DEPR: inplace kwarg in set_index (pandas-dev#48115)

* DEPR: inplace kwarg in set_index

* GH ref

doc/source/user_guide/indexing.rst

@@ -1723,13 +1723,12 @@ the given columns to a MultiIndex:
    frame
 
 Other options in ``set_index`` allow you not drop the index columns or to add
-the index in-place (without creating a new object):
+the index without creating a copy of the underlying data:
 
 .. ipython:: python
 
    data.set_index('c', drop=False)
-   data.set_index(['a', 'b'], inplace=True)
-   data
+   data.set_index(['a', 'b'], copy=False)
 
 Reset the index
 ~~~~~~~~~~~~~~~

doc/source/whatsnew/v1.5.0.rst

@@ -849,6 +849,7 @@ Other Deprecations
 - Deprecated unused arguments ``encoding`` and ``verbose`` in :meth:`Series.to_excel` and :meth:`DataFrame.to_excel` (:issue:`47912`)
 - Deprecated producing a single element when iterating over a :class:`DataFrameGroupBy` or a :class:`SeriesGroupBy` that has been grouped by a list of length 1; A tuple of length one will be returned instead (:issue:`42795`)
 - Fixed up warning message of deprecation of :meth:`MultiIndex.lesort_depth` as public method, as the message previously referred to :meth:`MultiIndex.is_lexsorted` instead (:issue:`38701`)
+- Deprecated the ``inplace`` keyword in :meth:`DataFrame.set_index`, use ``df = df.set_index(..., copy=False)`` instead (:issue:`48115`)
 - Deprecated the ``sort_columns`` argument in :meth:`DataFrame.plot` and :meth:`Series.plot` (:issue:`47563`).
 - Deprecated positional arguments for all but the first argument of :meth:`DataFrame.to_stata` and :func:`read_stata`, use keyword arguments instead (:issue:`48128`).
 

pandas/core/frame.py

@@ -5819,7 +5819,7 @@ def set_index(
         *,
         drop: bool = ...,
         append: bool = ...,
-        inplace: Literal[False] = ...,
+        inplace: Literal[False] | lib.NoDefault = ...,
         verify_integrity: bool = ...,
         copy: bool | lib.NoDefault = ...,
     ) -> DataFrame:
@@ -5844,7 +5844,7 @@ def set_index(
         keys,
         drop: bool = True,
         append: bool = False,
-        inplace: bool = False,
+        inplace: bool | lib.NoDefault = lib.no_default,
         verify_integrity: bool = False,
         copy: bool | lib.NoDefault = lib.no_default,
     ) -> DataFrame | None:
@@ -5869,6 +5869,9 @@ def set_index(
             Whether to append columns to existing index.
         inplace : bool, default False
             Whether to modify the DataFrame rather than creating a new one.
+
+            .. deprecated:: 1.5.0
+
         verify_integrity : bool, default False
             Check the new index for duplicates. Otherwise defer the check until
             necessary. Setting to False will improve the performance of this
@@ -5942,7 +5945,18 @@ def set_index(
         3 9       7  2013    84
         4 16     10  2014    31
         """
-        inplace = validate_bool_kwarg(inplace, "inplace")
+        if inplace is not lib.no_default:
+            inplace = validate_bool_kwarg(inplace, "inplace")
+            warnings.warn(
+                "The 'inplace' keyword in DataFrame.set_index is deprecated "
+                "and will be removed in a future version. Use "
+                "`df = df.set_index(..., copy=False)` instead.",
+                FutureWarning,
+                stacklevel=find_stack_level(inspect.currentframe()),
+            )
+        else:
+            inplace = False
+
         if inplace:
             if copy is not lib.no_default:
                 raise ValueError("Cannot specify copy when inplace=True")

pandas/core/reshape/merge.py

@@ -782,9 +782,9 @@ def get_result(self, copy: bool = True) -> DataFrame:
         if self.indicator:
             result = self._indicator_post_merge(result)
 
-        self._maybe_add_join_keys(result, left_indexer, right_indexer)
+        result = self._maybe_add_join_keys(result, left_indexer, right_indexer)
 
-        self._maybe_restore_index_levels(result)
+        result = self._maybe_restore_index_levels(result)
 
         self._maybe_drop_cross_column(result, self._cross)
 
@@ -851,7 +851,7 @@ def _indicator_post_merge(self, result: DataFrame) -> DataFrame:
         result = result.drop(labels=["_left_indicator", "_right_indicator"], axis=1)
         return result
 
-    def _maybe_restore_index_levels(self, result: DataFrame) -> None:
+    def _maybe_restore_index_levels(self, result: DataFrame) -> DataFrame:
         """
         Restore index levels specified as `on` parameters
 
@@ -869,7 +869,7 @@ def _maybe_restore_index_levels(self, result: DataFrame) -> None:
 
         Returns
         -------
-        None
+        DataFrame
         """
         names_to_restore = []
         for name, left_key, right_key in zip(
@@ -893,14 +893,15 @@ def _maybe_restore_index_levels(self, result: DataFrame) -> None:
                 names_to_restore.append(name)
 
         if names_to_restore:
-            result.set_index(names_to_restore, inplace=True)
+            result = result.set_index(names_to_restore, copy=False)
+        return result
 
     def _maybe_add_join_keys(
         self,
         result: DataFrame,
         left_indexer: np.ndarray | None,
         right_indexer: np.ndarray | None,
-    ) -> None:
+    ) -> DataFrame:
 
         left_has_missing = None
         right_has_missing = None
@@ -996,11 +997,12 @@ def _maybe_add_join_keys(
                             for level_name in result.index.names
                         ]
 
-                        result.set_index(idx_list, inplace=True)
+                        result = result.set_index(idx_list, copy=False)
                     else:
                         result.index = Index(key_col, name=name)
                 else:
                     result.insert(i, name or f"key_{i}", key_col)
+        return result
 
     def _get_join_indexers(self) -> tuple[npt.NDArray[np.intp], npt.NDArray[np.intp]]:
         """return the join indexers"""
@@ -1768,7 +1770,8 @@ def get_result(self, copy: bool = True) -> DataFrame:
         result = self._reindex_and_concat(
             join_index, left_join_indexer, right_join_indexer, copy=copy
         )
-        self._maybe_add_join_keys(result, left_indexer, right_indexer)
+
+        result = self._maybe_add_join_keys(result, left_indexer, right_indexer)
 
         return result
 

pandas/io/parsers/arrow_parser_wrapper.py

@@ -117,7 +117,7 @@ def _finalize_output(self, frame: DataFrame) -> DataFrame:
                     # String case
                     if item not in frame.columns:
                         raise ValueError(f"Index {item} invalid")
-            frame.set_index(self.index_col, drop=True, inplace=True)
+            frame = frame.set_index(self.index_col, drop=True, copy=False)
             # Clear names if headerless and no name given
             if self.header is None and not multi_index_named:
                 frame.index.names = [None] * len(frame.index.names)

pandas/io/pytables.py

@@ -4663,7 +4663,7 @@ def read(
                     columns.insert(0, n)
         s = super().read(where=where, columns=columns, start=start, stop=stop)
         if is_multi_index:
-            s.set_index(self.levels, inplace=True)
+            s = s.set_index(self.levels, copy=False)
 
         s = s.iloc[:, 0]
 

pandas/io/sql.py

@@ -152,7 +152,7 @@ def _wrap_result(
     frame = _parse_date_columns(frame, parse_dates)
 
     if index_col is not None:
-        frame.set_index(index_col, inplace=True)
+        frame = frame.set_index(index_col, copy=False)
 
     return frame
 
@@ -979,7 +979,7 @@ def _query_iterator(
                 self._harmonize_columns(parse_dates=parse_dates)
 
                 if self.index is not None:
-                    self.frame.set_index(self.index, inplace=True)
+                    self.frame = self.frame.set_index(self.index, copy=False)
 
                 yield self.frame
 
@@ -1020,7 +1020,7 @@ def read(
             self._harmonize_columns(parse_dates=parse_dates)
 
             if self.index is not None:
-                self.frame.set_index(self.index, inplace=True)
+                self.frame = self.frame.set_index(self.index, copy=False)
 
             return self.frame
 

pandas/tests/frame/methods/test_combine_first.py

@@ -387,8 +387,8 @@ def test_combine_first_string_dtype_only_na(self, nullable_string_dtype):
             {"a": ["962", "85"], "b": [pd.NA] * 2}, dtype=nullable_string_dtype
         )
         df2 = DataFrame({"a": ["85"], "b": [pd.NA]}, dtype=nullable_string_dtype)
-        df.set_index(["a", "b"], inplace=True)
-        df2.set_index(["a", "b"], inplace=True)
+        df = df.set_index(["a", "b"], copy=False)
+        df2 = df2.set_index(["a", "b"], copy=False)
         result = df.combine_first(df2)
         expected = DataFrame(
             {"a": ["962", "85"], "b": [pd.NA] * 2}, dtype=nullable_string_dtype

pandas/tests/frame/methods/test_set_index.py

@@ -40,9 +40,11 @@ def test_set_index_copy(self):
 
         msg = "Cannot specify copy when inplace=True"
         with pytest.raises(ValueError, match=msg):
-            df.set_index("A", inplace=True, copy=True)
+            with tm.assert_produces_warning(FutureWarning, match="The 'inplace'"):
+                df.set_index("A", inplace=True, copy=True)
         with pytest.raises(ValueError, match=msg):
-            df.set_index("A", inplace=True, copy=False)
+            with tm.assert_produces_warning(FutureWarning, match="The 'inplace'"):
+                df.set_index("A", inplace=True, copy=False)
 
     def test_set_index_multiindex(self):
         # segfault in GH#3308
@@ -197,7 +199,10 @@ def test_set_index_drop_inplace(self, frame_of_index_cols, drop, inplace, keys):
 
         if inplace:
             result = df.copy()
-            return_value = result.set_index(keys, drop=drop, inplace=True)
+            with tm.assert_produces_warning(
+                FutureWarning, match="The 'inplace' keyword"
+            ):
+                return_value = result.set_index(keys, drop=drop, inplace=True)
             assert return_value is None
         else:
             result = df.set_index(keys, drop=drop)

pandas/tests/frame/test_api.py

@@ -244,7 +244,8 @@ def _check_f(base, f):
 
         # set_index
         f = lambda x: x.set_index("a", inplace=True)
-        _check_f(data.copy(), f)
+        with tm.assert_produces_warning(FutureWarning, match="The 'inplace' keyword"):
+            _check_f(data.copy(), f)
 
         # reset_index
         f = lambda x: x.reset_index(inplace=True)

pandas/tests/frame/test_query_eval.py

@@ -436,7 +436,8 @@ def test_date_index_query(self):
         df = DataFrame(np.random.randn(n, 3))
         df["dates1"] = date_range("1/1/2012", periods=n)
         df["dates3"] = date_range("1/1/2014", periods=n)
-        return_value = df.set_index("dates1", inplace=True, drop=True)
+        with tm.assert_produces_warning(FutureWarning, match="The 'inplace' keyword"):
+            return_value = df.set_index("dates1", inplace=True, drop=True)
         assert return_value is None
         res = df.query("index < 20130101 < dates3", engine=engine, parser=parser)
         expec = df[(df.index < "20130101") & ("20130101" < df.dates3)]
@@ -449,7 +450,8 @@ def test_date_index_query_with_NaT(self):
         df["dates1"] = date_range("1/1/2012", periods=n)
         df["dates3"] = date_range("1/1/2014", periods=n)
         df.iloc[0, 0] = pd.NaT
-        return_value = df.set_index("dates1", inplace=True, drop=True)
+        with tm.assert_produces_warning(FutureWarning, match="The 'inplace' keyword"):
+            return_value = df.set_index("dates1", inplace=True, drop=True)
         assert return_value is None
         res = df.query("index < 20130101 < dates3", engine=engine, parser=parser)
         expec = df[(df.index < "20130101") & ("20130101" < df.dates3)]
@@ -463,7 +465,8 @@ def test_date_index_query_with_NaT_duplicates(self):
         d["dates3"] = date_range("1/1/2014", periods=n)
         df = DataFrame(d)
         df.loc[np.random.rand(n) > 0.5, "dates1"] = pd.NaT
-        return_value = df.set_index("dates1", inplace=True, drop=True)
+        with tm.assert_produces_warning(FutureWarning, match="The 'inplace' keyword"):
+            return_value = df.set_index("dates1", inplace=True, drop=True)
         assert return_value is None
         res = df.query("dates1 < 20130101 < dates3", engine=engine, parser=parser)
         expec = df[(df.index.to_series() < "20130101") & ("20130101" < df.dates3)]
@@ -794,7 +797,8 @@ def test_date_index_query(self):
         df = DataFrame(np.random.randn(n, 3))
         df["dates1"] = date_range("1/1/2012", periods=n)
         df["dates3"] = date_range("1/1/2014", periods=n)
-        return_value = df.set_index("dates1", inplace=True, drop=True)
+        with tm.assert_produces_warning(FutureWarning, match="The 'inplace' keyword"):
+            return_value = df.set_index("dates1", inplace=True, drop=True)
         assert return_value is None
         res = df.query(
             "(index < 20130101) & (20130101 < dates3)", engine=engine, parser=parser
@@ -809,7 +813,8 @@ def test_date_index_query_with_NaT(self):
         df["dates1"] = date_range("1/1/2012", periods=n)
         df["dates3"] = date_range("1/1/2014", periods=n)
         df.iloc[0, 0] = pd.NaT
-        return_value = df.set_index("dates1", inplace=True, drop=True)
+        with tm.assert_produces_warning(FutureWarning, match="The 'inplace' keyword"):
+            return_value = df.set_index("dates1", inplace=True, drop=True)
         assert return_value is None
         res = df.query(
             "(index < 20130101) & (20130101 < dates3)", engine=engine, parser=parser
@@ -824,7 +829,8 @@ def test_date_index_query_with_NaT_duplicates(self):
         df["dates1"] = date_range("1/1/2012", periods=n)
         df["dates3"] = date_range("1/1/2014", periods=n)
         df.loc[np.random.rand(n) > 0.5, "dates1"] = pd.NaT
-        return_value = df.set_index("dates1", inplace=True, drop=True)
+        with tm.assert_produces_warning(FutureWarning, match="The 'inplace' keyword"):
+            return_value = df.set_index("dates1", inplace=True, drop=True)
         assert return_value is None
         msg = r"'BoolOp' nodes are not implemented"
         with pytest.raises(NotImplementedError, match=msg):

pandas/tests/groupby/test_apply.py

@@ -678,7 +678,7 @@ def test_apply_groupby_datetimeindex():
     result = df.groupby("Name").sum()
 
     expected = DataFrame({"Name": ["A", "B", "C"], "Value": [10, 50, 90]})
-    expected.set_index("Name", inplace=True)
+    expected = expected.set_index("Name", copy=False)
 
     tm.assert_frame_equal(result, expected)
 

pandas/tests/groupby/test_function.py

@@ -98,7 +98,7 @@ def test_builtins_apply(keys, f):
 
     if f != sum:
         expected = gb.agg(fname).reset_index()
-        expected.set_index(keys, inplace=True, drop=False)
+        expected = expected.set_index(keys, copy=False, drop=False)
         tm.assert_frame_equal(result, expected, check_dtype=False)
 
     tm.assert_series_equal(getattr(result, fname)(), getattr(df, fname)())
@@ -454,7 +454,7 @@ def test_groupby_non_arithmetic_agg_types(dtype, method, data):
     df_out = DataFrame(exp)
 
     df_out["b"] = df_out.b.astype(out_type)
-    df_out.set_index("a", inplace=True)
+    df_out = df_out.set_index("a", copy=False)
 
     grpd = df.groupby("a")
     t = getattr(grpd, method)(*data["args"])

pandas/tests/indexes/multi/test_reshape.py

@@ -35,7 +35,7 @@ def test_insert(idx):
         idx.insert(0, ("foo2",))
 
     left = pd.DataFrame([["a", "b", 0], ["b", "d", 1]], columns=["1st", "2nd", "3rd"])
-    left.set_index(["1st", "2nd"], inplace=True)
+    left = left.set_index(["1st", "2nd"], copy=False)
     ts = left["3rd"].copy(deep=True)
 
     left.loc[("b", "x"), "3rd"] = 2
@@ -65,7 +65,7 @@ def test_insert(idx):
         ],
         columns=["1st", "2nd", "3rd"],
     )
-    right.set_index(["1st", "2nd"], inplace=True)
+    right = right.set_index(["1st", "2nd"], copy=False)
     # FIXME data types changes to float because
     # of intermediate nan insertion;
     tm.assert_frame_equal(left, right, check_dtype=False)

pandas/tests/indexing/multiindex/test_indexing_slow.py

@@ -60,15 +60,18 @@ def validate(mi, df, key):
         assert key[: i + 1] in mi.index
         right = df[mask].copy()
 
+        msg = "The 'inplace' keyword in DataFrame.set_index is deprecated"
         if i + 1 != len(key):  # partial key
             return_value = right.drop(cols[: i + 1], axis=1, inplace=True)
             assert return_value is None
-            return_value = right.set_index(cols[i + 1 : -1], inplace=True)
+            with tm.assert_produces_warning(FutureWarning, match=msg):
+                return_value = right.set_index(cols[i + 1 : -1], inplace=True)
             assert return_value is None
             tm.assert_frame_equal(mi.loc[key[: i + 1]], right)
 
         else:  # full key
-            return_value = right.set_index(cols[:-1], inplace=True)
+            with tm.assert_produces_warning(FutureWarning, match=msg):
+                return_value = right.set_index(cols[:-1], inplace=True)
             assert return_value is None
             if len(right) == 1:  # single hit
                 right = Series(

pandas/tests/indexing/multiindex/test_multiindex.py

@@ -131,7 +131,7 @@ def test_multiindex_complex(self):
                 "z": non_complex_data,
             }
         )
-        result.set_index(["x", "y"], inplace=True)
+        result = result.set_index(["x", "y"], copy=False)
         expected = DataFrame(
             {"z": non_complex_data},
             index=MultiIndex.from_arrays(

pandas/tests/io/pytables/test_append.py

@@ -137,7 +137,7 @@ def test_append_series(setup_path):
         mi["B"] = np.arange(len(mi))
         mi["C"] = "foo"
         mi.loc[3:5, "C"] = "bar"
-        mi.set_index(["C", "B"], inplace=True)
+        mi = mi.set_index(["C", "B"], copy=False)
         s = mi.stack()
         s.index = s.index.droplevel(2)
         store.append("mi", s)
@@ -326,7 +326,7 @@ def test_append_with_different_block_ordering(setup_path):
                 a = df.pop("A")
                 df["A"] = a
 
-            df.set_index("index", inplace=True)
+            df = df.set_index("index", copy=False)
 
             store.append("df", df)