dmphillippo / multinma

Network meta-analysis of individual and aggregate data in Stan
https://dmphillippo.github.io/multinma
35 stars 17 forks source link

mixed likelihoods #42

Open gbrlrgrs opened 5 months ago

gbrlrgrs commented 5 months ago

Hello @dmphillippo. Any way of using different likelihoods / links on different elements of a dataset contributing to a shared model? Motivation: if I use cloglog--binomial on dichotomous data, I end up on log-hazard scale so, in principle, there's no reason why I couldn't combine with explicitly reported rate data (log--Poisson model), with both estimating log(HR)s. But combine_network() complains of a type mismatch. Is there any way around that?