Backport PR #47327 on branch 1.4.x (REGR: Fix fillna making a copy wh…

…en dict was given as fill value and inplace is set) (#47448)

* Backport PR #47327: REGR: Fix fillna making a copy when dict was given as fill value and inplace is set

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

doc/source/whatsnew/v1.4.3.rst

@@ -29,6 +29,7 @@ Fixed regressions
 - Fixed regression in :meth:`DataFrame.to_csv` raising error when :class:`DataFrame` contains extension dtype categorical column (:issue:`46297`, :issue:`46812`)
 - Fixed regression in representation of ``dtypes`` attribute of :class:`MultiIndex` (:issue:`46900`)
 - Fixed regression when setting values with :meth:`DataFrame.loc` updating :class:`RangeIndex` when index was set as new column and column was updated afterwards (:issue:`47128`)
+- Fixed regression in :meth:`DataFrame.fillna` and :meth:`DataFrame.update` creating a copy when updating inplace (:issue:`47188`)
 - Fixed regression in :meth:`DataFrame.nsmallest` led to wrong results when ``np.nan`` in the sorting column (:issue:`46589`)
 - Fixed regression in :func:`read_fwf` raising ``ValueError`` when ``widths`` was specified with ``usecols`` (:issue:`46580`)
 - Fixed regression in :func:`concat` not sorting columns for mixed column names (:issue:`47127`)

pandas/core/frame.py

@@ -7594,7 +7594,7 @@ def update(
             if mask.all():
                 continue
 
-            self[col] = expressions.where(mask, this, that)
+            self.loc[:, col] = expressions.where(mask, this, that)
 
     # ----------------------------------------------------------------------
     # Data reshaping

pandas/core/generic.py

@@ -6482,7 +6482,9 @@ def fillna(
                     if k not in result:
                         continue
                     downcast_k = downcast if not is_dict else downcast.get(k)
-                    result[k] = result[k].fillna(v, limit=limit, downcast=downcast_k)
+                    result.loc[:, k] = result[k].fillna(
+                        v, limit=limit, downcast=downcast_k
+                    )
                 return result if not inplace else None
 
             elif not is_list_like(value):

pandas/tests/frame/methods/test_fillna.py

@@ -265,6 +265,7 @@ def test_fillna_downcast_false(self, frame_or_series):
         result = obj.fillna("", downcast=False)
         tm.assert_equal(result, obj)
 
+    @td.skip_array_manager_invalid_test
     @pytest.mark.parametrize("columns", [["A", "A", "B"], ["A", "A"]])
     def test_fillna_dictlike_value_duplicate_colnames(self, columns):
         # GH#43476
@@ -654,6 +655,17 @@ def test_fillna_inplace_with_columns_limit_and_value(self):
         df.fillna(axis=1, value=100, limit=1, inplace=True)
         tm.assert_frame_equal(df, expected)
 
+    @td.skip_array_manager_invalid_test
+    @pytest.mark.parametrize("val", [-1, {"x": -1, "y": -1}])
+    def test_inplace_dict_update_view(self, val):
+        # GH#47188
+        df = DataFrame({"x": [np.nan, 2], "y": [np.nan, 2]})
+        result_view = df[:]
+        df.fillna(val, inplace=True)
+        expected = DataFrame({"x": [-1, 2.0], "y": [-1.0, 2]})
+        tm.assert_frame_equal(df, expected)
+        tm.assert_frame_equal(result_view, expected)
+
 
 def test_fillna_nonconsolidated_frame():
     # https://github.com/pandas-dev/pandas/issues/36495

pandas/tests/frame/methods/test_update.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+import pandas.util._test_decorators as td
+
 import pandas as pd
 from pandas import (
     DataFrame,
@@ -146,3 +148,14 @@ def test_update_with_different_dtype(self):
 
         expected = DataFrame({"a": [1, 3], "b": [np.nan, 2], "c": ["foo", np.nan]})
         tm.assert_frame_equal(df, expected)
+
+    @td.skip_array_manager_invalid_test
+    def test_update_modify_view(self):
+        # GH#47188
+        df = DataFrame({"A": ["1", np.nan], "B": ["100", np.nan]})
+        df2 = DataFrame({"A": ["a", "x"], "B": ["100", "200"]})
+        result_view = df2[:]
+        df2.update(df)
+        expected = DataFrame({"A": ["1", "x"], "B": ["100", "200"]})
+        tm.assert_frame_equal(df2, expected)
+        tm.assert_frame_equal(result_view, expected)