Series.replace fails on categoricals with list

[bnaul]

Looks like this was introduced in #27026: not sure of the details of what was incorrect in the old version but I do know that this at least didn't error:

cat = pd.CategoricalIndex(['a', 'b']).to_series()
cat.replace('a', 'A')  # works
cat.replace(['a'], 'A')
~/model/.venv/lib/python3.7/site-packages/pandas/core/arrays/categorical.py in replace(self, to_replace, value, inplace)
   2440         inplace = validate_bool_kwarg(inplace, "inplace")
   2441         cat = self if inplace else self.copy()
-> 2442         if to_replace in cat.categories:
   2443             if isna(value):
   2444                 cat.remove_categories(to_replace, inplace=True)

~/model/.venv/lib/python3.7/site-packages/pandas/core/indexes/base.py in __contains__(self, key)
   3898     @Appender(_index_shared_docs["contains"] % _index_doc_kwargs)
   3899     def __contains__(self, key) -> bool:
-> 3900         hash(key)
   3901         try:
   3902             return key in self._engine

TypeError: unhashable type: 'list'

Seems like most of the methods used in this implementation already handle list-like inputs so it should be a pretty easy fix..?

Output of pd.show_versions()

pandas : 1.0.0

[bnaul]

cc @JustinZhengBC just in case you have an easy fix off the top of your head 🙂

[dsaxton]

Also interesting that tuples of values get ignored instead of raising an error:

In [11]: s                                                                                                                                                            
Out[11]: 
0    0
1    1
2    2
3    3
4    4
dtype: category
Categories (5, int64): [0, 1, 2, 3, 4]

In [12]: s.replace((0,), 999)                                                                                                                                         
Out[12]: 
0    0
1    1
2    2
3    3
4    4
dtype: category
Categories (5, int64): [0, 1, 2, 3, 4]

In [13]: s.astype(int).replace((0,), 999)                                                                                                                             
Out[13]: 
0    999
1      1
2      2
3      3
4      4
dtype: int64

[JustinZhengBC]

At a glance it looks like we can just check if to_replace and value are list-like and do recursive calls accordingly. Tuples probably are doing nothing because it's checking if the tuple itself is a category, rather than the values within it. I'll see if I can get a fix in tonight