Description
Is your feature request related to a problem?
I would like to be able to use a subclass of UserDict with custom __getitem__ implementation as a mapper for my dataframe columns. My use case is that I want to have a mapper with default remapping rules when the key is not present.
Unfortunately, as the implementation checks for membership of the key before actually calling __getitem__, I am only able to do the behaviour by wrapping the UserDict with a lambda.
Current implementation of get_rename_function in pandas:
def get_rename_function(mapper):
"""
Returns a function that will map names/labels, dependent if mapper
is a dict, Series or just a function.
"""
if isinstance(mapper, (abc.Mapping, ABCSeries)):
def f(x):
if x in mapper:
return mapper[x]
else:
return x
else:
f = mapper
return f
My proposition is the following:
def get_rename_function(mapper):
"""
Returns a function that will map names/labels, dependent if mapper
is a dict, Series or just a function.
"""
if isinstance(mapper, (abc.Mapping, ABCSeries)):
def f(k):
return mapper.get(k, k)
else:
f = mapper
return f
which functionally should result in the same behaviour.
The advantage for me is that __getitem__ is called and I can process k to return my default mapping.
I don't know however if this would have a significant performance impact. If so, alternatively I'd propose the following:
def get_rename_function(mapper):
"""
Returns a function that will map names/labels, dependent if mapper
is a dict, Series or just a function.
"""
if not callable(mapper) and isinstance(mapper, (abc.Mapping, ABCSeries)):
def f(x):
if x in mapper:
return mapper[x]
else:
return x
else:
f = mapper
return f
Then I could implement the __call__ method. The downside however is that this could be a breaking change for some people that pass a subclass of a Mapping as well.
What do you think?