In my Stan model, I fixed one parameter to zero to satisfy some identifiability constraints.
When I sample, I obtained the desired sampling but I obtain warning messages that the largest Rhat is NA because of the fixed parameter.
Warning messages: 1: The largest R-hat is NA, indicating chains have not mixed. Running the chains for more iterations may help. See http://mc-stan.org/misc/warnings.html#r-hat 2: Bulk Effective Samples Size (ESS) is too low, indicating posterior means and medians may be unreliable. Running the chains for more iterations may help. See http://mc-stan.org/misc/warnings.html#bulk-ess 3: Tail Effective Samples Size (ESS) is too low, indicating posterior variances and tail quantiles may be unreliable. Running the chains for more iterations may help. See http://mc-stan.org/misc/warnings.html#tail-ess
mean se_mean sd 2.5% 25% 50% 75% 97.5% n_eff Rhat valid Q5 Q50 Q95 MCSE_Q2.5 MCSE_Q25 MCSE_Q50 MCSE_Q75 MCSE_Q97.5 MCSE_SD Bulk_ESS k 0 0 0 0 0 0 0 0 NA 1 1 0 0 0 NA NA NA NA NA NA 4000 Tail_ESS k 4000
In the forum, I saw some “tricks” to avoid this message, for example:
- Include the rstan sampling call in
suppressWarnings(). This will indeed suppress warnings but might suppress also other important warnings as well.
- Include the parameter constraint in the
modelblock instead of the
transformed parametersblock. That would indeed work but I would like to keep the constrained parameter in the set of my parameters.
My general question is then: is it possible to fix one parameter in the parameter space without having this “false” warning?
Thank you for your help!