Inconsistent approximation default for Auto ARIMA

[alech97]

statsforecast/statsforecast/models.py

Line 238 in b4b59c9

approximation: bool = False) -> np.ndarray:

approximation is set to False by default in auto_arima(), but auto_arima_f()'s default handles this logic similarly to R's auto.arima() already:

statsforecast/statsforecast/arima.py

Line 1438 in b4b59c9

approximation = len(x) > 150 or period > 12

[AzulGarza]

Hi @alech97! We decided to set approximation = False to be comparable with pmdarima. That package (if I remember correctly) doesn't have the option to use approximation. Even without approximation, our implementation is a lot faster. 😄
The use of approximation=True reduces the time even more, but may also reduce the accuracy.

[mergenthaler]

@all-contributors please add @alech97 for question

[allcontributors[bot]]

@mergenthaler

I've put up a pull request to add @alech97! 🎉

[MarcoGorelli]

The use of approximation=False reduces the time even more, but may also reduce the accuracy.

I presume you meant "The use of approximation=True"?

[AzulGarza]

Yes, @MarcoGorelli that is correct! Thanks for catching that :)