marvin-python-toolbox
marvin-python-toolbox copied to clipboard
Implement a functional approach on the DASFE sdk
Currently, one is needed to know a convention of parameters names to set when implementing the components of the DASFE pattern. For instance, in the Preparator class, the user needs to know that the previous phase object is in a variable called self.initial_dataset and his job is to set the self.dataset.
It would be more intuitive if we apply a functional approach to the DASFE. In this case the Preparator execute method would be like:
def execute(initial_dataset, other_params):
...
## returns the prepared dataset
return any_name_var