Another motivation for CmdStanR

I just noticed the page on CmdStanR at, and I think it’s worth mentioning one other motivation for having something like CmdStanR: stability. RStan seems to destabilize R in some way, causing it to unexpectedly crash at inopportune moments. It doesn’t happen a lot; usually it happens after enough time has passed that I’ve relaxed my guard and stopped being diligent about doing saveRDS() immediately after running HMC. :-( I expect that having Stan run in an entirely separate process should prevent this kind of thing from happening in CmdStanR.