Thanks Christian! Sorry I didn’t formulate my question well. What I’d like to do is to change the number of points in credible data sets generated by sample_posterior_predictive. In my example this number is always 5 (same as in the observed data Y), but I’d like to increase it to 10 (just as an example). In this particular case I should be able to “lump” two credible data sets with 5 observations each to get what I want, but that seems like a poor practice in a general case. Did I clarify it at all?