Nils Braun

@arturdaraujo - thanks for looking into this! I would be very interested in how the numba implementation performs on the calculators with high computation cost (as marked by this flag)....

> Numba natively uses multiprocessing, it is set from the get-go. Correct! But as also stated in the links you shared it does not play well with the "normal" `multiprocessing`...

Hi @hn2 ! Thanks for filing the issue and sorry for the late response. Unfortunately, I can not reproduce the issue. Do you maybe have a minimal example? I am...

@dor132 , @morestart, @hn2 - has one of you a reproducible example with the data for me? Anything I can use for debugging? Thank you! Or does anyone maybe have...

Hi @YWU99u - which version of tsfresh, python and statsmodel are you using? Unfortunately, I can not reproduce this issue.

Hi @XudongWang97! Thanks for sharing those details. The respective file was changes approximately 2 years ago to not contain the "unbiased" parameter anymore, so I am confused why this shows...

Thanks for looking into this. If you have some time, could you check the package versions in both environments (e.g. with pip freeze) and see if there are significant differences?...

Hi @kiraminnie - can you check the version of tsfresh you are using? It looks like this is a (rather) old version.

Yes, tsfresh 0.18.0 is from 2021 (!), so it definitely will not work with any recent packages. I would recommend updating to the newest version (0.20.2). In general, the newest...

Hi @bulldog5046 - sorry for the late response. The problem in your case is, that your target is integer-valued, but has many different values. Our internal automatic ml target deduction...