SVTagging

SV Tagging for compressed SUSY search.

Workflow

1. Produce custom NANO AOD from MINI AOD (with extra SV variables)
2. Produce reduced ntuples from NANO AOD (with extra SV variables)
3. Produce efficiency input file (on condor)
4. Create efficiency plots

Produce Custom NANO AOD

This section describes how to produce custom NANO AOD from MINI AOD. These commands are specific to submitting crab jobs from CMS LPC.

For bash users, if not already done, add these lines to your ~/.bash_profile:

# CMS environment
source /cvmfs/cms.cern.ch/cmsset_default.sh
# CRAB environment
source /cvmfs/cms.cern.ch/common/crab-setup.sh

Then do

source ~/.bash_profile

Copy the following working area on CMS LPC:

# Working are on cmslpc for producing NANO AOD samples:
# /uscms/home/caleb/nobackup/KU_Compressed_SUSY/SV_Tagging/erichjs/cms_nano/CMSSW_10_2_22/src

mkdir ~/nobackup/SV_Tagging
cd ~/nobackup/SV_Tagging
rsync -az /uscms/home/caleb/nobackup/KU_Compressed_SUSY/SV_Tagging/erichjs/cms_nano/CMSSW_10_2_22 .

After setting up a working area, compile it:

cd ~/nobackup/SV_Tagging/CMSSW_10_2_22/src
cmsenv
scram b -j8
voms-proxy-init --valid 192:00 -voms cms

First, test that the code runs locally, for example:

cmsRun SUS-RunIISummer16NanoAODv7-cfg.py

If the code works locally, then you can submit to crab, for example:

crab submit -c crab-TTJets-DiLept-FullSim-2016.py

To check the status of crab jobs:

crab status -d <directory>

If there are failed jobs, they can be resubmitted like this:

crab resubmit -d <directory>

Crab Submit Script

You may use this crab submit script... though it should be tested first.

python submitTreesLPC.py -c <config> -d <dir> -f <datasets>

where <config> is a CMS config file, <dir> is a directory name for the crab submission, and <datasets> is a text file listing full dataset names with one dataset per line.

For example,

python submitTreesLPC.py -c SUS-RunIISummer16NanoAODv7-cfg.py -d TTJets-DiLept-FullSim-2016-Test -f Datasets-TTJets-DiLept-FullSim-2016.txt

Create Efficiency Plots

Set up working area

Login to a cmslpc Alma Linux 9 node (cmslpc-el9):

ssh -Y <username>@cmslpc-el9.fnal.gov

List available CMSSW versions and compatible scram architectures.

scram list CMSSW

Add this line to your ~/.bash_profile (or just run this command if you prefer):

export SCRAM_ARCH=el9_amd64_gcc12

If you edited ~/.bash_profile, logout and log back in to cmslpc-el9, or do:

source ~/.bash_profile

Then run these set up commands in a directory of your choice in ~/nobackup:

cd ~/nobackup/<working_area_path>
cmsrel CMSSW_13_3_0
cd CMSSW_13_3_0/src
cmsenv
git clone git@github.com:ku-cms/SVTagging.git
cd SVTagging

Copy the input data files (only required once):

rsync -az /uscms/home/caleb/nobackup/KU_Compressed_SUSY/CMSSW_10_6_5/src/SVTagging/data .

Create plots

Create efficiency plots with this command:

python3 python/make_sv_eff_plots.py

This script saves .pdf, .C, and .root files to an output directory.

You can copy these output files to your computer using rsync. For example, you may run a command with this syntax from a terminal on your computer:

rsync -az <username>@cmslpc-el9.fnal.gov:<path_to_output_directory> .