Antoine Guillaume
Antoine Guillaume
I'm OK with the use of the `parallel_config` option and let user warp estimators if they want to change from the default. But if we want to default to threads,...
So for options we have : - Add a parameter to specify backend, which defaults to threads for all estimators with parallel capabilities - Add an environment variable to specify...
Expected results seem to differ for some test configurations, not sure why yet
I'm currently clueless about why these fails, I cannot seem to reproduce the issue on my Ubuntu either ... If anyone has any leads, I'd be grateful, otherwise i'll come...
I think I found the issue. I was using a NumPy random Generator numba inside a parallel loop, but the use of such a generator is not thread-safe according to...
What were the remaining things to do here ? In terms of testing, it seems that there is some kind of non-deterministic thing happens, which given the algorithm shouldn't be...
Definitely a feature I would like to see, if I can be of any help don't hesitate to ping me.
 Results are somewhat equivalent on univariate (default train/test) Number of shapelet sampled is below the default (10k) for small datasets as expected, will push the fix.
Hi, if by that branch you mean the branch from 2473 yes you can.
Hey @Kaustbh, PR is in, you can raise yours, thanks !