get_params option?

[FABoulanger]

Hello,

do you know how to access the parameters of the models in order to have them in a dictionary?
In scikit learn there is model.get_params, is there an equivalent?
thank you

[HansikaPH]

Thanks for this suggestion. At the moment, there is no function accessible to the users to do this. One option would be to directly access the model parameters as in m.model.season_params.items() for the seasonality related parameters etc. But, having such a function to do that would be certainly useful. If you would like to give it a try in implementing it, we would be happy to help and collaborate.

[ourownstory]

@FABoulanger This is actually a great suggestion and makes for a great first issue. I added it as issue #252 in case you or someone would like to contribute it.

Further, for hyperparameters, you can retrieve them by calling asdict from dataclasses, see on stackoverflow, Option 5, as they are all dataclasses (code here).
e.g. trend_hyperparams = dataclasses.asdict(m.config_trend)