analysis

This repository contains the analysis scripts for event selection, plotting e.t.c

histogram making

Suppose you want to write to an output directory out using all of the cuts from reco_selection.py, for some root ntuple that is located at ../ntuples/input.root, but only the first 10k events. Then, you can run:

python3 analyzeDelphes_RecoHistograms.py -o out --all -n 10000 ../ntuples/input.root

The output will then be written to out/input.root with an accompanying log file hist.log that lists the selections that were used, along with their efficiencies. If you want to run on multiple input files, you can specify them separately: python3 analyzeDelphes_RecoHistograms.py -o out --all -n 10000 ../ntuples/input.root ../ntuples/input2.root Or you can specify ../ntuples/*.root. Or, if you have madgraph output directories that have a single run (Events/run_01/unweighted_events.root), you can specify the madgraph directory instead of unweighted_events.root, so that the correct name of the process is used for the output root file. If you want to force the script to overwrite an existing output file, you can add -f If you want to specify a subset of the cuts, you can specify them with -c, e.g., -c 0,1,3 will use the 0, 1, and 3 cuts, given the order in reco_selection.py The output histograms will be automatically scaled according to the luminosity in lumi.yaml and cross_section.yaml. lumi.yaml maps from energy in TeV (3, 6, 10, 30) to integrated luminosity. You can specify the energy with -e (default, 6) cross_section.yaml maps from the input process name to a cross section. The input process name is inferred from the stem of the input root file, or from the name of the directory, if a madgraph directory is specified. Cross sections should be in fb, luminosities should be in fb^-1 If you specify multiple files, they will be run in parallel. Progress will be shown with a rich progress bar, if you have rich installed in your python environment.

Cut values

You can specify cut values to analyzeDelphes_RecoHistograms.py using the variable names specified in reco_selection.py. For example, add the following to set the pt min to 200 GeV:

--pt_min=200

Note the =!

Condor submission

To submit a histogramming job to condor, you can use condor_submit.py (assuming you have your environment set up for condor submission, and that you have set up your environment such that the usual histogram scripts are runnable).

Example:

python condor_submit.py --python_args "indir -f -o outdir -n 1000 -c 1,4 --ncpus 1"

To submit many jobs, you can write, e.g., a simple bash script:

for PT_MIN in 100 200 300 400
do
    python condor_submit.py --python_args "indir -f -o outdir -n 1000 -c 1,4 --ncpus 1 --pt_min=$PT_MIN"
done

where indir and outdir are replaced with absolute paths. Note that indir is a positional argument. If submitting multiple jobs in a loop, make sure to use a different output directory for each job, to avoid overwriting your results.

To test that your jobs will work correctly, you can try passing the --python_args to analyzeDelphes_RecoHistograms.py directly, to check that the job starts without error. For example:

python analyzeDelphes_RecoHistograms.py indir -f -o outdir -n 10 -c 1,4 --ncpus 1 --pt_min=$PT_MIN

where we have -n 10 to ensure a fast job. You can check that this runs without error, and if so, submit larger condor jobs as described above.