L1T Kalman Filter Upgrade

Code related to Kalman Filter algorithm upgrades of the CMS Level 1 Global Muon Trigger.

C++ algorithm found at cmssw: L1Trigger/Phase2L1GMT in this link: https://github.com/cms-sw/cmssw/tree/master/L1Trigger/Phase2L1GMT

Step 1 is to setup an up-to-date cmsenv, and run a cmsDriver command that runs the emulator and will produce a python file (--fileout argument), and a processed data file. The example below processes 10 events with -n for a DY sample through --filein argument and runs L1 and L1P2GT processes for the emulator through -s argument. Per these insturctions for the emulator recipe: https://cmsl1tofflinesw.docs.cern.ch/Phase2/phase2recipe/ look for a recent version of CMSSW with scram then setup an env with it. Since this was developing mostly the Kalman Filter code, I use cms-addpkg L1Trigger/Phase2L1GMT:

scram list CMSSW_16_1_X_*
cmsrel CMSSW_16_1_X_
cd CMSSW_16_1_X_*/src
cmsenv
git cms-init
git cms-addpkg L1Trigger/Phase2L1GMT
scram b -j 8
git checkout -b mybranch

cmsDriver.py -s L1,L1P2GT --conditions auto:phase2_realistic_T35 --geometry ExtendedRun4D121 --era Phase2C17I13M9 --eventcontent FEVTDEBUGHLT --datatier GEN-SIM-DIGI-RAW-MINIAOD --customise SLHCUpgradeSimulations/Configuration/aging.customise_aging_1000,Configuration/DataProcessing/Utils.addMonitoring,L1Trigger/Configuration/customisePhase2.addHcalTriggerPrimitives --filein /store/mc/Phase2Spring24DIGIRECOMiniAOD/DYToLL_M-50_TuneCP5_14TeV-pythia8/GEN-SIM-DIGI-RAW-MINIAOD/PU200_Trk1GeV_140X_mcRun4_realistic_v4-v1/2810000/5bed6cad-6cdb-4a5e-87af-08cc8c8e0ff8.root --fileout file:output_Phase2_L1T_test.root --python_filename run.py --inputCommands="keep *, drop l1tPFJets_*_*_*, drop l1tTrackerMuons_l1tTkMuonsGmt*_*_HLT" --mc -n 10 --nThreads 1 

That will create your run.py file with many processes and algorithms which make processing the events VERY slow. For Kalman Filter Algorithm development, changes to the C++ code happen regularly and waiting long durations to simply process 10 events is not feasible. To solve this problem, build a "reduced" run.py called run_reduced.py which is a fraction of the processes before. Only the necessary algorithms, processes and modules/collections are kept to study the KMTF quickly. One must run the previous cmsDriver.py command first since the new run_reduced.py will take the first output as input.

The idea is that the driver command is run once with all processes/collections, etc, then C++ algorithm modifications can be made Edits to the C++ algorithm can be made, then events reprocessed quickly with:

cmsRun run_reduced.py

If that works fine, the next step is to process a full dataset which uses submitChain.py together with run.py. The output data files which will ultimately be sent to your eos area configured inside run.py. The code is currently setup to process a Drell-Yan full dataset and points to my eos area in run.py. change these and also change the settings in submitChain.py to keep more collection modules or remove some depending on the goals:

voms-proxy-init --voms cms --valid 192:00
python3 submitChain.py

If a new plot needs to be made, add the function in interactive_plotters.py. Each function takes data and station, where data comes from the main event_loop function in main.py.

Example that plots Δz vs pT for station 1:

$ python3 -i interactive_plotter.py
  successful event loop. events: 100
  plot_delta_z_vs_pT(data, 1)

To add more events and choose to not convert k, z out of digital units then plot propagation from MB1 to vertex using the non-converted data:

$ python3 -i interactive_plotter.py
  successful event loop. events: 100
  data=event_loop(10000, False, False)
  plot_st1_to_vtx_vs_k1(data, False, False)

Or choose to do the same except with converted z and k:

$ python3 -i interactive_plotter.py
  successful event loop. events: 100
  data=event_loop(10000)
  plot_st1_to_vtx_vs_k1(data)