I’m running a
stan_glm() model via
rstanarm on a multi-core server with +150 cores and +1 TB of RAM. It’s not clear to me that the number of cores I set is correct.
Let’s say I instantiate and run a Poisson GLM model with 10 chains and 90 cores:
library(rstanarm) options(mc.cores = 80) model = stan_glm(formula = A ~ B, data=my_data, family=poisson(), cores=80, chains=10)
When I run this though, it appears to approximately 10 cores.
How can I re-run this to use all 80 cores?
Operating System: Unix
Interface Version: rstanarm