I’m getting started with Pyro and am enjoying using it so far. I want to build a GLMM for dataset I have and am looking for recommendations on the best place to start.
I’ve found several of the Pyro resources on GLMMs:
- Optimal Experiment Design — Pyro documentation
- GitHub - pyro-ppl/brmp: Bayesian Regression Models in Pyro
- http://pyro.ai/numpyro/examples/ucbadmit.html
Are these the best places to start or does anyone knows of other good resources?
Specifically, I noticed the pyro.contrib.oed.glmm documentation mentioned that it’s being deprecated for brmp. However, it looks like there hasn’t been much recent activity on brmp.
Is brmp still the “future” or would I be better off following something like the numpyro ucbadmit example and building it manually?
I’d like to use the higher level abstraction from brmp but am open to building it out manually if necessary.