This module is used to derive reweighted fake factors and apply them to trees.
To make the relevant reweighted fake factor BDTs run this command.
python scripts/ff_reweight_ml.py --channel=mt --year=2017This will output the unnormalised fake factor BDTs into the BDTs folder as well as the normalisation json.
To apply the reweighted fake factors to a single root file run this command.
python scripts/add_ff_reweight_to_trees.py --input_location=/vols/cms/gu18/Offline/output/MSSM/mssm_2017_v10 --output_location=/vols/cms/gu18/Offline/output/MSSM/mssm_2017_v10_new --filename=SingleMuonB_mt_2017.root --channel=mt --year=2017 --splitting=100000 --offset=0The offset and splitting is used to fix memory problems. You may have to loop over offsets depending on the size of the tree. You will then need to hadd the relevant outputs with the following command.
hadd -f /vols/cms/gu18/Offline/output/MSSM/mssm_2017_v10_new/SingleMuonB_mt_2017.root /vols/cms/gu18/Offline/output/MSSM/mssm_2017_v10_new/SingleMuonB_mt_2017_*To batch add reweighted fake factors to a folder of root files run this command.
python scripts/batch_add_ff_reweight_to_trees.py --input_location=/vols/cms/gu18/Offline/output/MSSM/mssm_2017_v10 --output_location=/vols/cms/gu18/Offline/output/MSSM/mssm_2017_v10_newTo hadd the resulting output run this command.
python scripts/batch_add_ff_reweight_to_trees.py --output_location=/vols/cms/gu18/Offline/output/MSSM/mssm_2017_v10_new --haddReweighted with Boosed Decision Trees: article, paper, notebook, github repository