Energy scale and smearings application

Table of Contents

• Workflow to derive energy scales and smearings
• Use of energy scale and smearings in external frameworks
  • Energy scale and smearing corrections
• Tutorial
  • STEP1: time*eta dependent scale monitoring

Workflow to derive energy scales and smearings

Multistep procedure:

• first create a config file with the list of ntuples

Use of energy scale and smearings in external frameworks

The energy scale and smearings derived with ECALELF framework are provided as .dat files.

The module to read and get the energy scale and smearings for the different electron categories is EnergyScaleCorrection_class

Constructor of the class: 
  EnergyScaleCorrection_class(TString correctionFileName, 
                              TString smearingFileName=""); 

the method to get the correction is 
  float ScaleCorrection(int runNumber, bool isEBEle, double R9Ele, double etaSCEle, 
                        double EtEle,
                        int nPV, float nPVmean=0);

Energy scale and smearing corrections

Energy scale and smearings can be applied to the already existing ntuples. One can have different sets of energy scales and smearings, depending on

• the particular energy used for the invariant mass,
• the particular algorithm
• the number of steps in iterative procedure

For more details see the section dedicated to the energy scale and smearing derivation.

Each set of energy scale and smearing corrections is defined by a "type" name. You can add all the wanted set of scale and smearing correction files to the same validation config file, but only the one specified by the –corrEleType and –smearEleType options are used. The name of the trees of the corrections are defined as follows: <basename>_<type>, where <basename> is scaleEle for energy scale corrections and smearEle for energy smearings. This way you don't need to modify your validation config file each time you want to run with a different set of corrections, you just need to pass the appropriate type name at the command line.

Tutorial

From Calibration/

git fetch origin
git checkout -b updateFromFasanella origin/updateFromFasanella
scram b -j 16

cd ZFitter/
rm bin/ZFitter.exe
make

Create the directories:

mkdir test/{MC,dato}/{img,fitres} -p 

Have a look at the validation file containing data and MC ntuples: let's take data/validation/ICHEP_dataset_13_invfb.dat

Define your "single-electron" region in a specific dat file (data/regions/scaleStep2smearing_1.dat in this example)

Categorize in "di-electron" categories:

./bin/ZFitter.exe -f data/validation/ICHEP_dataset_13_invfb.dat --regionsFile=data/regions/scaleStep2smearing_1.dat --invMass_var=invMass_SC_corr --addBranch=smearerCat --saveRootMacro

At the end you'll have produced in tmp/ root files containing the categorization branch (called "smearerCat")

At the bottom of your validation file (data/validation/ICHEP_dataset_13_invfb.dat) paste the root files with the category branch:

d1      smearerCat_scaleStep2smearing_1 tmp/smearerCat_scaleStep2smearing_1_d1-ICHEP_dataset_13_invfb.root
d2      smearerCat_scaleStep2smearing_1 tmp/smearerCat_scaleStep2smearing_1_d2-ICHEP_dataset_13_invfb.root
d3      smearerCat_scaleStep2smearing_1 tmp/smearerCat_scaleStep2smearing_1_d3-ICHEP_dataset_13_invfb.root
s1      smearerCat_scaleStep2smearing_1 tmp/smearerCat_scaleStep2smearing_1_s1-ICHEP_dataset_13_invfb.root
s2      smearerCat_scaleStep2smearing_1 tmp/smearerCat_scaleStep2smearing_1_s2-ICHEP_dataset_13_invfb.root

(**Note: it would be wise to mv the smearerCat*.root files in another directory and never use tmp/ in real cases)

At this point you can fill your mass histograms divided in di-electron categories:

./bin/ZFitter.exe -f data/validation/ICHEP_dataset_13_invfb.dat --regionsFile=data/regions/scaleStep2smearing_1.dat --invMass_var=invMass_SC_corr --smearerFit --plotOnly

Now have a look at the histograms in test/dato/fitres/histos-scaleStep2smearing_1-Et_25.root

Quick view at the likelihood profiles:

./bin/ZFitter.exe -f data/validation/ICHEP_dataset_13_invfb.dat --regionsFile=data/regions/scaleStep2smearing_1.dat --invMass_var=invMass_SC_corr --smearerFit --plotOnly --profileOnly

Now have a look at the likelihoods in test/dato/fitres/outProfile-scaleStep2smearing_1-Et_25.root

STEP1: time*eta dependent scale monitoring

Fit data an MC mass distributions with a Breit-Wigner function convoluted with a crystal ball.

Have a look at the region file data/regions/scaleStep1.dat (there is no r9 requirement here, only eta) and the run range division in data/runRanges/run_range_for_tutorial.dat (there is only a run period here, for simplicity)

./bin/ZFitter.exe -f data/validation/ICHEP_dataset_13_invfb.dat --regionsFile data/regions/scaleStep1.dat  --runRangesFile data/runRanges/run_range_for_tutorial.dat --invMass_var invMass_SC_corr

Have a look at your mass spectra you just fitted

ls -rthl test/dato/fitres/
ls -rthl test/MC/fitres/
display test/dato/img/EB-absEta_0_1-runNumber_273158_273160-Et_25.pdf
display test/MC/img/EB-absEta_0_1-Et_25.pdf

Produce a latex table with the results of the fit

 ./script/makeTable.sh --regionsFile data/regions/scaleStep1.dat  --runRangesFile data/runRanges/run_range_for_tutorial.dat --outDirFitResMC=test/MC/fitres --outDirFitResData=test/dato/fitres --commonCut=Et_25 > test/dato/fitres/step1_ICHEP_dataset_13_invfb.tex

Produce a dat file with the scale shift

./script/tex2txt.sh test/dato/fitres/step1_ICHEP_dataset_13_invfb.tex | awk -F "\t" -f awk/recalibOutput.awk |grep -v '^%' > test/dato/fitres/step1_ICHEP_dataset_13_invfb.dat

Produce a summary plot with the scale shift behaviour over time:

./script/stability.sh -t  test/dato/fitres/step1_ICHEP_dataset_13_invfb.tex --outDirImgData test/dato/img/ -x runNumber -y peak