Add threading for variational/optimize

[rok-cesnovar]

Summary

Fixes #193

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):
Rok Češnovar, Uni. of Ljubljana

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)

[rok-cesnovar]

Not so sure about the name here. but threads_per_chain does not work for optimize/variational.

[jgabry]

Yeah that's a good question. Ideally we would use the same argument name but you're right that there are no chains here. I guess threads is fine. @mitzimorris or @bbbales2 any thoughts on what name we should use here?

[jgabry]

I put a few comments, but this looks good to me, thanks!

[rok-cesnovar]

Thanks. I guess the only thing to discuss is whether threads = X is a good argument for optimization/variational. It's a different name than sampling, but threads_per_chain does not work for optimization.

[jgabry]

Thanks. I guess the only thing to discuss is whether threads = X is a good argument for optimization/variational. It's a different name than sampling, but threads_per_chain does not work for optimization.

I think threads is good but we can wait a little while to see if anyone has a different suggestion.

[codecov-io]

Codecov Report

Merging #369 (75e7e37) into master (b9ce4d9) will decrease coverage by 2.21%.
The diff coverage is 72.34%.

@@            Coverage Diff             @@
##           master     #369      +/-   ##
==========================================
- Coverage   90.58%   88.37%   -2.22%     
==========================================
  Files          12       12              
  Lines        2571     2614      +43     
==========================================
- Hits         2329     2310      -19     
- Misses        242      304      +62     

Impacted Files	Coverage Δ
R/model.R	88.79% <69.04%> (-3.56%)	⬇️
R/read_csv.R	97.56% <100.00%> (-1.21%)	⬇️
R/run.R	96.55% <100.00%> (-2.30%)	⬇️
R/install.R	43.54% <0.00%> (-8.41%)	⬇️
R/utils.R	92.20% <0.00%> (-1.30%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b9ce4d9...75e7e37. Read the comment docs.

[bbbales2]

I'm happy with threads. I hope in the future we do a multiple-runs sorta VI (everything should be checked -- https://arxiv.org/abs/2009.00666), but that isn't here now.

[jgabry]

Yeah I think multiple-run VI would be a good thing in the future. I guess we can adjust as necessary when that time comes.

[jgabry]

Seems like we're going with threads at least until there's multi-run VI. That seems fine so I'm going to go ahead merge this because the only thing holding it up was that argument name.