PERF: Only clear cached .levels when setting MultiIndex.names (#59578)

* PERF: Only clear cached .levels when setting MultiIndex.names

* whatsnew number

Author: mroeschke

doc/source/whatsnew/v3.0.0.rst

@@ -503,6 +503,7 @@ Performance improvements
 - :meth:`Series.str.extract` returns a :class:`RangeIndex` columns instead of an :class:`Index` column when possible (:issue:`57542`)
 - :meth:`Series.str.partition` with :class:`ArrowDtype` returns a :class:`RangeIndex` columns instead of an :class:`Index` column when possible (:issue:`57768`)
 - Performance improvement in :class:`DataFrame` when ``data`` is a ``dict`` and ``columns`` is specified (:issue:`24368`)
+- Performance improvement in :class:`MultiIndex` when setting :attr:`MultiIndex.names` doesn't invalidate all cached operations (:issue:`59578`)
 - Performance improvement in :meth:`DataFrame.join` for sorted but non-unique indexes (:issue:`56941`)
 - Performance improvement in :meth:`DataFrame.join` when left and/or right are non-unique and ``how`` is ``"left"``, ``"right"``, or ``"inner"`` (:issue:`56817`)
 - Performance improvement in :meth:`DataFrame.join` with ``how="left"`` or ``how="right"`` and ``sort=True`` (:issue:`56919`)

pandas/core/indexes/multi.py

@@ -799,7 +799,7 @@ def dtypes(self) -> Series:
         """
         from pandas import Series
 
-        names = com.fill_missing_names([level.name for level in self.levels])
+        names = com.fill_missing_names(self.names)
         return Series([level.dtype for level in self.levels], index=Index(names))
 
     def __len__(self) -> int:
@@ -1572,7 +1572,7 @@ def _format_multi(
     def _get_names(self) -> FrozenList:
         return FrozenList(self._names)
 
-    def _set_names(self, names, *, level=None, validate: bool = True) -> None:
+    def _set_names(self, names, *, level=None) -> None:
         """
         Set new names on index. Each name has to be a hashable type.
 
@@ -1583,8 +1583,6 @@ def _set_names(self, names, *, level=None, validate: bool = True) -> None:
         level : int, level name, or sequence of int/level names (default None)
             If the index is a MultiIndex (hierarchical), level(s) to set (None
             for all levels).  Otherwise level must be None
-        validate : bool, default True
-            validate that the names match level lengths
 
         Raises
         ------
@@ -1603,13 +1601,12 @@ def _set_names(self, names, *, level=None, validate: bool = True) -> None:
             raise ValueError("Names should be list-like for a MultiIndex")
         names = list(names)
 
-        if validate:
-            if level is not None and len(names) != len(level):
-                raise ValueError("Length of names must match length of level.")
-            if level is None and len(names) != self.nlevels:
-                raise ValueError(
-                    "Length of names must match number of levels in MultiIndex."
-                )
+        if level is not None and len(names) != len(level):
+            raise ValueError("Length of names must match length of level.")
+        if level is None and len(names) != self.nlevels:
+            raise ValueError(
+                "Length of names must match number of levels in MultiIndex."
+            )
 
         if level is None:
             level = range(self.nlevels)
@@ -1627,8 +1624,9 @@ def _set_names(self, names, *, level=None, validate: bool = True) -> None:
                     )
             self._names[lev] = name
 
-        # If .levels has been accessed, the names in our cache will be stale.
-        self._reset_cache()
+        # If .levels has been accessed, the .name of each level in our cache
+        # will be stale.
+        self._reset_cache("levels")
 
     names = property(
         fset=_set_names,

pandas/tests/indexing/multiindex/test_chaining_and_caching.py

@@ -5,6 +5,7 @@
 from pandas import (
     DataFrame,
     MultiIndex,
+    RangeIndex,
     Series,
 )
 import pandas._testing as tm
@@ -68,3 +69,19 @@ def test_indexer_caching(monkeypatch):
         s[s == 0] = 1
     expected = Series(np.ones(size_cutoff), index=index)
     tm.assert_series_equal(s, expected)
+
+
+def test_set_names_only_clears_level_cache():
+    mi = MultiIndex.from_arrays([range(4), range(4)], names=["a", "b"])
+    mi.dtypes
+    mi.is_monotonic_increasing
+    mi._engine
+    mi.levels
+    old_cache_keys = sorted(mi._cache.keys())
+    assert old_cache_keys == ["_engine", "dtypes", "is_monotonic_increasing", "levels"]
+    mi.names = ["A", "B"]
+    new_cache_keys = sorted(mi._cache.keys())
+    assert new_cache_keys == ["_engine", "dtypes", "is_monotonic_increasing"]
+    new_levels = mi.levels
+    tm.assert_index_equal(new_levels[0], RangeIndex(4, name="A"))
+    tm.assert_index_equal(new_levels[1], RangeIndex(4, name="B"))