Andrew Johnson

> What's the eventual goal? I'm updating the handling of unconstrained pars in `cmdstanr` to be structured using the [`draws` formats from `posterior`](https://cran.r-project.org/web/packages/posterior/vignettes/posterior.html#draws-formats) (motivation in [this issue](https://github.com/stan-dev/cmdstanr/issues/730)). For that I...

> I wrote back on the original issue asking why. I don't understand the motivation and have the same concerns as @mitzimorris expressed about the proposed solution. Note that the...

And also to clarify that the proposed solution in the linked post was using the existing structure for specifying initial values that is used by both `rstan` and `cmdstanr`

I think it's fine to update the RNG to check that N is non-nonnegative, instead of positive - since this is what the [binomial_rng does](https://github.com/stan-dev/math/blob/e73651b991c8a08f045ad09e0c0124eeaaae8c90/stan/math/prim/prob/binomial_rng.hpp#L43)

According to the log, it's running four parallel build jobs. Is it possible to reduce that in the autobuilders? That would be the easiest way to reduce the amount of...

Sorry about the delay, I'll be able to test this locally today and will see what a fix could be

The culprit here is the `src/grammar_inst.cpp` file, which has several thousand lines of code to be compiled. A workaround here is to split the file into smaller parts, so that...

That's a great question! I've raised it with the NHSRplotthedots/MDC team for a definitive answer: https://github.com/nhs-r-community/NHSRplotthedots/issues/200

> @andrjohns did you get anywhere with NHSRplotthedots/MDC team? If not I might see if I can follow it up as well. Unfortunately nothing yet, there hasn't been any response...

> @andrjohns. Update after meeting on 17/04/2024. Am going to speak directly to MDC director about this. Brilliant, thanks for following this up! Once we have an 'official' word, I'll...