Tom Wenseleers

I can have a try - though I should say I'm not expert in the inner workings of `easystats`. E.g. I haven't checked whether `easystats `assumes predictions of multinomial models...

Well there seem to be some other issues related to support of `nnet::multinom` multinomial models - e.g. `insight::get_predicted_ci()` doesn't work yet, `insight::get_predicted()` works but has the predictions come out in...

Yes that's the question if that's what you want - for a multivariate type model like multinomial it's standard that predictions for each outcome level would come out in different...

Just to chime in here: `predict.multinom` indeed only has `type="probs"` (=`"response"`) implemented (annoyingly enough). However, to go from the response type to the link type as returned by `mclogit::predict.mblogit` with...

If you would be interested to allow for prediction intervals you might like to look into [https://rdrr.io/cran/ciTools/](https://rdrr.io/cran/ciTools/), which supports (log-) linear, (log-) linear mixed, generalized linear models, generalized linear mixed...

Many thanks for this - I found your notes to be really helpful! Interestingly though, for a very small problem I was testing here now (3 variables, 1000 cases, Poisson...

Ha that's great - many thanks for that! I see you use the same initialization strategy now as in R's implementation, using eval(family$initialize)! On my 4 core laptop with Intel...

Might be worth changing the irls function that you lay out to something like this, i.e. with correct initialization : ``` irls = function(X, y, weights=rep(1,nrow(X)), family=poisson(log), maxit=25, tol=1e-16) {...

Many thanks for that, that sounds great! And yes, the workaround you mentioned I was using right now - I just thought it would be a small thing to fix...

No worries, take your time, there's no rush! Below is an example to show that with some datasets `nnet:::predict.multinom` with `type="probs"` drops the reference level, while for others it gives...