Allow standalone generate_quantities using non-HMC fit

[WardBrian]

Submisison Checklist

• Run tests: ./runCmdStanTests.py src/test
• Declare copyright holder and open-source license: see below

Summary:

This closes #1099. Currently, generate_quantities assumes there are exactly 7 columns of non-parameters before the parameter draws in the CSV, as this is the case in HMC. However, no further assumptions are made on the draws for this method.

This PR simply changes from using a hardcoded value of 7 to first counting the number of columns which end with "__", then proceeding as before.

Following the existing design in terms of the warmup draws, this makes no attempt to avoid running generate_quantities on pre-converged optimization paths. It will run for each row in the CSV, leaving it up to downstream interfaces to sort this out.

Intended Effect:

generate_quantities can now accept files for fitted_params which are the results of optimization or VI.

How to Verify:

A test is added which uses a CSV file generated from LBFGS.

Side Effects:

Should be none. The only changes are in command.hpp, none of the upstream IO needed adjusting, since the specific information we need from the CSV is agnostic.

Interfaces like cmdstanpy/cmdstanr will need updates to support this

Documentation:

I believe that this PR actually makes the existing documentation more correct. The current documentation states

This method requires sub-argument fitted_params which takes as its value an existing Stan CSV file

without specifying that this CSV file must be from sampling.

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):

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/)