Skip to content

Latest commit

 

History

History
90 lines (65 loc) · 2.58 KB

README.md

File metadata and controls

90 lines (65 loc) · 2.58 KB
Raw

Run3BparkingHLTStudies

The instructions below have been verified with python3 (CMSSW_12_0_X onwards).

1. log-in to lxplus

cmsrel CMSSW_12_1_0
cd CMSSW_12_1_0/src
cmsenv

2. Setup package

git clone [email protected]:gitytakahas/Run3BparkingHLTStudies.git
cd Run3BparkingHLTStudies

3. (Can be skipped) derive efficiecny for single-mu after folding in analysis efficiency.

cd $CMSSW_BASE/src/Run3BparkingHLTStudies/single-mu/
python3 draw_roc.py --pr --weight

This will print out relevant numbers to be later used to make the ROC curve. If you just need L1 times HLT trigger eff. just remove --weight option. These numbers have been already embedded into the draw_roc.py script, which is the one used for drawing the ROC curve.

4. (Can be skipped) create efficiency map for di-e after folding in analysis efficiency.

cd $CMSSW_BASE/src/Run3BparkingHLTStudies/
python3 compare_eff.py --weight

This will create the efficiency map to be later used to make the ROC curve. If you don't want to fold in analysis efficiency, you can just remove --weight option. Obtained maps are already stored in eos so this step can be skipped unless you want to create the new one. If you want to make comparison plots for several basic distributions at the HLT level, just add --plot option.

5. (Can be skipped) create rate map for di-electron trigger

cd $CMSSW_BASE/src/Run3BparkingHLTStudies/
python3 compare_rate.py

This will create the HLT rate map (as a function of number of pileup) to be later used to make the ROC curve. Obtained maps are already stored in eos so this step can be skipped unless you want to create the new one.

6. Create ROC curve

cd $CMSSW_BASE/src/Run3BparkingHLTStudies/
python3 draw_roc.py --weight

The script will loop over relevant npu:

  • pu = 17 (0.6E34)
  • pu = 25 (0.9E34)
  • pu = 30 (1.1E34)
  • pu = 36 (1.3E34)
  • pu = 42 (1.5E34)
  • pu = 48 (1.7E34)
  • pu = 56 (2.0E34)

and create the ROOT files containing each ROC curve, to be later used in the next step (estimate.py). If you want to have different npu, you can just edit the last few lines of draw_roc.py.

5. Estimate number of Kee candidates

Provides various menus:

  • original menus in AN-21-160 v3
  • "hybrid" menus using L1 prescales
  • menus paramterised vs NPU and Nbunches
  • menus paramterised vs NPU and Nbunches and using Run 3 observed rates (default)

... and various tables of useful metrics.

python3 metrics.py

Provides estimates per LHC fill (based on a pseudo LHC luminosity profiles):

python3 estimate.py