Mistake in type hinting Exponential Smoothing model - seasonality parameter

[rijkvandermeulen]

I believe the type hinting for the seasonal parameter in the Exponential Smoothing model is incorrect (ModelMode iso SeasonalityMode):

class ExponentialSmoothing(ForecastingModel):
    def __init__(
        self,
        trend: Optional[ModelMode] = ModelMode.ADDITIVE,
        damped: Optional[bool] = False,
        seasonal: Optional[ModelMode] = SeasonalityMode.ADDITIVE,
        seasonal_periods: Optional[int] = None,
        random_state: int = 0,
        **fit_kwargs,
    ):

I'll open a PR to make this quick fix. :)

BTW: have you ever considered adding mypy to your pipeline and pre-commit hooks to spot issues like this?

[gdevos010]

I can't speak for @hrzn, but I personally would love to see mypy added to the pre-commit hooks and pipeline. Mypy currently shows 119 errors which will need to be resolved first. Would you be able to help fix those?

[hrzn]

I can't speak for @hrzn, but I personally would love to see mypy added to the pre-commit hooks and pipeline. Mypy currently shows 119 errors which will need to be resolved first. Would you be able to help fix those?

I agree this would be nice 😍 We need a bit of work to fix those issues before can add the pre-commit hook indeed.

[rijkvandermeulen]

@hrzn @gdevos010 makes sense! I can have a look at the existing mypy errors somewhere in the coming weeks.

[madtoinou]

Fixed by #1186.