PSR: Return early if no draws

[andrjohns]

Submission Checklist

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

Summary

stan::analyze::compute_potential_scale_reduction still attempts to calculate the PSR if there are 0 draws passed, which is currently causing a segfault when calling stansummary on the output from optimize:

andrew@Andrews-MacBook-Air bernoulli % ./bernoulli optimize data file=bernoulli.data.json
...

andrew@Andrews-MacBook-Air bernoulli % ../../bin/stansummary output.csv
Warning: non-fatal error reading adaptation data
Assertion failed: (index >= 0 && index < size()), function operator(), file DenseCoeffsBase.h, line 181.
zsh: abort      ../../bin/stansummary output.csv

Intended Effect

No segfault

How to Verify

Call stansummary with this applied.

Side Effects

N/A

Documentation

N/A

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): Andrew Johnson

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/)

[syclik]

@andrjohns, the change looks great.

Before it gets merge, just a question: is quiet_NaN the right behavior here? I don't know how this gets used next, so just wanting to sure this isn't going to trigger other problems. (I'm guessing that NaN is the right choice; just wanting to double check.)

[WardBrian]

The only place this is used is for user-facing output (from stansummary) so I believe NaN is a reasonable choice. We also return NaN if all chains are constant a bit further down in the function