I am running a Bayesian model on a high performance computing cluster, and I keep running into the following error: "Negative length vectors are not allowed"
It occurs whether I run the model in parallel or not.
error with rjags::run.jags -- parallel
error with rjags::autorun.jags -- parallel
error with R2jags::jags -- non parallel
I also get the error if I parallelize and run jags simply across the nodes. I do not think the error is due to model specification or the data being fed to the model, as I have successfully run the model on a local r session (for a short MCMC run).
Has anybody run into this error before?
any help is greatly appreciated. thanks