Hi all - R and cmdstan user here trying his hands on Julia and Turing.
Came across TuringGLM and I like its smooth interface akin to what brms is to RStan.
I do want to learn the raw Turing code though, and was wondering if there is a way to export the code generated under the hood by TuringGLM? I can then use that as a base and tinker with it for a more complicated structure for example.
i.e. Is there an equivalent function to the make_stancode function in brms?
Thanks in advance!
As far as I understand it TuringGLM does not generate code in the way brms does.
It’s actually more comparable to rstanarm in that it specifies some models that can be manipulated by prior choices etc.
Anyways the model code for those models is still available in TuringGLM.jl/src/turing_model.jl at main · TuringLang/TuringGLM.jl · GitHub
where all the _model functions contain the respective Turing models.
I recently proposed TuringRegressions.jl, which includes brms-style export, changes to and re-import of model code, if it’s of interest?