andrzejnovak / TnPSF

W TagAndProbe rhalphalib implementation

Setup combine and rhalphalib

Follow official combine instruction or setup with conda using this branch for python2 or main for python3.

Setup rhalphalib and then

git clone https://github.com/andrzejnovak/TnPSF.git
cd TnPSF

Generate variations

For each root file generate variations (only matched - catp2).

python scalesmear.py -i templates/ref17/wtemplates_n2cvb.root  --plot
python scalesmear.py -i templates/ref17/wtemplates_cvl.root  --plot

New files will have a name convention of <input_name>_var.root.

Generate combine/rhalphalib workspace and fit

python sf.py --fit single -t templates/ref17/wtemplates_n2cvb_var.root -o FitSingle
cd FitSingle

or for two-cut setup:

python sf.py --fit double -t templates/ref17/wtemplates_n2cvb_var.root --t2 templates/ref17/wtemplates_cvl_var.root -o FitDouble
cd FitDouble

and run the fit

combine -M FitDiagnostics --expectSignal 1 -d model_combined.root --cminDefaultMinimizerStrategy 0 --robustFit=1 --saveShapes --saveWithUncertainties --rMin 0.5 --rMax 1.5

To make plots from FitDiagnostics output, run within the fit folder:

python ../results.py --year 2018