Analysis Framework

Workflow

Below is a short description of the workflow of the analysis framework:

• Production of Patuples from data and MC AOD or RECO samples.
• Production of small ntuples from Patuples.
• Analysis of ntuples.

Production of Patuples

We use the patuples produced by the B2G group. The recipe and location of the patuples are given in this page.

Production of Analysis ntuples

Below is the recipe to install the analysis ntuple code:

cd CMSSW_5_3_3/src
cmsenv
cvs co -r V00-03-03 -d Yumiceva/Top7TeV UserCode/Yumiceva/Top7TeV

Run a test job:

cd Yumiceva/Top7TeV/test
cmsRun TuplesFromPAT.py

Reference selection for objects:

• • B2G twiki
  • TOP twiki

Description of package:

• For data:
  • Location of JSON files:

ls /afs/cern.ch/cms/CAF/CMSCOMM/COMM_DQM/certification/

• • Definition of primary datasets (PD) can be found here
  • Triggers:

HLT_IsoMu24_eta2p1_v* appropriate for SingleMu PD
HLT_IsoMu17_eta2p1_TriCentralPFJet30_v* appropriate for HadMu PD

Preparing crab files

WORK IN PROGRESS

• Define the samples you want to run in python dictionaries within the script setupProd.py

• For data, you need the dataset name, and json file.

• Configuration files for muons:

cd Yumiceva/Top7TeV/test
python setupProd.py /uscmst1b_scratch/lpc1/cmsroc/yumiceva/4tops2013 muons

Create, Submit and check job status.

• Create and submit jobs:
  • After the script setupProd.py has been run for either muons or electrons, this script create two text files with the crab commands to create and submit jobs.

• Check status of jobs:

python statusProd.py production/ status

• Get output:

python statusProd.py production/ getoutput

• Get json lumi report:

python statusProd.py production/ report

* Get merged output ntuple:

python statusProd.py production/ hadd

Luminosity calculation

Follow the scripts in this recipe

python statusProd.py production/ lumi

Location of ntuples

Data ntuples

MC ntuples

• Signal:

• Background:

Analysis of ntuples

Recipes

• Install the analysis software:

cd CMSSW_5_3_3/src
cmsenv
cvs co -r V00-01-00 -d Yumiceva/TreeAnalyzer UserCode/Yumiceva/TreeAnalyzer 

* Setup ROOT to run PROOF:

• • Copy to your area and modify according to your needs the ROOT setup files:

cp ~yumiceva/RooAlias.C ~/.
cp ~yumiceva/RooCMS.C ~/.
cp ~yumiceva/RooLogon.C ~/.
cp ~yumiceva/.rootrc ~/.

• • Edit rootrc file to point the PROOF sandbox to your scratch area. Change the line:

Proof.Sandbox: /uscmst1b_scratch/lpc1/cmsroc/yumiceva/proof

• Run test analysis code:

cd Yumiceva/TreeAnalyzer/test
root -l -b -q loadLibraries.C runSimpleAnalysis.C\(\"data\"\)

• The histograms will be store in a ROOT file called "results_data.root".

Analysis Code

• The analysis code is run using the PROOF feature in ROOT. The code can be found in the files: SimpleAnalyzer.h and SimpleAnalyzer.C
• The code is setup and compiled using the script "runSimpleAnalysis.C".

Debugging and developing analysis code

• Enable log files:

cd Yumiceva/TreeAnalyzer/test
root -l loadLibraries.C /uscms/home/yumiceva/lpc1/4tops2013_ntuples_v1/muons/Run2012A-recover-06Aug2012-v1.root
root [2] .x runSimpleAnalysis.C("data",1)