I am doing some simulation work where in each simulation iteration I call pre-compiled Stan models (using the rstan sampling function). When I go to a new factor-level combination of my simulation inputs I like to look carefully at the MCMC diagnortics for 30 or so trials. In my simulation R function I inserted calls to pairs and traceplot just after the call to sampling. The pairs plot come out fine, but not the traceplot results. Traceplot works fine when I save a stanfit object and call traceplot from the R command line. What am I doing wrong?
I am running on Windows 10 using using R 3.6.3 and rstan 2.21.2.