The main agenda of this meeting is to study the E906 simulation setup and compare it with the E1039 setup. The E906 simulation setup is presented by Kenichi in:
https://seaquest-docdb.fnal.gov/cgi-bin/private/ShowDocument?docid=5431
What we learn so far:
- Our E1039 generator is phase space and did not include all physics related to the Drell-Yan production, for example there is Drell-Yan angular distribution that has not been included in our generator.
- Therefore, we need to improve our generator (or simulation) in general. The things that we need to improve are summarized in the email exchange between me and Haiwang below:
As we discussed last week. As the data taking approaching, there are many important tasks related to software/simulation emerging. We could really use the help from the UVA group. And thanks a lot volunteering.
We discussed two major tasks during the analysis meeting last week.
1) Port the generator tools from DPSim to e1039-core. This includes the event weighting mentioned by Kenichi and the generator level speed booster cut mentioned by Kun. My impression is we could study the generator functions in this file by Kun: https://github.com/liuk/DPSim/blob/dev/src/DPPrimaryGeneratorAction.cxx
And port the functions to e1039-core as generators like this: https://github.com/E1039-Collaboration/e1039-core/blob/master/simulation/g4main/PHG4ParticleGun.cc
This task could be relatively short termed, about 1-2 month.
2) Fast simulation for background study.
This task could be approached in 3 steps in my opinion:
1, Using minimum bias event to generate a 5D (x, y, px, py, pz) phase space map for each particle species at the downstream surface of FMag together with probability of each particle species. This could be done using macros like this: https://github.com/E1039-Collaboration/e1039-analysis/blob/master/PrgTrkDev/Fun4PatternDB.C This will output a root file "trk_eval.root". It contains information at St1 NOT surface of FMag at this time, we could modify that later.
2, A particle generator that is able to sample from the 5D phase space maps generated from the 1st step. This could be done making a particle generator similar like this: https://github.com/E1039-Collaboration/e1039-core/blob/master/simulation/g4main/PHG4ParticleGun.cc#L45 but being able to sample in 5D space.
3, Tuning the simulation to match data using various observables.
We could start from study those as a start.
This task could start with a short term feasibility study, ~1 month. If we decided to do this. This would be a long term task in my opinion.
Andrew also expressed interests in this study. So maybe someone from the UVA group could collaborate with Andrew on this task.
These are just my initial thinking on these two topics.
What do you think about this?
We could discuss details in the simulation meetings or the analysis meetings.
Cheers,
Haiwang