Extend get_dims and get_param_names in model_base

[WardBrian]

Submission Checklist

• Run unit tests: ./runTests.py src/test/unit
• Run cpplint: make cpplint
• Declare copyright holder and open-source license: see below

Summary

Closes stan-dev/stanc3#1240. New flags are added to these methods to allow excluding the tparams or generated quantities, and these are used in initialization.

Copyright and Licensing

Please list the copyright holder for the work you are submitting (this will be you or your assignee, such as a university or company):

Simons Foundation

By submitting this pull request, the copyright holder is agreeing to license the submitted work under the following licenses:

• Code: BSD 3-clause (https://opensource.org/licenses/BSD-3-Clause)
• Documentation: CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

[WardBrian]

It seems like we have a bit of a chicken-egg problem here. If we delete the older function definition from stanc before updating model_base, then the stanc3 tests will all fail, but if we generate both versions in stanc3, then PRs against Stan will fail due to the ambiguity of existing calls.

We reach a happy state only after model_base is updated and stanc3 generates only one signature, I believe. @SteveBronder - thoughts?

[WardBrian]

Jenkins passes when using stan-dev/stanc3#1245 up until the comparisons tests, which by definition can't run since the different versions can't use the same compiler output

[bob-carpenter]

This looks good on the C++ side. Thanks for adding docs and tests.

I'll defer to others more in touch with our CI to merge.

[wds15]

An @bgoodri and @hsbadr have a quick look at this and evaluate on how difficult this change will be to manage for rstan?

[bob-carpenter]

have a quick look at this and evaluate on how difficult this change will be to manage for rstan?

On the code side, probably nothing because of the default arguments, but at most two extra boolean arguments to get_param_names and get_dims calls.

So it's just a matter of whether it'll cause further complications with staging on CRAN for RStan and StanHeaders.

[bgoodri]

All past models will have C++ code without those flags so somehow I would have to avoid passing them in that situation.

[wds15]

OK... I was just worried about the rare need for binary compatibility in the model class, but it sounds as if we are good.

[bob-carpenter]

@bgoodri: Can you avoid passing them? I couldn't tell if you were saying it would be OK for RStan or a problem.

[WardBrian]

Thanks to some work by @serban-nicusor-toptal this is now able to pass all the tests including the performance tests when tested against stan-dev/stanc3#1245.

I'd love to see this merged before the next release