Phillip Alday

This is not currently possible and it's not exactly trivial to implement in general. I need to update the docs here, but I have some [lecture notes with comments on...

Oh, I meant to post the link in my original comment, but here it is now 😄 https://palday.github.io/economics2024/01-intro.html#limitations-of-mixedmodels.jl

> That document states: > > > MixedModels.jl does support constraining the residual variance to known scalar value, which is useful in meta-analysis. > > Can that value be zero?...

Wait, is the actual desired constraint simply a diagonal covariance structure? Something like the `||` in lme4?

I think this is stale/resolved now, but feel free to re-open if that's wrong!

I don't know if the example here is artificial, but I'm not sure I think it makes sense to try to copy the result of `aes(x, y)` into Julia. Here's...

What are you trying to accomplish? There is nothing in MixedModels.jl that directly corresponds to [`StatsModels.ModelMatrix.assign`](https://github.com/JuliaStats/StatsModels.jl/blob/368601ef811066e3631eae81a0c5a538ce34e962/src/modelframe.jl#L192-L213). Depending on what you're trying to accomplish, I can potentially provide better guidance.

There is already a VIF implementation for mixed models: https://palday.github.io/MixedModelsExtras.jl/stable/api/#MixedModelsExtras.vif. The implementation there would also work for GLM.jl models with an appropriate method for `termnames`. Indeed the type restriction for...

@nalimilan I have a slight preference to have `gvif` and `vif` as separate functions, much like `r2` and `adjr2` are. We don't have to issue the warning for categorical models...

I think the MixedModels portion of this is complete and the necessary stubs have been upstreamed to StatsAPI/StatsModels. Anything else is a GLM.jl issue, so I'm closing this. :smile: