaiting float extension dtype)
+        (["Int64", "uint64"], "object"),
+        (["Int64", "bool"], "Int64"),
+        (["UInt8", "bool"], "UInt8"),
+    ],
+)
+def test_concat_series_with_numpy(to_concat_dtypes, result_dtype):
+
+    s1 = pd.Series([0, 1, pd.NA], dtype=to_concat_dtypes[0])
+    s2 = pd.Series(np.array([0, 1], dtype=to_concat_dtypes[1]))
+    result = pd.concat([s1, s2], ignore_index=True)
+    expected = pd.Series([0, 1, pd.NA, 0, 1], dtype=object).astype(result_dtype)
+    tm.assert_series_equal(result, expected)
+
+    # order doesn't matter for result
+    result = pd.concat([s2, s1], ignore_index=True)
+    expected = pd.Series([0, 1, 0, 1, pd.NA], dtype=object).astype(result_dtype)
+    tm.assert_series_equal(result, expected)