Add comments on when to use which TypeVar
#8212
Conversation
From the chars of pydata#8208
xarray/core/types.py
Outdated
| T_Xarray = TypeVar("T_Xarray", "DataArray", "Dataset") | ||
|
|
||
| # Use `T_DataWithCoords` for: | ||
| # - functions that return either `DataArray` or `Dataset`, and |
There was a problem hiding this comment.
I think for this usecase you should use T_DataArrayOrSet.
You can use this one, when the code that uses the input only uses methods that are available in DataWithCoords.
There was a problem hiding this comment.
Yes OK. I had tried to consolidate in #8213.
Is the rule for what to use really "the code that uses the input only uses methods that are available in DataWithCoords" though? And then if we add an extra method, we change everything to use T_DataArrayOrSet, given they're not always compatible?
(Maybe this isn't an easy answer, and I appreciate your review)
There was a problem hiding this comment.
I tried another pass. If it's not quite there yet, feel free to just rewrite and/or push off for the moment
max-sixty
force-pushed
the
types-comments
branch
from
f84120c to
ac49c57
Compare
| # Maybe we rename this to T_Data or something less Fortran-y? | ||
| T_Xarray = TypeVar("T_Xarray", "DataArray", "Dataset") | ||
| # For working directly with `DataWithCoords`. It will only allow using methods defined | ||
| # on `DataWithCoords`. |
There was a problem hiding this comment.
I think it's already good.
Personally I have found it better to use than T_DataArrayOrSet because it can better resolve the unions. But ofc it only works under the limitations above.
|
Thanks - I think this is very helpful! |
From the chars of #8208