Skip to content

README_STEAMRateStudy_Example.md

Latest commit

 

History

History
193 lines (121 loc) · 4.9 KB

README_STEAMRateStudy_Example.md

File metadata and controls

193 lines (121 loc) · 4.9 KB

Example: STEAM HLT Rate Study

Please refer to the STEAM page for most up-to-date instructions.

Step 1: Setting up the environment

cmsrel CMSSW_15_1_0_patch1
cd CMSSW_15_1_0_patch1/src
cmsenv
scram build -j 4
git clone https://github.com/cms-steam/SteamRatesEdmWorkflow.git
cd SteamRatesEdmWorkflow/Prod/

Step 2: Generating the EDM ROOT files

hltGetConfiguration /dev/CMSSW_15_1_0/GRun --full --offline \
 --no-output --data --process MYHLT --type GRun \
 --prescale 2p0E34+ZeroBias+HLTPhysics \
 --globaltag 150X_dataRun3_HLT_v1 --max-events -1  > hlt.py
edmConfigDump hlt.py > hlt_config.py
voms-proxy-init --voms cms --valid 168:00
user=asimsek; cert=$(find /tmp/x509up_u* -user "$user" -print -quit); cp "$cert" "/afs/cern.ch/user/${user:0:1}/$user/private/"

Test Locally:

Tip

You can use createFileList.py script to create a list from the given datasets and run numbers.

cmsRun run_steamflow_cfg.py nEvents=1 &> log.txt &

Send to Condor:

python3 cmsCondorData.py run_steamflow_cfg.py /afs/cern.ch/work/a/asimsek/private/TriggerRates/CMSSW_15_1_0_patch1/src/ /eos/cms/store/user/asimsek/NPSTriggerRates/01Nov2025 -n 1 -q workday -p /afs/cern.ch/user/a/asimsek/private/x509up_u75207
./sub_total.jobb

Tip

You can check your jobs with condor_q command.

Step 3: Running the counting script

cd $CMSSW_BASE/src/SteamRatesEdmWorkflow/Rates/

Warning

Don't forget to change the inputFilesDir, cmsswDir, json_file, maps, and flavour variables inside the config_makeCondorJobsData.py.
Set maps = "nomaps" to use the Draw.py script.
Although the jobs on condor is usually fast, they might take a bit longer than maps = "nomaps" on HTCondor (flavour = "tomorrow")

python3 config_makeCondorJobsData.py

Test Locally:

source Jobs/Job_0/sub_0.sh

Send to Condor:

./sub_total.jobb

Step 4: Misc

Calculate HLT_PS:

Note

Find the dataset and HLT prescale values from OMS
All 8 Dataset_EphemeralHLTPhysics* dataset has the same PS value (PS=8) — one of them enough.
Prescale of the HLT_EphemeralPhysics_v is 220.

HLT_PS = (PS) x (HLT_EphemeralPhysics)
HLT_PS = 8 x 220
HLT_PS = 1760

Calculate Lumi, N_LS, Avg. Pileup:

cd $CMSSW_BASE/src/SteamRatesEdmWorkflow/Rates/
export PATH=$HOME/.local/bin:/cvmfs/cms-bril.cern.ch/brilconda/bin:$PATH
python3 -m pip install --user brilws

brilcalc lumi -u 1e33/cm2s -b "STABLE BEAMS" -i Json/json_398183.txt

Warning

The normtag (normtag_PHYSICS.json) for your JSON may not be available yet!
If available, add: --normtag /cvmfs/cms-bril.cern.ch/cms-lumi-pog/Normtags/normtag_PHYSICS.json
use -u /fb instead of -u 1e33/cm2s when needed.

Avg Instantaneous lumi:

Note

Divide the total delivered luminosity (1e33/cm2s) by the number of LS (nls) in your JSON.
This is your average input luminosity (Lumi_In), which you should write into the config file.
The final output needs to be presented in 1e34 /cm^2/s.

Lumi_In = 9357 / 427  ## 1e33 /cm^2/s
Lumi_In = 2.19        ## 1e34 /cm^2/s

Deadtime:

The average deadtime can be calculated using the ratio of total recorded luminosity and total delivered luminosity.

Avg_DeadTime = (1 - (total recorded lumi/total delivered lumi))*100
Avg_DeadTime = (1 - (8929.191427781/9357.778527154))*100
Avg_DeadTime = 4.58%

Avg Pileup:

Important

The recommended cross section for Run 3 = 69.2 mb (69.2e-27).
However, use 80mb (80e-27) as recommended here — matches with the OMS PU.
Find Fill Number from OMS Run Report - cms_run=11190
Find number of bunches OMS Fill Report - NBunches = 2448
LHC Revolution Frequency = 11245 Hz

pileup = (Lumi_In x pp_inelastic_xsec x NBunches) / (LHC_Frequency)
pileup = ((2.19e34 cm^-2 s^-1) x (80e-27 cm^2)) / (2448 x 11245 s^-1)
pileup = 63.64

Step 5: Merging and scaling

Warning

Don't forget to change the lumi_in, lumi_target, hlt_ps, and maps variables inside the config_mergeOutputsData.py.
Lumi_In, HLT_PS, Deadtime, etc. calculation steps are given below.
Set maps = "nomaps" to use the Draw.py script.

python3 config_mergeOutputsData.py