fix pickling

Author: jorisvandenbossche

pandas/compat/pickle_compat.py

@@ -22,6 +22,7 @@
     PeriodArray,
     TimedeltaArray,
 )
+from pandas.core.generic import NDFrame
 from pandas.core.internals import BlockManager
 
 if TYPE_CHECKING:
@@ -90,6 +91,10 @@ def load_reduce(self) -> None:
                 cls = args[0]
                 stack[-1] = NDArrayBacked.__new__(*args)
                 return
+            elif args and issubclass(args[0], NDFrame):
+                cls = args[0]
+                stack[-1] = cls.__new__(cls)
+                return
             raise
 
     dispatch[pickle.REDUCE[0]] = load_reduce  # type: ignore[assignment]

pandas/core/frame.py

@@ -47,7 +47,6 @@
     properties,
 )
 from pandas._libs.hashtable import duplicated
-from pandas._libs.internals import SetitemMixin
 from pandas._libs.lib import is_range_indexer
 from pandas.compat import PYPY
 from pandas.compat._constants import REF_COUNT
@@ -511,7 +510,7 @@
 
 
 @set_module("pandas")
-class DataFrame(NDFrame, OpsMixin, SetitemMixin):
+class DataFrame(NDFrame, OpsMixin):
     """
     Two-dimensional, size-mutable, potentially heterogeneous tabular data.
 

pandas/core/generic.py

@@ -27,6 +27,7 @@
 from pandas._config import config
 
 from pandas._libs import lib
+from pandas._libs.internals import SetitemMixin
 from pandas._libs.lib import is_range_indexer
 from pandas._libs.tslibs import (
     Period,
@@ -222,7 +223,7 @@
 }
 
 
-class NDFrame(PandasObject, indexing.IndexingMixin):
+class NDFrame(PandasObject, indexing.IndexingMixin, SetitemMixin):
     """
     N-dimensional analogue of DataFrame. Store multi-dimensional in a
     size-mutable, labeled data structure
@@ -252,6 +253,11 @@ class NDFrame(PandasObject, indexing.IndexingMixin):
     # ----------------------------------------------------------------------
     # Constructors
 
+    # override those to avoid inheriting from SetitemMixin (cython generates
+    # them by default)
+    __new__ = object.__new__
+    __reduce__ = object.__reduce__
+
     def __init__(self, data: Manager) -> None:
         object.__setattr__(self, "_mgr", data)
         object.__setattr__(self, "_attrs", {})
@@ -305,7 +311,7 @@ def _from_mgr(cls, mgr: Manager, axes: list[Index]) -> Self:
         The axes must match mgr.axes, but are required for future-proofing
         in the event that axes are refactored out of the Manager objects.
         """
-        obj = object.__new__(cls)
+        obj = cls.__new__(cls)
         NDFrame.__init__(obj, mgr)
         return obj
 

pandas/core/series.py

@@ -32,7 +32,6 @@
     properties,
     reshape,
 )
-from pandas._libs.internals import SetitemMixin
 from pandas._libs.lib import is_range_indexer
 from pandas.compat import PYPY
 from pandas.compat._constants import REF_COUNT
@@ -234,7 +233,7 @@
 # class "NDFrame")
 # definition in base class "NDFrame"
 @set_module("pandas")
-class Series(base.IndexOpsMixin, NDFrame, SetitemMixin):  # type: ignore[misc]
+class Series(base.IndexOpsMixin, NDFrame):  # type: ignore[misc]
     """
     One-dimensional ndarray with axis labels (including time series).