feature/SetColumnDtype #53
Conversation
|
|
||
| def __init__(self, columns, dtype): | ||
|
|
||
| super().__init__(columns, copy=True) |
There was a problem hiding this comment.
This needs a string input check
There was a problem hiding this comment.
dtype can be a string or a dtype object. Have added clarity to docstring in 36a9dde
|
|
||
| return X | ||
|
|
||
| def __validate_dtype(self, dtype: str): |
There was a problem hiding this comment.
In my opinion it would be be nicer to do this validation without having to use a try/except clause - I think that the errors from pd.api.types.pandas_dtype are maybe expressive enough. Not sure that having the freedom to add self.classname is worth adding a try except. (feel free to ignore this one as it's quite picky)
There was a problem hiding this comment.
The self.classname in the error can be really handy when you have dozens of steps in a pipeline. Always nice to know where things are breaking. I agree it is slightly overwrought but since it has been done now I'm inclined to leave it in.
tests/misc/test_SetColumnDtype.py
Outdated
| from tubular.misc import SetColumnDtype | ||
|
|
||
|
|
||
| class TestSetColumnDtype(object): |
There was a problem hiding this comment.
The tests are not in the usual TestInit, TestTransform etc format
There was a problem hiding this comment.
reorganised in cf59b9c
| @@ -0,0 +1,447 @@ | |||
| { | |||
There was a problem hiding this comment.
I think that this could do with more markdown - not that clear what is going on in the examples
There was a problem hiding this comment.
added in 2874ad5
… testSetColumnDtypeTransform as usual. Added a few more tests.
added SetColumnDtype transformer, test module and example
Transformer to set column dtypes