jrotter2/emtfanalyzer

Analysis tools for EMTF

EMTFAnalyzer

Analysis tools for EMTF.

Setup

The EMTFAnalyzer is designed to run on lxplus, uses crab for job submissions, and pulls input files from EOS. The version of CMSSW is critical since trigger primitives and other aspects of the EMTF will change.

cmsrel CMSSW_12_1_0_pre3
cd CMSSW_12_1_0_pre3/src/
cmsenv
git cms-init
git cms-rebase-topic dildick:CMSSW_12_1_X-EMTF-CCLUT
git clone [email protected]:dildick/EMTFAnalyzer.git
scram b -j 4

Structure

The EMTFAnalyzer repository has two main directories, AWBTools which has not been kept upto date, and NTupleMaker which is used for Run3 BDT Training Preperation. NTupleMaker is then divided into crab, src, python, and test.

The crab directory has not been kept upto date and most scripts have been replaced by scripts in test.

In the src directory are Branch and Matcher helper classes.

In the python directory are the actual NTuple process definitions.

In the test directory are where the crab configurations and PSets are for the jobs.

Code

The EMTFAnalyzer analyzes EDM collections and dumps them into flat ntuples. It also makes links between EMTFTracks and EMTFHits through indices. The EMTFAnalyzer was updated by Sven to take into account GEM hits from GE1/1 for Run-3. And it was also updated for Run-3 CSC trigger primitives when the CCLUT algorithm is set in OTMBs and/or TMBs.

A GEMEMTFMatcher plugin was written to match EMTF tracks to GEM clusters, convert them to EMTFHits, and insert them into the EMTFHitCollection. This plugin is useful when you want to match EMTF tracks with GEM hits, but when the CMSSW track finder does not yet do this for you by default. The matching is based on a delta-phi match between ME1/1 and GE1/1. In future GEM-EMTF studies you may want to test this module before assuming it works out of the box. Part of the code was written by Denis Rathjens (TAMU).

Steps for Generating NTuples

1.) The first step to generating NTuples for BDT Training is to generate a MC sample. This is done using NTupleMaker/test/crab_SingleMuFlatOneOverPt1To1000GeV_GEN_SIM.py which will get saved to EOS - this step is only the generation and simuluation of the MC.

2.) The second step is to run the emulation of this MC Sample to digitization and trigger primitives is done using NTupleMaker/test/step2_DIGI_L1.py which will also be saved to EOS.

3.) Then the third step is to run the NTuplization step which will generate NTuples based on NTupleMake/python/FlatNtuple_cff.py and will be sent to crab using NTupleMaker/test/Run3_MC_NTuple.py.

4.) Lastly, these NTuples need to be hadded together to be used as input into the BDT.