LFV Z Analysis

Description:

A modularized ntuplizer that reads standard NanoAOD and provides a slimmer version with cleaned objects and complex variables. The output is again a standard NanoAOD format file, and can be processed with any compatible histogramming tool. It is based on NanoAOD tools.

The modules are implemented in NanoAODTools/python/postprocessing/modules/CUmodules/:

• GenAnalyzer.py: Scans the GenParticle collection and creates the configured decay. eg Z->mue is 23->13,11. For cascade decays we can add several modules together.
• GenCount.py: Counts the number of processed events, storing the total number and the negative number of events in a ROOT histogram in the event output
• GenLepCount.py: Counts the number of primary leptons in the event, useful for separating Z->tau+tau from e+e/mu+mu and also for identifying WW/ttbar->tau+tau
• GenZllAnalyzer.py: Scans the GenParticle collection and attempts to find a Z->ll' decay. If a Z boson isn't found, it will look for exactly two primary charged leptons
• GenIdenticalMothersDiscriminator.py: In case of two identical particles this can be used to disentangle them.
• GenRecoMatcher.py: Takes genparticles as inputs and matches them with reco
• HTSkimmer.py: Creates the HT variable based on configuration.
• JetLepCleaner.py: Removes overlapping jets (or leptons) with leptons (or jets).
• JetSkimmer.py: Skims the jet collection
• LeptonSkimmer.py: Skims the lepton collections
• SelectionFilter.py: Applies some selection logic to filter events to different channels (e.g. e+mu vs mu+tau) and filter out events that don't satisfy any channel selections

Building instructions:

cmsrel CMSSW_10_6_29
cd CMSSW_10_6_29/src
cmsenv
git clone https://github.com/michaelmackenzie/Z_LFV_analysis PhysicsTools
scram b -j8

Run Locally:

Process a NanoAOD file using the LFVAnalyzer.py:

cd $CMSSW_BASE/src/PhysicsTools/NanoAODTools/
DATASET="<dataset to analyze, e.g. /SingleMuon/Run2016H-02Apr2020-v1/NANOAOD>"
FILE=`das_client -query="file dataset=${DATASET} <instance=prod/phys03 if needed>" | head -n 1`
xrdcp -f ${FILE} ./NanoAOD.root
python python/LFVAnalyzer.py NanoAOD.root <"data", "MC", or "Embedded"> <"2016", "2017", or "2018">
root.exe tree.root

Run on CRAB:

cd $CMSSW_BASE/src
cd PhysicsTools/NanoAODTools/crab/
python multi_submit.py

Run on LPC:

Submit jobs:

cd $CMSSW_BASE/src/PhysicsTools/NanoAODTools/condor/
#configure submitBatch_Legacy.py output_dir user name, samplesToSubmit, and doYears lists
./submitBatch_Legacy.py

Check job statuses:

cd $CMSSW_BASE/src/PhysicsTools/NanoAODTools/condor/
./query_grid.sh [--summary for less information]

When the batch job finishes, check job success:

cd $CMSSW_BASE/src/PhysicsTools/NanoAODTools/condor/
./check_batch_job.sh batch/<job directory>/
#if there are failed jobs, resubmit them with:
./check_batch_job.sh batch/<job directory>/ --resubmit
#use --help (-h) for more options

Merge the output files into single file datasets:

cd $CMSSW_BASE/src/PhysicsTools/NanoAODTools/condor/
python prepare_batch.py nano_batchout/<job directory>/ <--mc_data ["MC" or "data"]> <--veto [veto tag]> <--tag [to process tag]> <--dryrun> <--year [e.g. 2016]>
#use --help (-h) for more options