Problems in defining weakly informative priors for a categorical multilevel mixture model

Just add 1e-9 instead of clip