You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi!
Since a lot of AB tests involve more than 2 groups, have you thought about adding support for more 2 groups? Or do you advise using the current structure and just compare all groups to A?
Thank you for creating the package.
Best,
Miha
The text was updated successfully, but these errors were encountered:
Hi guys - totally open to a change like this if it's handled elegantly with as little disruption as possible to current API/docs. I assume it won't be possible to do it 100% seamlessly and keep all examples the same and therefore will be considered a pretty major release.
There is also the possibility of adding hierarchical partial pooling and hyperpiors, if we have more than 2 groups. This would mean changing the backend and possibly losing some of the closed form conjugate solutions. But perhaps worth augmenting the library with that as well. Maybe as an option that externs to some MCMC library.
Personally I have no time to do either of these 2 but I would be happy to support one or both of these approaches with high level feedback and in-depth code reviews if somebody else wants to give it a shot.
Hi!
Since a lot of AB tests involve more than 2 groups, have you thought about adding support for more 2 groups? Or do you advise using the current structure and just compare all groups to A?
Thank you for creating the package.
Best,
Miha
The text was updated successfully, but these errors were encountered: