REF: make Block.delete return a new Block (pandas-dev#45689)

Author: jbrockmendel

pandas/core/internals/blocks.py

@@ -380,23 +380,19 @@ def set_inplace(self, locs, values: ArrayLike) -> None:
         """
         self.values[locs] = values
 
-    def delete(self, loc) -> None:
+    def delete(self, loc) -> Block:
         """
-        Delete given loc(-s) from block in-place.
+        Return a new Block with the given loc(s) deleted.
         """
         # Argument 1 to "delete" has incompatible type "Union[ndarray[Any, Any],
         # ExtensionArray]"; expected "Union[_SupportsArray[dtype[Any]],
         # Sequence[_SupportsArray[dtype[Any]]], Sequence[Sequence
         # [_SupportsArray[dtype[Any]]]], Sequence[Sequence[Sequence[
         # _SupportsArray[dtype[Any]]]]], Sequence[Sequence[Sequence[Sequence[
         # _SupportsArray[dtype[Any]]]]]]]"  [arg-type]
-        self.values = np.delete(self.values, loc, 0)  # type: ignore[arg-type]
-        self.mgr_locs = self._mgr_locs.delete(loc)
-        try:
-            self._cache.clear()
-        except AttributeError:
-            # _cache not yet initialized
-            pass
+        values = np.delete(self.values, loc, 0)  # type: ignore[arg-type]
+        mgr_locs = self._mgr_locs.delete(loc)
+        return type(self)(values, placement=mgr_locs, ndim=self.ndim)
 
     @final
     def apply(self, func, **kwargs) -> list[Block]:
@@ -1504,18 +1500,11 @@ def fillna(
 
         return [self.make_block_same_class(values=new_values)]
 
-    def delete(self, loc) -> None:
-        """
-        Delete given loc(-s) from block in-place.
-        """
+    def delete(self, loc) -> Block:
         # This will be unnecessary if/when __array_function__ is implemented
-        self.values = self.values.delete(loc)
-        self.mgr_locs = self._mgr_locs.delete(loc)
-        try:
-            self._cache.clear()
-        except AttributeError:
-            # _cache not yet initialized
-            pass
+        values = self.values.delete(loc)
+        mgr_locs = self._mgr_locs.delete(loc)
+        return type(self)(values, placement=mgr_locs, ndim=self.ndim)
 
     @cache_readonly
     def array_values(self) -> ExtensionArray:

pandas/core/internals/managers.py

@@ -1133,8 +1133,12 @@ def value_getitem(placement):
                 if len(val_locs) == len(blk.mgr_locs):
                     removed_blknos.append(blkno_l)
                 else:
-                    blk.delete(blk_locs)
-                    self._blklocs[blk.mgr_locs.indexer] = np.arange(len(blk))
+                    nb = blk.delete(blk_locs)
+                    blocks_tup = (
+                        self.blocks[:blkno_l] + (nb,) + self.blocks[blkno_l + 1 :]
+                    )
+                    self.blocks = blocks_tup
+                    self._blklocs[nb.mgr_locs.indexer] = np.arange(len(nb))
 
         if len(removed_blknos):
             # Remove blocks & update blknos accordingly
@@ -1869,8 +1873,10 @@ def idelete(self, indexer) -> SingleBlockManager:
 
         Ensures that self.blocks doesn't become empty.
         """
-        self._block.delete(indexer)
+        nb = self._block.delete(indexer)
+        self.blocks = (nb,)
         self.axes[0] = self.axes[0].delete(indexer)
+        self._cache.clear()
         return self
 
     def fast_xs(self, loc):

pandas/tests/internals/test_internals.py

@@ -285,27 +285,36 @@ def test_copy(self):
 
     def test_delete(self):
         newb = self.fblock.copy()
-        newb.delete(0)
-        assert isinstance(newb.mgr_locs, BlockPlacement)
+        locs = newb.mgr_locs
+        nb = newb.delete(0)
+        assert newb.mgr_locs is locs
+
+        assert nb is not newb
+
         tm.assert_numpy_array_equal(
-            newb.mgr_locs.as_array, np.array([2, 4], dtype=np.intp)
+            nb.mgr_locs.as_array, np.array([2, 4], dtype=np.intp)
         )
-        assert (newb.values[0] == 1).all()
+        assert not (newb.values[0] == 1).all()
+        assert (nb.values[0] == 1).all()
 
         newb = self.fblock.copy()
-        newb.delete(1)
-        assert isinstance(newb.mgr_locs, BlockPlacement)
+        locs = newb.mgr_locs
+        nb = newb.delete(1)
+        assert newb.mgr_locs is locs
+
         tm.assert_numpy_array_equal(
-            newb.mgr_locs.as_array, np.array([0, 4], dtype=np.intp)
+            nb.mgr_locs.as_array, np.array([0, 4], dtype=np.intp)
         )
-        assert (newb.values[1] == 2).all()
+        assert not (newb.values[1] == 2).all()
+        assert (nb.values[1] == 2).all()
 
         newb = self.fblock.copy()
-        newb.delete(2)
+        locs = newb.mgr_locs
+        nb = newb.delete(2)
         tm.assert_numpy_array_equal(
-            newb.mgr_locs.as_array, np.array([0, 2], dtype=np.intp)
+            nb.mgr_locs.as_array, np.array([0, 2], dtype=np.intp)
         )
-        assert (newb.values[1] == 1).all()
+        assert (nb.values[1] == 1).all()
 
         newb = self.fblock.copy()
 
@@ -319,15 +328,15 @@ def test_delete_datetimelike(self):
         blk = df._mgr.blocks[0]
         assert isinstance(blk.values, TimedeltaArray)
 
-        blk.delete(1)
-        assert isinstance(blk.values, TimedeltaArray)
+        nb = blk.delete(1)
+        assert isinstance(nb.values, TimedeltaArray)
 
         df = DataFrame(arr.view("M8[ns]"))
         blk = df._mgr.blocks[0]
         assert isinstance(blk.values, DatetimeArray)
 
-        blk.delete([1, 3])
-        assert isinstance(blk.values, DatetimeArray)
+        nb = blk.delete([1, 3])
+        assert isinstance(nb.values, DatetimeArray)
 
     def test_split(self):
         # GH#37799