Nils Braun

What do you mean with the input needs df.compute()? If you use dask as input to the rolling function, you can also hand over a non-evaluated dask data frame (not...

Oh - I am very sorry. I somehow assumed #731 is already finished and implemented but have just remembered that this is not the case. Hm, well then you are...

Great, @theinem ! Three really interesting features. As @MaxBenChrist mentioned, could you add some unittests for your new feature calculators? We try to test every calculator :-) Also, you new...

Thanks for the info, @theinem. As mentioned in https://github.com/blue-yonder/tsfresh/pull/825, this is not a problem for us!

That is a very good suggestion! Would you like to do a PR? It is mostly used in the `MultiprocessingDistributor` and in the `calculate_relevance_table` (and probably a bunch of docstrings).

> I can try it. However it might take some weeks because I'm quite busy right now. That would be awesome! If this is not fast enough for you, I...

You could have a look into https://github.com/blue-yonder/tsfresh/pull/852/files to get some starter :-)

These are all very valid points you raise. I think it is probably the pivoting. I do not know exactly how many ids you have, but I guess the resulting...

Just a short comment: it seems like the pivoting can be improved by a large factor. Still sorting out the details, but it was totally worth looking into this. Thanks...

#705 removed a bit of computational effort (~half) from this part - but the issue is still not solved. Additional open things to do: * we sort the index of...