Michael Panchenko

@ivanappliedai will probably address this in the near future

Blocked by #956

Will be addressed in #937

Thanks for reporting this! I was also wondering why v_s_ was computed additionally, doesn't seem like a good idea. I don't understand why your change affected performance. Also, interesting observation...

Sorry for the late reply @spacegoing. I recently was added to the maintainers and am now going through the issues. The reaction rate should go up soon. Could you pls...

I see. One should make use of the truncated flag to repeat computation as little as possible. And these methods should be covered by tests first. See also #893, which...

Also related to #881 - having an episode_id in the batch would allow finding the right `bath.obs` at the and of the episode

> Sorry. That one was a misleading comment LoL. > > I run 10 runs for each config and there wasn't much difference. I guess that was an unlucky v.s....

@arvganesh I'm going through older issues from time to time and now noticed this one. It's related to #933 and #895, both of which are good issues for a first...

Finally, the combination of sourcing `conda.sh` directly in `.zshrc`, @iwei20's fix eliminating carriage returns with @mlaves fix for escaping backslashes, conda works with oh-my-zsh under git bash. Thank you all...