Brian Ward

Is this just for convenience or do we expect different performance? I thought to_matrix was essentially free for eigen types

IMO I think we should allow target+= in TParams. @bob-carpenter?

It was a quick change to allow `target +=` (#980). The idea of a boolean to allow target adjustments to depend on a jacobian parameter is probably a larger scope...

Yeah, we don't need to update examples, I'm more just thinking that if there are any statements like "You can only use target += in the model block" they will...

@betanalpha - would you propose that we disallow the calling of `_lp` functions in transformed parameters then? I think that would be a rather large breaking change if anyone was...

I’ve let #971 fall out of date a little since it didn’t seem like there was a consensus on merging it, but composing offset/multiplier with other transforms is definitely still...

I think this is intentional behavior. We currently only support SoA in user defined functions if we can inline them (see #1237), and it is unclear whether or not the...

At the moment, the optimizer has no way of knowing whether the reason it is doing the memory pattern optimizations is because it is from `-fsoa` or `--O1` -- both...

Useful links: - [Core_kernel.Debug](https://ocaml.janestreet.com/ocaml-core/v0.13/doc/core_kernel/Core_kernel/Debug/index.html) - [Ocamldebug](https://ocaml.org/manual/debugger.html) - [Print_s](https://ocaml.janestreet.com/ocaml-core/v0.13/doc/core_kernel/Core_kernel/index.html#val-print_s) (this isn't really explained, but `print_s [%sexp (e : typed_expression)]` will print the s-expression form of e, or any type with the...

We actually have —explain enabled by default, but the results end up in a rather unhelpful place (`./_build/default/src/frontend`) There’s definitely a lot more stuff we can document on the getting...