Brian Ward

Thanks a ton @amas0, those are all super helpful to see laid out like that. > I think the question comes down to whether there's interest in including polars as...

Hi @amas0 @eroesch — are either of you interested in the next steps here (first spreading the new CSV reading to other methods, later updating all the existing comment scanning...

After #806 the first bullet point here is officially done (thanks again @amas0!) It sounds like you're also interested in bullet point 2 (the JSON output of `save_cmdstan_config`), so I...

`@gpu` would make the most sense to be implemented 1) by someone other than myself who understands them more, and 2) as part of something like #1353. In particular, at...

I can confirm that `model info` doesn't report this, and by going back and trying the first commit the `info` command was added in, it seems as though it _never_...

This will be moved to a function independent of the `CmdStanModel` object, like compilation itself has already been. Good catch though, we might want to do this as a 1.2.2...

@bob-carpenter profiling is done with [an object](https://github.com/stan-dev/math/blob/92075708b1d1796eb82e3b284cd11e544433518e/stan/math/rev/core/profiling.hpp#L133) that uses the constructor and destructor to registers some callbacks on the autodiff stack for timing, so it should be 0 cost if...

I see two potential solutions to this: 1. Wrap the body of the constructor of the `profile` class and the entire destructor in an `#ifdef` 2. Add a flag to...

There are two issues here: - in general, cmdstan’s current build system is just not set up to work without write access to its own directory. See https://github.com/stan-dev/cmdstan/issues/1175 for some...

That might be necessary if different models are needed at different points during the cluster run. In my experience (with cmdstanpy, not cmdstanr, but I imagine the same is possible)...