Backport PR #56179 on branch 2.1.x (BUG: to_numeric casting to ea for new string dtype) (#56263)

Backport PR #56179: BUG: to_numeric casting to ea for new string dtype

Co-authored-by: Patrick Hoefler <61934744+phofl@users.noreply.github.com>

Author: meeseeksmachine

doc/source/whatsnew/v2.1.4.rst

@@ -23,6 +23,7 @@ Bug fixes
 ~~~~~~~~~
 - Bug in :meth:`DataFrame.apply` where passing ``raw=True`` ignored ``args`` passed to the applied function (:issue:`55753`)
 - Bug in :meth:`Index.__getitem__` returning wrong result for Arrow dtypes and negative stepsize (:issue:`55832`)
+- Fixed bug in :func:`to_numeric` converting to extension dtype for ``string[pyarrow_numpy]`` dtype (:issue:`56179`)
 - Fixed bug in :meth:`DataFrame.__setitem__` casting :class:`Index` with object-dtype to PyArrow backed strings when ``infer_string`` option is set (:issue:`55638`)
 - Fixed bug in :meth:`DataFrame.to_hdf` raising when columns have ``StringDtype`` (:issue:`55088`)
 - Fixed bug in :meth:`Index.insert` casting object-dtype to PyArrow backed strings when ``infer_string`` option is set (:issue:`55638`)

pandas/core/tools/numeric.py

@@ -224,7 +224,8 @@ def to_numeric(
                 set(),
                 coerce_numeric=coerce_numeric,
                 convert_to_masked_nullable=dtype_backend is not lib.no_default
-                or isinstance(values_dtype, StringDtype),
+                or isinstance(values_dtype, StringDtype)
+                and not values_dtype.storage == "pyarrow_numpy",
             )
         except (ValueError, TypeError):
             if errors == "raise":
@@ -239,6 +240,7 @@ def to_numeric(
         dtype_backend is not lib.no_default
         and new_mask is None
         or isinstance(values_dtype, StringDtype)
+        and not values_dtype.storage == "pyarrow_numpy"
     ):
         new_mask = np.zeros(values.shape, dtype=np.bool_)
 

pandas/tests/tools/test_to_numeric.py

@@ -4,12 +4,15 @@
 from numpy import iinfo
 import pytest
 
+import pandas.util._test_decorators as td
+
 import pandas as pd
 from pandas import (
     ArrowDtype,
     DataFrame,
     Index,
     Series,
+    option_context,
     to_numeric,
 )
 import pandas._testing as tm
@@ -67,10 +70,14 @@ def test_empty(input_kwargs, result_kwargs):
     tm.assert_series_equal(result, expected)
 
 
+@pytest.mark.parametrize(
+    "infer_string", [False, pytest.param(True, marks=td.skip_if_no("pyarrow"))]
+)
 @pytest.mark.parametrize("last_val", ["7", 7])
-def test_series(last_val):
-    ser = Series(["1", "-3.14", last_val])
-    result = to_numeric(ser)
+def test_series(last_val, infer_string):
+    with option_context("future.infer_string", infer_string):
+        ser = Series(["1", "-3.14", last_val])
+        result = to_numeric(ser)
 
     expected = Series([1, -3.14, 7])
     tm.assert_series_equal(result, expected)