I’ve seen the Mixedmodels.jl package has thre differente branches:
- master
- nlmm
- dropmulαβ!
How are they different? What’s new there?
If you add MixedModels you install the latest release version of the package.
Think of master as the most recent version of the next release version. You can install this version when there is a bug fix or a new feature in the master that hasn’t been officially released with add MixedModels#master. I only do that when the developers of the package themselves suggest it as a temporary solution.
The other branches are usually works in progress that haven’t been merged with master. For example if you find a bug which you want to fix and share your code with others, you should create a branch (e.g. Juan/fixbug) and ask for a pull request (essentially the permission to be merged to master and be available for the next release). Your changes will get reviewed and if accepted your branch will get merged to master and then it gets deleted.
Sometimes there branches have already been merged with master or are already in previous releases but do not get deleted for various reasons.
I would suggest to ignore the other branches and install the package as per instructions which uses the latest release. These branches might have bugs, performance regressions etc that need to be ironed out before release.
1 Like
I understand but I was just wondering whether there where new experimental features included on those branches.