Daniel Sabanes Bove

Could we add our own `suffstats` method to fix this? At least from https://math.stackexchange.com/questions/321952/beta-distribution-sufficient-statistic it seems this should be relatively straightforward?

Awesome, thanks so much @dmetivie ! The stochastic EM works perfectly fine here.

> In classical EM this is what version that is used at [each `M` steps](https://github.com/dmetivie/ExpectationMaximization.jl/blob/a865533287a960bdc43e664c9b46771e05a70781/src/classic_em.jl#L48-L49). > > And in the link above, you can check that it is currently not...

Hi @dmetivie , so I am happy that I got the weighted likelihood maximization working for the Beta distribution. However, somehow when using that in the backend for the classic...

Just tested again in a fresh folder, and the problem can be boiled down to this plot:  The green histogram is the fit from `fit_mle()` applied with a 2...

Thanks @dmetivie , that is super helpful! In that case, I guess I could also even pass mixtures with different number of components in the vector to `fit_mle()`? And it...

Thanks @jpdunc23 , that would be super helpful. I am currently trying to use `future.batchtools` with the LSF backend to run some `simChef` simulations on our HPC, and I am...

@hpages @LiNk-NY thx a lot for your work on improving this - and pls let us know if we can help in any way (besides raising issues, thx to @gogonzo...