BUG: RecursionError when attempting to replace np.nan values (#45725)

doc/source/whatsnew/v1.5.0.rst

@@ -248,6 +248,7 @@ Conversion
 - Bug in :meth:`Series.astype` and :meth:`DataFrame.astype` from floating dtype to unsigned integer dtype failing to raise in the presence of negative values (:issue:`45151`)
 - Bug in :func:`array` with ``FloatingDtype`` and values containing float-castable strings incorrectly raising (:issue:`45424`)
 - Bug when comparing string and datetime64ns objects causing ``OverflowError`` exception. (:issue:`45506`)
+- Bug when attempting to replace ``numpy.nan`` values causing ``RecursionError``
 
 Strings
 ^^^^^^^

pandas/core/internals/blocks.py

@@ -669,6 +669,7 @@ def replace(
         #  go through replace_list
 
         values = self.values
+        value = self._standardize_fill_value(value) # GH#45725
 
         if isinstance(values, Categorical):
             # TODO: avoid special-casing

pandas/tests/frame/methods/test_replace.py

@@ -661,6 +661,18 @@ def test_replace_simple_nested_dict_with_nonexistent_value(self):
         result = df.replace({"col": {-1: "-", 1: "a", 4: "b"}})
         tm.assert_frame_equal(expected, result)
 
+    def test_replace_numpy_nan(self):
+        # GH#45725 ensure np.nan can be replaced with pd.NA
+        df = pd.DataFrame({"A": [np.nan], "B": [np.nan]}, dtype=object)
+        result = df.replace({np.nan: pd.NA})
+        expected = pd.DataFrame({"A": [pd.NA], "B": [pd.NA]}, dtype=object)
+        tm.assert_frame_equal(result, expected)
+
+        # same thing but with None
+        result = df.replace({np.nan: None})
+        expected = pd.DataFrame({"A": [None], "B": [None]}, dtype=object)
+        tm.assert_frame_equal(result, expected)
+
     def test_replace_value_is_none(self, datetime_frame):
         orig_value = datetime_frame.iloc[0, 0]
         orig2 = datetime_frame.iloc[1, 0]

pandas/tests/series/methods/test_replace.py

@@ -36,6 +36,20 @@ def test_replace_explicit_none(self):
         assert expected.iloc[-1] is None
         tm.assert_series_equal(result, expected)
 
+    def test_replace_numpy_nan(self):
+        # GH#45725 ensure np.nan can be replaced
+        ser = pd.Series([np.nan, np.nan], dtype=object)
+        result = ser.replace({np.nan: pd.NA})
+        expected = pd.Series([pd.NA, pd.NA], dtype=object)
+        tm.assert_series_equal(result, expected)
+        assert result.dtype == object
+
+        # same thing but with None
+        result = ser.replace({np.nan: None})
+        expected = pd.Series([None, None], dtype=object)
+        tm.assert_series_equal(result, expected)
+        assert result.dtype == object
+
     def test_replace_noop_doesnt_downcast(self):
         # GH#44498
         ser = pd.Series([None, None, pd.Timestamp("2021-12-16 17:31")], dtype=object)