TIDAL

git clone --recurse-submodules [email protected]:Ksavva1021/TIDAL.git

CP Analysis Instructions

Code to produce datacards is in the Draw directory (make sure to setup Draw utilities).

The config for the CP analysis is:

Draw/scripts/cpdecay_datacards.yaml

To create the datacards run:

python Draw/scripts/makeDatacards.py --config Draw/scripts/cpdecay_datacards.yaml --batch

Once all jobs are complete, hadd them for combine with:

python Draw/scripts/hadd_cp_datacards.py -i Draw/Plots/Path/You/Chose/*/*/*/*.root -o Draw/Plots/Path/You/Chose/added_histo.root

Setup Instructions

Note: SVFit is a separate package so, install and load in a clean terminal

Install the environment:

micromamba env create -f environment.yml

Setting up Draw Utilities (Separate from SVFIT, clean terminal):

Setting up Draw utilities (Multidraw):

micromamba activate TIDAL
source setup_multidraw.sh
pip install -e .

Using Draw:

Activate the environment and load ROOT:

micromamba activate TIDAL
source load_package.sh # Select Option 1

Setting up SVFIT:

micromamba activate TIDAL
source setup_svfit.sh

To check that SVFIT was set up properly. Run interactively in python3:

from TauAnalysis.ClassicSVfit.wrapper.pybind_wrapper import FastMTT
x = FastMTT()

Using SVFIT:

Activate the environment and load ROOT+SVFIT:

micromamba activate TIDAL
source load_package.sh # Select Option 2

For the time being SVFIT is only set up for the $\tau_h \tau_h$ channel. There are extra command-line arguments inside the scripts\run_svfit.py script but, you only need --source_dir and --use_condor

Example Command - This will submit jobs to the batch (again reminder only for $\tau_h \tau_h$ channel at the moment, skips the rest):

python3 scripts/run_svfit.py --source_dir /vols/cms/ks1021/offline/HiggsDNA/IC/output/test/Run3_2022/ --use_condor

Output file is called svfit.parquet and contains run, lumi, event, svfit mass, svfit error. These are saved in the directory where the merged.parquet file is found e.g.

/vols/cms/ks1021/offline/HiggsDNA/IC/output/test/Run3_2022/tt/DYto2L_M-50_madgraphMLM/nominal/svfit.parquet