API: MultiIndex.names|codes|levels returns tuples (#57042)

* MultiIndex.names|codes|levels returns tuples

* Fix typing

* Add whatsnew note

* Fix stacking

* Fix doctest, test

* Fix other test

* Remove example

doc/source/user_guide/groupby.rst

@@ -137,15 +137,6 @@ We could naturally group by either the ``A`` or ``B`` columns, or both:
 
    ``df.groupby('A')`` is just syntactic sugar for ``df.groupby(df['A'])``.
 
-If we also have a MultiIndex on columns ``A`` and ``B``, we can group by all
-the columns except the one we specify:
-
-.. ipython:: python
-
-   df2 = df.set_index(["A", "B"])
-   grouped = df2.groupby(level=df2.index.names.difference(["B"]))
-   grouped.sum()
-
 The above GroupBy will split the DataFrame on its index (rows). To split by columns, first do
 a transpose:
 

doc/source/whatsnew/v3.0.0.rst

@@ -86,6 +86,7 @@ See :ref:`install.dependencies` and :ref:`install.optional_dependencies` for mor
 Other API changes
 ^^^^^^^^^^^^^^^^^
 - 3rd party ``py.path`` objects are no longer explicitly supported in IO methods. Use :py:class:`pathlib.Path` objects instead (:issue:`57091`)
+- :attr:`MultiIndex.codes`, :attr:`MultiIndex.levels`, and :attr:`MultiIndex.names` now returns a ``tuple`` instead of a ``FrozenList`` (:issue:`53531`)
 -
 
 .. ---------------------------------------------------------------------------

pandas/_libs/index.pyi

@@ -2,7 +2,10 @@ import numpy as np
 
 from pandas._typing import npt
 
-from pandas import MultiIndex
+from pandas import (
+    Index,
+    MultiIndex,
+)
 from pandas.core.arrays import ExtensionArray
 
 multiindex_nulls_shift: int
@@ -70,13 +73,13 @@ class MaskedUInt8Engine(MaskedIndexEngine): ...
 class MaskedBoolEngine(MaskedUInt8Engine): ...
 
 class BaseMultiIndexCodesEngine:
-    levels: list[np.ndarray]
+    levels: tuple[np.ndarray]
     offsets: np.ndarray  # ndarray[uint64_t, ndim=1]
 
     def __init__(
         self,
-        levels: list[np.ndarray],  # all entries hashable
-        labels: list[np.ndarray],  # all entries integer-dtyped
+        levels: tuple[Index, ...],  # all entries hashable
+        labels: tuple[np.ndarray],  # all entries integer-dtyped
         offsets: np.ndarray,  # np.ndarray[np.uint64, ndim=1]
     ) -> None: ...
     def get_indexer(self, target: npt.NDArray[np.object_]) -> npt.NDArray[np.intp]: ...

pandas/core/groupby/groupby.py

@@ -5698,7 +5698,7 @@ def _insert_quantile_level(idx: Index, qs: npt.NDArray[np.float64]) -> MultiInde
         idx = cast(MultiIndex, idx)
         levels = list(idx.levels) + [lev]
         codes = [np.repeat(x, nqs) for x in idx.codes] + [np.tile(lev_codes, len(idx))]
-        mi = MultiIndex(levels=levels, codes=codes, names=idx.names + [None])
+        mi = MultiIndex(levels=levels, codes=codes, names=list(idx.names) + [None])
     else:
         nidx = len(idx)
         idx_codes = coerce_indexer_dtype(np.arange(nidx), idx)

pandas/core/indexes/base.py

@@ -179,7 +179,6 @@
     disallow_ndim_indexing,
     is_valid_positional_slice,
 )
-from pandas.core.indexes.frozen import FrozenList
 from pandas.core.missing import clean_reindex_fill_method
 from pandas.core.ops import get_op_result_name
 from pandas.core.ops.invalid import make_invalid_op
@@ -1767,8 +1766,8 @@ def _get_default_index_names(
 
         return names
 
-    def _get_names(self) -> FrozenList:
-        return FrozenList((self.name,))
+    def _get_names(self) -> tuple[Hashable | None, ...]:
+        return (self.name,)
 
     def _set_names(self, values, *, level=None) -> None:
         """
@@ -1866,7 +1865,7 @@ def set_names(self, names, *, level=None, inplace: bool = False) -> Self | None:
                     ('python', 2019),
                     ( 'cobra', 2018),
                     ( 'cobra', 2019)],
-                   names=['species', 'year'])
+                   names=('species', 'year'))
 
         When renaming levels with a dict, levels can not be passed.
 
@@ -1875,7 +1874,7 @@ def set_names(self, names, *, level=None, inplace: bool = False) -> Self | None:
                     ('python', 2019),
                     ( 'cobra', 2018),
                     ( 'cobra', 2019)],
-                   names=['snake', 'year'])
+                   names=('snake', 'year'))
         """
         if level is not None and not isinstance(self, ABCMultiIndex):
             raise ValueError("Level must be None for non-MultiIndex")
@@ -1959,19 +1958,19 @@ def rename(self, name, inplace: bool = False) -> Self | None:
 
         >>> idx = pd.MultiIndex.from_product([['python', 'cobra'],
         ...                                   [2018, 2019]],
-        ...                                  names=['kind', 'year'])
+        ...                                  names=('kind', 'year'))
         >>> idx
         MultiIndex([('python', 2018),
                     ('python', 2019),
                     ( 'cobra', 2018),
                     ( 'cobra', 2019)],
-                   names=['kind', 'year'])
+                   names=('kind', 'year'))
         >>> idx.rename(['species', 'year'])
         MultiIndex([('python', 2018),
                     ('python', 2019),
                     ( 'cobra', 2018),
                     ( 'cobra', 2019)],
-                   names=['species', 'year'])
+                   names=('species', 'year'))
         >>> idx.rename('species')
         Traceback (most recent call last):
         TypeError: Must pass list-like as `names`.
@@ -2135,22 +2134,22 @@ def droplevel(self, level: IndexLabel = 0):
         >>> mi
         MultiIndex([(1, 3, 5),
                     (2, 4, 6)],
-                   names=['x', 'y', 'z'])
+                   names=('x', 'y', 'z'))
 
         >>> mi.droplevel()
         MultiIndex([(3, 5),
                     (4, 6)],
-                   names=['y', 'z'])
+                   names=('y', 'z'))
 
         >>> mi.droplevel(2)
         MultiIndex([(1, 3),
                     (2, 4)],
-                   names=['x', 'y'])
+                   names=('x', 'y'))
 
         >>> mi.droplevel('z')
         MultiIndex([(1, 3),
                     (2, 4)],
-                   names=['x', 'y'])
+                   names=('x', 'y'))
 
         >>> mi.droplevel(['x', 'y'])
         Index([5, 6], dtype='int64', name='z')
@@ -4865,7 +4864,9 @@ def _join_level(
         """
         from pandas.core.indexes.multi import MultiIndex
 
-        def _get_leaf_sorter(labels: list[np.ndarray]) -> npt.NDArray[np.intp]:
+        def _get_leaf_sorter(
+            labels: tuple[np.ndarray, ...] | list[np.ndarray]
+        ) -> npt.NDArray[np.intp]:
             """
             Returns sorter for the inner most level while preserving the
             order of higher levels.
@@ -6627,7 +6628,7 @@ def isin(self, values, level=None) -> npt.NDArray[np.bool_]:
         MultiIndex([(1,   'red'),
                     (2,  'blue'),
                     (3, 'green')],
-                   names=['number', 'color'])
+                   names=('number', 'color'))
 
         Check whether the strings in the 'color' level of the MultiIndex
         are in a list of colors.
@@ -7608,7 +7609,7 @@ def ensure_index_from_sequences(sequences, names=None) -> Index:
     >>> ensure_index_from_sequences([["a", "a"], ["a", "b"]], names=["L1", "L2"])
     MultiIndex([('a', 'a'),
                 ('a', 'b')],
-               names=['L1', 'L2'])
+               names=('L1', 'L2'))
 
     See Also
     --------