BRMS formulation of `stan_jm`

[mattwilliamson13]

Hello,
I've been trying to fit some joint longitudinal/survival models using stan_jm (which is great), but I need to be able to access some additional properties and compute different estimates beyond those that are exported via the existing utility functions. I've tried to do this by simply modifying the functions that R uses to send to the pre-compiled rstanarm model, but the inability to trace the anonymized arguments to the actual stan code is starting to make this impossible. I'm hoping that some version of the stan_jm modeling approach could be made available in brms because:

• brms syntax is (in my opinion) combines flexibility with readability in a way that would help me diagnose whether the model I'm fitting is actually the model I meant to fit.
• development for the stan_jm utilities and plot functions (as well as survival models more generally) in rstanarm seems to have stopped
• The ability to integrate the brms objects with the existing universe of brms helper and diagnostic functions would significantly expand the ability to interrogate these joint models.

I know there are lots of feature requests out there, but given the advances in survival modeling in brms in the last year or so, I was hoping this might be a relatively light lift. Or perhaps there is already a way to implement this with the existing notation (though my lest check of the discourse channel suggested otherwise).

Thanks for an excellent package and any help you might be able to provide.