Small fixes for Category

[koalp]

This PR solves issue with the Category type.

Makes numpy_pandas_coercible return a Series of bool

It solves #1073

More informative error message when coercing category

#931

The error message would now be

pandera.errors.SchemaError: 
Error while coercing 'category' to type CategoricalDtype(categories=[1, 2], ordered=False): 
Could not coerce <class 'pandas.core.series.Series'> data_container into type category:

instead of

pandera.errors.SchemaError: 
Error while coercing 'category' to type category:
Could not coerce <class 'pandas.core.series.Series'> data_container into type category:

• Should a custom f-string be used instead of repr(self.dtype) ? As repr is multiline whene there is too much categories, it may be better to use a oneline string.

closes

closes #1073
closes #931

[cosmicBboy]

hi @koalp thanks for the PR! This is blocked by a big pandera internals re-write: #913, I'll ping you here once that's merged, as the changes you've added will need to be ported over to the new internals.