Hi
I am running Stan models through brms on a computing cluster.
After a recent rebuild of the cluster management system, I encountered a problem in running brms code to fit models. The code has worked previously.
I am working in R 4.3.1. Loading ‘brms’ package (version 2.21.0).
To try to locate the problem, I ran the tutorial model code – from here:
fit1 <- brm(count ~ zAge + zBase * Trt + (1|patient),
data = epilepsy, family = poisson())
The result is:
Error in stanc(file = file, model_code = model_code, model_name = model_name, :
0
Semantic error in 'string', line 5, column 12 to column 13:
-------------------------------------------------
3: }
4: data {
5: int<lower=1> N; // total number of observations
^
6: array[N] int Y; // response variable
7: int<lower=1> K; // number of population-level effects
-------------------------------------------------
Integer literal cannot be larger than 2_147_483_647.
This error message is the same as the message I get running my own model code on the cluster.
I haven’t run brms on my own computer (old version of R and of brms which works fine).
Any advice on what the problem may be would be much appreciated.