[question] can hydra-sklearn-pipeline support function parameters for FunctionTransformer of sklearn

[muliliao]

my config is:
target: hydra_sklearn_pipeline.make_pipeline
recursive: False

steps_config: # use yaml list syntax to preserve to order
- custom_transformer:
target: sklearn.preprocessing.FunctionTransformer
func: np.log1p

then it will report error like this:
File "D:\ProgramFiles\Python\Python37\lib\site-packages\sklearn\preprocessing_function_transformer.py", line 171, in _transform
return func(X, **(kw_args if kw_args else {}))
TypeError: 'str' object is not callable

it's because it treat the function "np.log1p" as string.
so is there any approach to support function parameters?
thanks.

[elisim]

Thank you for reading the post, and submitting this issue :)

Maybe you can use _partial_ initialization? like the example here.
Alternatively, consider implementing a new transformer class (e.g. MyLogTransformer) and initialize it with Hydra.