BUG: mode not sorting values for arrow backed strings (pandas-dev#55621)

* BUG: mode not sorting values for arrow backed strings

* Fix tests

* Change to pa_installed variable

* Update pyarrow.py

* Fix

* Fix

(cherry picked from commit bb2d2e0)

doc/source/whatsnew/v2.1.2.rst

@@ -31,6 +31,7 @@ Bug fixes
 - Fixed bug in :meth:`Index.insert` raising when inserting ``None`` into :class:`Index` with ``dtype="string[pyarrow_numpy]"`` (:issue:`55365`)
 - Fixed bug in :meth:`Series.all`  and :meth:`Series.any` not treating missing values correctly for ``dtype="string[pyarrow_numpy]"`` (:issue:`55367`)
 - Fixed bug in :meth:`Series.floordiv` for :class:`ArrowDtype` (:issue:`55561`)
+- Fixed bug in :meth:`Series.mode` not sorting values for arrow backed string dtype (:issue:`55621`)
 - Fixed bug in :meth:`Series.rank` for ``string[pyarrow_numpy]`` dtype (:issue:`55362`)
 - Fixed bug in :meth:`Series.str.extractall` for :class:`ArrowDtype` dtype being converted to object (:issue:`53846`)
 - Fixed bug where PDEP-6 warning about setting an item of an incompatible dtype was being shown when creating a new conditional column (:issue:`55025`)

pandas/core/arrays/arrow/array.py

@@ -1886,6 +1886,7 @@ def _mode(self, dropna: bool = True) -> Self:
         if pa.types.is_temporal(pa_type):
             most_common = most_common.cast(pa_type)
 
+        most_common = most_common.take(pc.array_sort_indices(most_common))
         return type(self)(most_common)
 
     def _maybe_convert_setitem_value(self, value):

pandas/tests/extension/test_arrow.py

@@ -1474,7 +1474,7 @@ def test_quantile(data, interpolation, quantile, request):
 
 @pytest.mark.parametrize(
     "take_idx, exp_idx",
-    [[[0, 0, 2, 2, 4, 4], [0, 4]], [[0, 0, 0, 2, 4, 4], [0]]],
+    [[[0, 0, 2, 2, 4, 4], [4, 0]], [[0, 0, 0, 2, 4, 4], [0]]],
     ids=["multi_mode", "single_mode"],
 )
 def test_mode_dropna_true(data_for_grouping, take_idx, exp_idx):
@@ -1492,7 +1492,7 @@ def test_mode_dropna_false_mode_na(data):
     expected = pd.Series([None], dtype=data.dtype)
     tm.assert_series_equal(result, expected)
 
-    expected = pd.Series([None, data[0]], dtype=data.dtype)
+    expected = pd.Series([data[0], None], dtype=data.dtype)
     result = expected.mode(dropna=False)
     tm.assert_series_equal(result, expected)
 

pandas/tests/groupby/test_groupby.py

@@ -5,11 +5,11 @@
 import numpy as np
 import pytest
 
-from pandas.compat import pa_version_under7p0
 from pandas.errors import (
     PerformanceWarning,
     SpecificationError,
 )
+import pandas.util._test_decorators as td
 
 import pandas as pd
 from pandas import (
@@ -2518,10 +2518,7 @@ def test_groupby_column_index_name_lost(func):
     "infer_string",
     [
         False,
-        pytest.param(
-            True,
-            marks=pytest.mark.skipif(pa_version_under7p0, reason="arrow not installed"),
-        ),
+        pytest.param(True, marks=td.skip_if_no("pyarrow")),
     ],
 )
 def test_groupby_duplicate_columns(infer_string):
@@ -2751,13 +2748,20 @@ def test_rolling_wrong_param_min_period():
         test_df.groupby("name")["val"].rolling(window=2, min_period=1).sum()
 
 
-def test_by_column_values_with_same_starting_value():
+@pytest.mark.parametrize(
+    "dtype",
+    [
+        object,
+        pytest.param("string[pyarrow_numpy]", marks=td.skip_if_no("pyarrow")),
+    ],
+)
+def test_by_column_values_with_same_starting_value(dtype):
     # GH29635
     df = DataFrame(
         {
             "Name": ["Thomas", "Thomas", "Thomas John"],
             "Credit": [1200, 1300, 900],
-            "Mood": ["sad", "happy", "happy"],
+            "Mood": Series(["sad", "happy", "happy"], dtype=dtype),
         }
     )
     aggregate_details = {"Mood": Series.mode, "Credit": "sum"}