Add virtual function model_compile_info to the model_base.hpp

[rok-cesnovar]

Submission Checklist

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

Summary

Adds the virtual function model_compile_info to the model_base, which will be used to retrieve the info about the compiler version and flags used.

Related Cmdstan PR stan-dev/cmdstan#896 and issue stan-dev/cmdstan#887

Stanc3 now adds this function to the model:

  std::vector<std::string> model_compile_info() const {
    std::vector<std::string> stanc_info;
    stanc_info.push_back("stanc_version = stanc3 e010e06c");
    stanc_info.push_back("stancflags = --warn-pedantic");
    return stanc_info;
  }

This PR adds a virtual function model_compile_info() and the generation of the same function (sans flags) in order to still support compiling models with STANC2=true. As compiling with STANC2 is only for debugging purpose, I did not implement the stancflags functionality for stanc2 as well.

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, Univ. 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)
• Documentation: CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

[mitzimorris]

@rok-cesnovar - branch names should include issue number

[rok-cesnovar]

I keep forgetting on the branch names. I made the branches before the issues and then forgot to rename them.
I also got really used to the Github CLI, where you can just do gh pr checkout 2932.

Anyhow this should be ready for review. I added the generation of the model_compile_info to stanc2 for 2 reasons:

• the tests still work with stanc2
• compiling with STANC2=true would break

The relevant issue is stan-dev/cmdstan#887

[bob-carpenter]

Is there a reason this can't be a regular function instead of a virtual one? It just seems like something that would fit into the basic CRTP structure of the generated model class.

[stan-buildbot]

---

Name	Old Result	New Result	Ratio	Performance change( 1 - new / old )
gp_pois_regr/gp_pois_regr.stan	4.03	4.0	1.01	0.8% faster
low_dim_corr_gauss/low_dim_corr_gauss.stan	0.02	0.02	0.97	-2.68% slower
eight_schools/eight_schools.stan	0.09	0.09	1.03	2.48% faster
gp_regr/gp_regr.stan	0.19	0.19	1.0	0.39% faster
irt_2pl/irt_2pl.stan	5.28	5.37	0.98	-1.72% slower
performance.compilation	87.33	85.07	1.03	2.59% faster
low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan	7.69	7.72	1.0	-0.44% slower
pkpd/one_comp_mm_elim_abs.stan	20.49	21.55	0.95	-5.16% slower
sir/sir.stan	93.14	97.63	0.95	-4.82% slower
gp_regr/gen_gp_data.stan	0.04	0.04	0.99	-1.41% slower
low_dim_gauss_mix/low_dim_gauss_mix.stan	3.13	3.03	1.03	3.3% faster
pkpd/sim_one_comp_mm_elim_abs.stan	0.32	0.32	0.98	-1.94% slower
arK/arK.stan	1.79	1.79	1.0	-0.05% slower
arma/arma.stan	0.69	0.59	1.16	13.78% faster
garch/garch.stan	0.53	0.52	1.02	1.98% faster
Mean result: 1.00682847769

---

Jenkins Console Log
Blue Ocean
Commit hash: 8c5ef352d01ac83747fa4ff6dc3dfb400f0c0547

---

Machine information

ProductName: Mac OS X ProductVersion: 10.11.6 BuildVersion: 15G22010

CPU:
Intel(R) Xeon(R) CPU E5-1680 v2 @ 3.00GHz

G++:
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin15.6.0
Thread model: posix

Clang:
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin15.6.0
Thread model: posix

[rok-cesnovar]

Is there a reason this can't be a regular function instead of a virtual one? It just seems like something that would fit into the basic CRTP structure of the generated model class.

If I dont put the virtual method in model_base I get the following error

In file included from src/cmdstan/main.cpp:1:0:
src/cmdstan/command.hpp: In function ‘int cmdstan::command(int, const char**)’:
src/cmdstan/command.hpp:131:51: error: ‘class stan::model::model_base’ has no member named ‘model_compile_info’
     std::vector<std::string> compile_info = model.model_compile_info();
                                                   ^~~~~~~~~~~~~~~~~~

@bob-carpenter would you mind reviewing this? I updated the top-level comment with more details.

[stan-buildbot]

---

Name	Old Result	New Result	Ratio	Performance change( 1 - new / old )
gp_pois_regr/gp_pois_regr.stan	4.03	4.08	0.99	-1.07% slower
low_dim_corr_gauss/low_dim_corr_gauss.stan	0.02	0.02	1.0	0.02% faster
eight_schools/eight_schools.stan	0.09	0.09	1.0	-0.0% slower
gp_regr/gp_regr.stan	0.19	0.19	0.98	-1.85% slower
irt_2pl/irt_2pl.stan	5.33	5.33	1.0	-0.15% slower
performance.compilation	85.9	85.0	1.01	1.05% faster
low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan	7.69	7.74	0.99	-0.55% slower
pkpd/one_comp_mm_elim_abs.stan	23.26	22.66	1.03	2.56% faster
sir/sir.stan	92.18	98.31	0.94	-6.64% slower
gp_regr/gen_gp_data.stan	0.04	0.04	1.0	-0.47% slower
low_dim_gauss_mix/low_dim_gauss_mix.stan	3.01	3.1	0.97	-3.07% slower
pkpd/sim_one_comp_mm_elim_abs.stan	0.32	0.33	0.97	-2.81% slower
arK/arK.stan	1.78	1.78	1.0	0.2% faster
arma/arma.stan	0.66	0.59	1.12	10.43% faster
garch/garch.stan	0.54	0.52	1.04	3.91% faster
Mean result: 1.00241562789

---

Jenkins Console Log
Blue Ocean
Commit hash: 8c5ef352d01ac83747fa4ff6dc3dfb400f0c0547

---

Machine information

ProductName: Mac OS X ProductVersion: 10.11.6 BuildVersion: 15G22010

CPU:
Intel(R) Xeon(R) CPU E5-1680 v2 @ 3.00GHz

G++:
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin15.6.0
Thread model: posix

Clang:
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin15.6.0
Thread model: posix

[stan-buildbot]

---

Name	Old Result	New Result	Ratio	Performance change( 1 - new / old )
gp_pois_regr/gp_pois_regr.stan	4.0	4.2	0.95	-4.87% slower
low_dim_corr_gauss/low_dim_corr_gauss.stan	0.02	0.02	0.99	-1.52% slower
eight_schools/eight_schools.stan	0.09	0.09	1.04	3.47% faster
gp_regr/gp_regr.stan	0.19	0.19	0.98	-1.73% slower
irt_2pl/irt_2pl.stan	5.32	5.41	0.98	-1.55% slower
performance.compilation	87.52	85.7	1.02	2.08% faster
low_dim_gauss_mix_collapse/low_dim_gauss_mix_collapse.stan	7.71	8.44	0.91	-9.46% slower
pkpd/one_comp_mm_elim_abs.stan	21.24	20.26	1.05	4.58% faster
sir/sir.stan	95.6	91.75	1.04	4.03% faster
gp_regr/gen_gp_data.stan	0.04	0.04	1.0	0.47% faster
low_dim_gauss_mix/low_dim_gauss_mix.stan	3.05	3.27	0.93	-6.96% slower
pkpd/sim_one_comp_mm_elim_abs.stan	0.32	0.34	0.96	-4.71% slower
arK/arK.stan	1.79	1.83	0.98	-2.28% slower
arma/arma.stan	0.72	0.61	1.18	15.26% faster
garch/garch.stan	0.52	0.63	0.82	-22.5% slower
Mean result: 0.989059282904

---

Jenkins Console Log
Blue Ocean
Commit hash: 17729d3

---

Machine information

ProductName: Mac OS X ProductVersion: 10.11.6 BuildVersion: 15G22010

CPU:
Intel(R) Xeon(R) CPU E5-1680 v2 @ 3.00GHz

G++:
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin15.6.0
Thread model: posix

Clang:
Apple LLVM version 7.0.2 (clang-700.1.81)
Target: x86_64-apple-darwin15.6.0
Thread model: posix

[rok-cesnovar]

Anyone up for reviewing this?

[mitzimorris]

LGTM!