wds15

May I ask what the status of the backend is? Is it still planned to include clustermq as a backend to future or is there already a way to get...

But with the `varmat` magic bullet it should make a performance difference if one allows matrices for `y`. So we would have to pick a convention. How about `mu` being...

in case mu is a matrix, then one needs a convention, sure... but I thought we also have the case of mu being a vector, no? (I am probably too...

For my understanding: Is the desire to take in matrices to get better performance? If so, then a prototype demonstrating the speed advantages would be a good motivation to sort...

There are two cases to discriminate: - The user defines `STAN_THREADS`... then then AD tape is thread local for which you pay a bit of performance (quite platform dependent). In...

From the ode tests it is the djoint solver which is eating up most of the execution time. The adjoint solver must be run with high accuracy to get the...

@avehtari this sounds great...shouldn't we have a different sigma for the intercept and all other regressors? I would appreciate that as I don't like that much data-driven priors which you...

Ripping code apart into testable things is not at all not smart from what I learned.

One different approach I did for one of my R packages where I wanted to run SBC is to proxy the test results from the SBC run. So what I...

How do you run the benchmark? In case you use `expose_stan_functions` then you do not calculate the gradients of that expression ... and that is the real bummer for a...