JuliaStats
Additional details in documentation for Grouping()
Hi, great package, I just have a small suggestion for the documentation.
I was using MixedModels to fit a model with many levels and got the following error:
┌ Warning: Random effects grouping variables with many levels can cause out-of-memory errors. Try manually specifying `Grouping()` contrasts for those
variables.
It took me a while to figure out how to use Grouping() and contrasts, until I saw the example in the tests below. I think it would be worth having something like fit(MixedModel, @formula(y ~ 1 + (1 | grp)), d, contrasts=Dict(:grp => Grouping())) in the documentation for Grouping.
https://github.com/JuliaStats/MixedModels.jl/blob/aae37e9dd724d6eb8ded292a9a8798300bb2a562/test/grouping.jl
Happy to create a pull request if you think this is worthwhile.
We hadn't given much more documentation previously because we had only encountered it in a few places in the wild and Grouping() works like any other contrasts / hints for the StatsModels formula interface.
That said, it would be a great thing to have it better documented. Do you have an idea for where to document it so that it's easily discoverable?
My suggestion would be to make sure the docstring for Grouping explains how to use it (with an example invocations for lm and fit) and add a pointer to the docstring in the warning message (something like "for more, see ?Grouping")
