Everything to be run in environment of your CMSSW branch. As an example, see the environment file.

Tested in CMSSW_12_1_0_pre4.

Directory "extract_parameters"

Contains tools to run MC (direct simulation) and to extract new association cut strings from the simulation.
Execute ./run_simu to run a new simulation. On line 9, you can change the number of simulated events. 1E7 is a decent statistics, but it takes ~2h on LXPLUS.
analyze.cc is the code to fit the hit distributions - adjust the simulated statistics on line 229. Run make && ./analyze to execute it - it takes few seconds. The results can be investigated by opening analyze.root in the ROOT browser.

Directory "efficiency"

Contains tools to evaluate the efficiency of the near-far association. The efficiency is evaluated with the standard CMSSW module CTPPSProtonReconstructionEfficiencyEstimatorData. The scripts allow to either use the default association cuts settings (model "default") or define customs settings/models (see below)
One can adjust settings in run_multiple.
- Line 3: set the statistics: 1E4 runs for a minute, 1E5 runs about 10 minutes but gives smoother results.
- Line 5: list of models to test.
One can adjust the configuration template template_cfg.py.
- Lines 37 to 66: definition of the models.
- Lines 69 to 72: symmetriation of the settings. The current optics estimate is L-R symmetric.
Execute ./run_multiple to run the tests. The results can be investigated by opening simu/<model>/<statistics>/mc_eff_data.root, in particular by checking the efficiency histograms arm <arm>/exp prot <multiplicity>/nsi = 5.0/p_eff2_vs_x_N. Generally, the "eff2" estimate is considered better.