best
best copied to clipboard
Simplify and document the API
This PR allows one to not care about the underlying PyMC3 sampling process, and to compare two groups as simply as a t-test would be done.
Thanks to the introduced BestResults
classes, there’s no need to pass the data again to the plot_all
function.
>>> best_out = best.analyze_two(y1, y2)
>>> best.plot_all(best_out)
Add analysis and plotting function for one-group analysis (for example for paired samples):
>>> best_out = best.analyze_one(y)
(This is part 2 of 4 pull requests that build on top of one another.)
Apologies: at the time of the pull request this branch already included the changes from pull request #8 . Now I force-pushed this branch and those that built upon it.
Could you please accept/reject/comment on it? Thank you.
Oh, I am really sorry I missed this and your other pull requests for some reason. I am planning to archive this repository in favor of linking to PyMC upstream and I was just checking the open issues and pull requests and was surprised to see your work. Since it has been some time since you filed this, what would be your preferred option at this point?