SQUASGED [ArrayManager] Add SingleArrayManager to back a Series

.github/workflows/ci.yml

@@ -163,6 +163,9 @@ jobs:
         pytest pandas/tests/resample/ --array-manager
         pytest pandas/tests/reshape/merge --array-manager
 
+        pytest pandas/tests/series/methods --array-manager
+        pytest pandas/tests/series/test_* --array-manager
+
         # indexing subset (temporary since other tests don't pass yet)
         pytest pandas/tests/frame/indexing/test_indexing.py::TestDataFrameIndexing::test_setitem_boolean --array-manager
         pytest pandas/tests/frame/indexing/test_where.py --array-manager

pandas/_libs/reduction.pyx

@@ -59,16 +59,20 @@ cdef class _BaseGrouper:
             cached_typ = self.typ(
                 vslider.buf, dtype=vslider.buf.dtype, index=cached_ityp, name=self.name
             )
+            self.has_block = hasattr(cached_typ._mgr, "_block")
         else:
             # See the comment in indexes/base.py about _index_data.
             # We need this for EA-backed indexes that have a reference
             # to a 1-d ndarray like datetime / timedelta / period.
             object.__setattr__(cached_ityp, '_index_data', islider.buf)
             cached_ityp._engine.clear_mapping()
             cached_ityp._cache.clear()  # e.g. inferred_freq must go
-            object.__setattr__(cached_typ._mgr._block, 'values', vslider.buf)
-            object.__setattr__(cached_typ._mgr._block, 'mgr_locs',
-                               slice(len(vslider.buf)))
+            if self.has_block:
+                object.__setattr__(cached_typ._mgr._block, 'values', vslider.buf)
+                object.__setattr__(cached_typ._mgr._block, 'mgr_locs',
+                                   slice(len(vslider.buf)))
+            else:
+                cached_typ._mgr.arrays[0] = vslider.buf
             object.__setattr__(cached_typ, '_index', cached_ityp)
             object.__setattr__(cached_typ, 'name', self.name)
 
@@ -108,6 +112,7 @@ cdef class SeriesBinGrouper(_BaseGrouper):
     cdef public:
         ndarray arr, index, dummy_arr, dummy_index
         object values, f, bins, typ, ityp, name
+        bint has_block
 
     def __init__(self, object series, object f, object bins):
 
@@ -201,6 +206,7 @@ cdef class SeriesGrouper(_BaseGrouper):
     cdef public:
         ndarray arr, index, dummy_arr, dummy_index
         object f, labels, values, typ, ityp, name
+        bint has_block
 
     def __init__(self, object series, object f, object labels,
                  Py_ssize_t ngroups):

pandas/_typing.py

@@ -58,6 +58,8 @@
     from pandas.core.internals import (
         ArrayManager,
         BlockManager,
+        SingleArrayManager,
+        SingleBlockManager,
     )
     from pandas.core.resample import Resampler
     from pandas.core.series import Series
@@ -184,3 +186,4 @@
 
 # internals
 Manager = Union["ArrayManager", "BlockManager"]
+SingleManager = Union["SingleArrayManager", "SingleBlockManager"]

pandas/core/construction.py

@@ -481,7 +481,6 @@ def sanitize_array(
     DataFrame constructor, as the dtype keyword there may be interpreted as only
     applying to a subset of columns, see GH#24435.
     """
-
     if isinstance(data, ma.MaskedArray):
         data = sanitize_masked_array(data)
 
@@ -555,6 +554,7 @@ def sanitize_array(
             inferred = lib.infer_dtype(subarr, skipna=False)
             if inferred in {"interval", "period"}:
                 subarr = array(subarr)
+                subarr = extract_array(subarr, extract_numpy=True)
 
     return subarr
 

pandas/core/generic.py

@@ -138,6 +138,7 @@
 from pandas.core.internals import (
     ArrayManager,
     BlockManager,
+    SingleArrayManager,
 )
 from pandas.core.internals.construction import mgr_to_mgr
 from pandas.core.missing import find_valid_index
@@ -5563,7 +5564,7 @@ def _protect_consolidate(self, f):
         Consolidate _mgr -- if the blocks have changed, then clear the
         cache
         """
-        if isinstance(self._mgr, ArrayManager):
+        if isinstance(self._mgr, (ArrayManager, SingleArrayManager)):
             return f()
         blocks_before = len(self._mgr.blocks)
         result = f()

pandas/core/groupby/generic.py

@@ -1152,16 +1152,18 @@ def py_fallback(values: ArrayLike) -> ArrayLike:
             result = result._consolidate()
             assert isinstance(result, (Series, DataFrame))  # for mypy
             mgr = result._mgr
-            assert isinstance(mgr, BlockManager)
-
-            # unwrap DataFrame to get array
-            if len(mgr.blocks) != 1:
-                # We've split an object block! Everything we've assumed
-                # about a single block input returning a single block output
-                # is a lie. See eg GH-39329
-                return mgr.as_array()
+            if isinstance(mgr, BlockManager):
+                # unwrap DataFrame to get array
+                if len(mgr.blocks) != 1:
+                    # We've split an object block! Everything we've assumed
+                    # about a single block input returning a single block output
+                    # is a lie. See eg GH-39329
+                    return mgr.as_array()
+                else:
+                    result = mgr.blocks[0].values
+                    return result
             else:
-                result = mgr.blocks[0].values
+                result = mgr.arrays[0]
                 return result
 
         def array_func(values: ArrayLike) -> ArrayLike:

pandas/core/indexing.py

@@ -1590,7 +1590,11 @@ def _setitem_with_indexer(self, indexer, value, name="iloc"):
 
         # if there is only one block/type, still have to take split path
         # unless the block is one-dimensional or it can hold the value
-        if not take_split_path and self.obj._mgr.blocks and self.ndim > 1:
+        if (
+            not take_split_path
+            and getattr(self.obj._mgr, "blocks", False)
+            and self.ndim > 1
+        ):
             # in case of dict, keys are indices
             val = list(value.values()) if isinstance(value, dict) else value
             blk = self.obj._mgr.blocks[0]

pandas/core/internals/__init__.py

@@ -1,5 +1,11 @@
-from pandas.core.internals.array_manager import ArrayManager
-from pandas.core.internals.base import DataManager
+from pandas.core.internals.array_manager import (
+    ArrayManager,
+    SingleArrayManager,
+)
+from pandas.core.internals.base import (
+    DataManager,
+    SingleDataManager,
+)
 from pandas.core.internals.blocks import (  # io.pytables, io.packers
     Block,
     CategoricalBlock,
@@ -34,7 +40,9 @@
     "DataManager",
     "ArrayManager",
     "BlockManager",
+    "SingleDataManager",
     "SingleBlockManager",
+    "SingleArrayManager",
     "concatenate_managers",
     # those two are preserved here for downstream compatibility (GH-33892)
     "create_block_manager_from_arrays",