jhgoh / tzwi

Aiming to have full analysis with NanoAOD

TZWi

Yet another CMS NanoAOD tools focusing on specific analyses.

• Management area: https://app.asana.com/0/1115311953973868/list

We'd like to cover:

• ttbar dilepton analysis
  • inclusive cross section, differential cross section
  • ttbar+bbbar cross section ratio to ttbar+jets
• ttbar rare decays
  • t->qZ, ttbar and single top

See also

• https://twiki.cern.ch/twiki/bin/view/CMSPublic/WorkBookNanoAOD
• NanoAOD dataformat descriptions
  • 2016 samples: https://cms-nanoaod-integration.web.cern.ch/integration/master-102X/mc80X_doc.html
  • 2017 samples: https://cms-nanoaod-integration.web.cern.ch/integration/master-102X/mc94Xv2_doc.html
  • 2018 samples: https://cms-nanoaod-integration.web.cern.ch/integration/master-102X/mc102X_doc.html

Installation

#unset SCRAM_ARCH ## just in case if you need this...
cmsrel CMSSW_10_2_20_UL
cd CMSSW_10_2_20_UL/src
cmsenv
git-cms-init
git cms-merge-topic cms-nanoAOD:master-102X
git checkout -b nanoAOD cms-nanoAOD/master-102X
git clone https://github.com/cms-nanoAOD/nanoAOD-tools.git PhysicsTools/NanoAODTools
git clone https://github.com/cms-kr/TZWi
scram b -j

(Optional) Customized NanoAOD production

cd TZWi/NanoAODProduction/test
./generateConfig.sh ## This will produce 3 sets of 3 cfg files...
crab submit...

List up NanoAOD samples

Update sample list, produce file lists

tzwi-updatedataset $CMSSW_BASE/src/TZWi/NanoAODProduction/data/datasets/NanoAOD/2016/*.yaml
tzwi-updatedataset $CMSSW_BASE/src/TZWi/NanoAODProduction/data/datasets/NanoAOD/2017/*.yaml

Run postprocessors

Assume we are working at KISTI Tier2/3 and cms-kr/hep-tools package is installed.

./01.1_submit.py

Wait for the jobs to be finished, check output files, resubmit failed jobs.

Tip to list up failed job and resubmit them:

./01.2_checkFailed.sh

You can process failed ones manually:

cat failed.txt | sed 's;nano_postproc.py;;g' | xargs -P$(nproc) -L1 nano_postproc.py

Tip to extract all ntuples:

find *NANOAOD*/ -name 'result_*.tgz' | awk '{print "xzf "$1" ./ntuple"}' | xargs -L1 -P$(nproc) tar

Make histograms

This step will draw all histograms including systematics variations using maximum 20 CPUs in parallel.

./02_make_histograms.py

To the plotting steps

Run the followings

./03_scalemerge.py
./04_drawPlots.py

Then you will have plots in "plots" directory.