There have been reports of discrepancies in the movingstd() function in Timelion, we should look into that.
More details here: https://github.com/elastic/timelion/issues/177
@polyfractal per our chat, feel free to work with @thomasneirynck on this if you have any insights. would be good to get your input.
similar issue logged here: https://discuss.elastic.co/t/timelions-movingstd-vs-kibanas-standard-deviation/50373
@thomasneirynck @polyfractal @alexfrancoeur since we don't have time to look into it, should we mark this function as "experimental" for now, like we do with some other functions in Timelion?
Should we maybe just yank it for now? At this point it's probably just classified as "buggy" rather than experimental, since it's giving potentially incorrect values.
I'm unfortunately still underwater with other work, too much to do at the moment on other projects and not enough time to spin up a kibana dev environment again :/
This has been fixed with https://github.com/elastic/kibana/pull/15911
Most helpful comment
@polyfractal per our chat, feel free to work with @thomasneirynck on this if you have any insights. would be good to get your input.