CLN: remove fill_tuple kludge #33310
Conversation
pandas/core/internals/managers.py
Outdated
| @@ -1396,7 +1394,7 @@ def _slice_take_blocks_ax0(self, slice_or_indexer, fill_tuple=None): | |||
| blklocs[mgr_locs.indexer], | |||
| axis=0, | |||
| new_mgr_locs=mgr_locs, | |||
| fill_tuple=None, | |||
| fill_value=lib.no_default, | |||
There was a problem hiding this comment.
should this be removed? isn't it an anti-pattern to explicitly pass this.
| @@ -1339,16 +1339,15 @@ def _slice_take_blocks_ax0(self, slice_or_indexer, fill_tuple=None): | |||
| if sl_type in ("slice", "mask"): | |||
| return [blk.getitem_block(slobj, new_mgr_locs=slice(0, sllen))] | |||
| elif not allow_fill or self.ndim == 1: | |||
| if allow_fill and fill_tuple[0] is None: | |||
| if allow_fill and fill_value is None: | |||
There was a problem hiding this comment.
is allow_fill now redundant here?
There was a problem hiding this comment.
allow_fill is tricky to remove (but this is a start to it)
There was a problem hiding this comment.
The allow_fill is from checking fill_value is not lib.default. So fill_value is None is still a different check I think.
There was a problem hiding this comment.
what @jorisvandenbossche said
There was a problem hiding this comment.
if fill_value is None, isn't allow_fill, by definition, True. so allow_fill is redundant here and if fill_value is None: would be the same? nbd though, i must be missing something. will look again. We no longer need the guard, that was previously required with the fill_tuple __getitem__ call.
No description provided.