Closed seanlseymour closed 5 years ago
I'd love to see the requirement for bottleneck
fall off. Packages consumers can still use it, and the codebase is already set up to eagerly load it if available in the environment, else fall back to numpy.
@twiecki and team, open to maintaining bottleneck support but dropping it as a requirement ? We've seen some packaging issues with it (off an on) over the years.
Yes, should be optional.
@seanlseymour I may work on a PR later this week, unless you have already started on it.
No, I haven’t started anything here. Would appreciate!
Sean
On Aug 21, 2019, at 3:30 PM, Calvin DeBoer notifications@github.com wrote:
@seanlseymour I may work on a PR later this week, unless you have already started on it.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.
Thank you!
On Aug 23, 2019, at 1:59 PM, Thomas Wiecki notifications@github.com wrote:
Closed #113 https://github.com/quantopian/empyrical/issues/113 via #115 https://github.com/quantopian/empyrical/pull/115.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/quantopian/empyrical/issues/113?email_source=notifications&email_token=ADAIU6LQFNJO2JUE5ATRTXTQGBFSXA5CNFSM4ILN2MJ2YY3PNVWWK3TUL52HS4DFWZEXG43VMVCXMZLOORHG65DJMZUWGYLUNFXW5KTDN5WW2ZLOORPWSZGOTHKHPYQ#event-2580838370, or mute the thread https://github.com/notifications/unsubscribe-auth/ADAIU6IVOYMHTOLRF4O5FDDQGBFSXANCNFSM4ILN2MJQ.
@cgdeboer I'm new maintainer of bottleneck and would appreciate it if you could open an issue with some details on the packaging issues you've encountered, as the upcoming release of 1.3.0 should solve them: https://github.com/pydata/bottleneck
Installing fails using Python 3.6 and later as provided in current conda distribution. Bottleneck package dependency seems to fail to install.