latinos/fakeratemeasurement

Small standalone script to calculate both the fake and prompt rates using the tight-to-loose method. The script is divided into different steps.

Important note

The code that can be found in the directory called "Fake-rates-code" is an alternative option to compute the fake rates. Both options have been completely synchronized in June 2019.

Step 1

It is possible to run by hand the script runNanoFakes.C on one particular file,

root -l -b -q 'runNanoFakes.C("2017", "nanoLatino_DYJetsToLL_M-50__part0")'

The normal thing to do is to batch submit several samples at once. The batch submission can be done reading the samples directly from a directory.

python submitJobs.py -d <full path of the directory> -y <year>

To change the year to produce the fake rates, the argument -y has to be given and the following changes have to be made. First, the electron/muon has to be changed at the bottom of the nanoFakes.h file, and then the triggers might need to be changed (to move from ele12 to ele8, for example). The prescale weigth should be updated automatically though.

Check job status

It is possible to check the status of the jobs by using the command

condor_q      
watch -n10 condor_q

Step 2

An hadd is necessary once all the jobs are done.

cd results
hadd hadd_data.root  nanoLatino_*_Run201*.root
hadd hadd_wjets.root nanoLatino_WJetsToLNu*.root
hadd hadd_zjets.root nanoLatino_DYJetsToLL*.root

Two different options are then available.

getFakeRate.C

This is a small script allowing the user to compute both the fake and prompt rates, for both muons and electrons. The output is a set of png histograms.

getLumiEff.C

This script uses the ZRegion defined in nanoFakes.C in order to calculate a basic data/MC ratio within the peak of the Z, to make sure that the results given by getFakeRate.C make sense.