Tomas Pereira de Vasconcelos

Also tested this against a bunch of large internal and public projects and found no issues 🚀

@fkiraly What you are saying makes sense however couldn't that also be achieved by just calling `.fit()`? What I forgot to mention is that the part that I think is...

Sorry for the delay! Yes, this is definitely not a "bug", my bad! Also my comment about: > _It also looks like unnecessary overhead that does not do anything, since...

I also came across this issue (`ValueError: Need at least 3 dates to infer frequency`) when using multi-indexed DataFrames containing multiple time-series instances. The issue can actually be replicated in...

If you don't see any issues with the "simpler solution" approach, I can open a quick PR and get testing.

@yarnabrina range index is not an option as datetime information is sometimes needed to generate more informed forecasts. Think calendar features such as Black Friday, Easter, etc. I'll get working...

@yarnabrina the series' frequency is inferred to _retain_ the frequency information on the new slice of data. This fails for the edge case where the original data itself has fewer...

> I am not sure if I understood it well. Is the suggestion to run infer_freq per each different lowest level series? @yarnabrina no, the other way around. That is...

> In the most common cases, different lowest level series will have observations corresponding to same dates/datetimes, so number of distinct observations for index will not change Maybe, but unfortunately...

> I understand that of course :) My question is why do we need to know frequency at all. Does any of the algorithm use "frequency" itself in their forecasting...