Potential Problems with rstan and /or brms packages in R 4.0.2

Great and thanks!

Yes, I know about that Installation Guide. It usually works well. I have a hunch that there are some issues with some combination of the latest rStan, R, RTools and brms all not playing nice together!

I suspect it may be a good idea to avoid the latest version of R for a while, and advise my students to install v3.6.x, etc. (at least on Windows10).

Anyway, you’ve been extremely helpful. If I ever run into you in person, I owe you a beer/coffee.

No worries! Yeah, while the getting started is helpful for the basic setup, it could really use a section for dealing with common issues that people come across.

Are there any updates on this topic? I am teaching two stats classes where I introduce them to Bayesian using brms and it has been a complete nightmare trying to get things to work (I’ve never had this issue before). Many students are having to give up. This is happening on both Macs and PCs, with an incredibly diverse set of error messages showing up. My own combination of operating system, R, brms, Xcode and rStan is sort of working, but only in RStudio (not the default R interface) and despite throwing an error message on every attempt to compile (it produces results despite the error). Does anyone have a comprehensive idea of how to get all the levels to play nice with each other? Or do we know what changed this summer that made all hell break lose? Thanks for any thoughts or guidance.

1 Like

I’ve updated the Windows installation instructions to avoid most of the common issues (seems to have the done the trick, given the reduction in support posts recently). I haven’t had time to properly run through the Mac install instructions yet (since I need to set up VMs to test properly), but will be getting through that in the next couple of weeks.

In the interim, R4.0 tends to play best with rstan across both pre-Catalina and Catalina OSes. Stephen Martin had a great writeup of installation/setup instructions in this post as well: R 4.0, rstan, and you

1 Like

Thank you so much!! I’ll point the students to these instructions and with a little luck things will settle down. Thank you for all you do. Cheers, Emily