Dominique Makowski

When the random factor matters, results are... weird: ## GAMM4 ``` r model Warning: Can't compute random effect variances. Some variance components equal zero. Your model may suffer from singulariy....

> maybe we can tag Gavin S.? Yes please! (don't think I know him)

Sad that we cannot bump on twitter 😬

Dear @gavinsimpson after your great post or random factors & gams, we have no option but to turn to your authority 😁 would you have by any chance some idea...

Here are 8 versions of theoretically equivalent models (with the help of @tjmahr 's new blogpost ☺️), with different engines and random effects entered as such or splines. ``` r...

Maybe our new jedi master @bwiernik has some thoughts on that ☺️

That sounds great indeed :) Looking forward to that! From a design standpoint, I don't know what you have in mind for bringing closer together easystats and domir, however since...

Regarding `estfun`: the [ipcr package](https://github.com/manuelarnold/ipcr/blob/master/R/get_scores.R) has named the function `get_scores()` and [maxLik](https://github.com/cran/maxLik/blob/master/man/gradient.Rd) calls it (I think) the "gradients"... all so weirdly unspecific!

nice work detective :) so how do you feel about adding such a function? I'd be okay with this, if only I was convinced that it could be useful in...

Awesome @strengejacke! Yeah, the implementation for mixed models makes sense, the only problem is that we don't have a ground-truth to compare it to as it is not done in...