Bob Carpenter

> the API guarantees for the timing were never well-defined in the first place. That's why I've never used those internal timings. The timing I always care about is in...

Thanks for the proposal. The approach you propose is similar to what C++ does under the hood with variadic arguments (as @bgoodri alludes to). This has been generalized type safely...

Thanks. That makes the problem clear. Are you really going to have different number of predictors for the two entries so you can't just make it matrix multiplication? Anyway, it's...

Yes. I'm proposing adding: ``` vector[] to_array_of_2vectors(vector x, vector y); ``` where the two vectors have the same size and ``` to_array_of_2vectors(x, y)[i, 1] == x[i] to_array_of_2vectors(x, y)[i, 2] ==...

I'm only concerned about the external client interface, not how it's implemented. But having said that, if it's all scalars under the hood, how will vectorization help? I'd like to...

I take it you meant `binormal` and not `normal` in those first two!

This doesn't quite match the current (2.18) syntax, in which local variables in the model block may not have constraints. It also doesn't match the semantics, in that constraints on...

I started a discussion on Discourse about how to evaluate parallelization of log density or MCMC algorithm. https://discourse.mc-stan.org/t/evaluating-parallelization-performance/10962

> On Sep 12, 2019, at 7:48 PM, wds15 wrote: > > He made the suggestion in the parallel HMC thread... and I do agree. > > I do want...

> On Sep 12, 2019, at 11:30 PM, wds15 wrote: > > Why can’t we have a multi chain warmup which are terminated once done doing warmup. Then the CmdStan...