Skip to content

Latest commit

 

History

History
130 lines (81 loc) · 5.19 KB

README.md

File metadata and controls

130 lines (81 loc) · 5.19 KB
Raw

Muon Alignment Physics Validation

Checkout the physics validation code

  1. Follow the instructions in: https://github.com/cms-mual/Alignment/blob/CMSSW_9_2_6/README.md
  2. There you will also see how to take MuAlPhysicsValidation (basically just do https://github.com/cms-mual/MuAlPhysicsValidation.git in your favorite release).

The idea:

  1. You need to take RAWRECO SingleMuon events and reconstruct muons using the Muon Geometry and/or APEs you want to test.
  2. Run the Analyzer, in order to produce Ntuples with muon and dimuon information.
  3. Use the Ntuples to plot the reconstruction performance, overimposing the results obtained using several geometries and/or APEs.

Refitting

  1. cd MuAlPhysicsValidation/MuAlRefit

  2. cp -r MuAlRefit_Run2017B_RAWreco_2016Geom_01 CRAB_MYSTUDY; cd CRAB_MYSTUDY

  3. Change the crab configuration file: crab_data.py:

    Change the requestName and outLFNDirBase Change the Input dataset: inputDataset and the JSON file (if data) The python command you will execute: psetName inputFiles: inputFiles (CRAB cannot see external files if you do not specify them. If you want to use special APE, Geometries or GPR, you have to give here the path. CRAB will import them and copy them in the same place he runs cmsRun. So do not simply put the path name in the python code you will run.)

1d. One your crab file if fine, change the python code you will run. You can rename it with a better name:

mv refit_PromptGT_TrackerMay2016_MuAl_DT2016D6DOF_CSC2016D3DOF_cfg.py refit_PromptGT_MYNAME_cfg.py

1e. In refit_PromptGT_MYNAME_cfg.py you want to customize:

The GT.
The JSON file.
The APE, GPR, muon geometry and tracker geometry (all external files have to be declared in the CRAB file as inputFiles).

1f. Submit CRAB jobs:

cmsenv
voms-proxy-init -voms cms
source /cvmfs/cms.cern.ch/crab3/crab.sh
crab submit -c crab_data.py
crab status (and all the CRAB commands to check status, resubmit etc...)

1g. Once the jobs are finished, the output will be in EOS, as spdecified by outLFNDirBase location in the crab cfg file.

Analysis

2a. Change directory to MuAlAnalyzer:

cd ../MuAlAnalyzer

2b. Create a Working directory for this comparison (it could be the folder name on EOS from the previous step)

mkdir MuAlRefit_MYSTUDY

2c. Create a file list for the sample to be analyzed in Create_Input.sh; this will be the EOS folder that contains the Refit files and a python file that will create the file list. You will specify these on lines 2 and 3 of Create_Input.sh:

Folder="/store/group/alca_muonalign/<user>/<crab_output_location>"
fileTXT="MuAlRefit_MYSTUDY/MuAlRefit_MYSTUDY_list.py"

Then, execute the script.

bash Create_Input.sh 
cd MuAlRefit_MYSTUDY

2d. Now copy here the python file you will change to run your comparison muAlAnalyzer_Data_cfg.py.

cp ../muAlAnalyzer_Data_cfg.py .
cp ../createJobs.py .
python createJobs.py $FILELIST$ $N_JOBS$

FILELIST is the one you created and N_JOBS is the job splitting you want to use; recommend N_JOBS ~800 jobs and to remain in the range [500-900] jobs.

2e. Now you can submit the jobs:

source submit.sh

2f. hadd the output ROOT files into a single ROOT file and clean space.

hadd FINALFILE.root out_*root
rm -rf out_*root
rm -rf LSF*

Plotting

3a. Change directory to PerformancePlots:

cd ../PerformancePlots

3b. Use myPlot_2016E_3vs6DOF.py as a model for a new pyhton command to make the plots

cp myPlot_2016E_3vs6DOF.py myPlot_MINE.py

3c. You need to change:

maxEntries -> for test set to low value, for final plots set to -1 (all events)
samples -> change the path to the root files you created in the previous step
combineHistos -> decide what samples are shown in the same plot (E.g. [0,1] to show the first and second only).

3d. Launch the command:

python myPlot_MINE.py

Plotting with Performance Plots 2.0

This is an alternative procedure for step 3.

3'a. Change directory to PerformancePlots_2_0

cd ../PerformancePlots_2_0

3'b. Point Step1_make_2d_plots.py at the root file you hadded in the previous step, also specifying an output file:

python Step1_make_2d_plots.py ../MuAlAnalyzer/MuAlRefit_MYSTUDY/FINALFILE.root MuAlRefit_Step1_plots std -b

This will create a folder with some 2D plots in it named MuAlRefit_Step1_plots as well as a ROOT file MuAlRefit_Step1_plots.root. To make comparison plots with another analysis, you may repeat this step to create multiple root files to be used in step 3'c.

3'c. Edit Step2_makeProfiles.py to make comparison plots over groups of analysis datasets. You need to edit the following fields:

tot_fileList
tot_fileListName
tot_colors
Combinations
tot_outputFolderName

Specify the list of root files created in step 3'b in tot_fileList, labeling each dataset in tot_fileListName appropriately, and assigning colors and the grouping in tot_colors and Combinations, respectively. Finally, create an output folder to store all your comparison plots in the tot_outputFolderName field. Then, run the code:

python Step2_makeProfiles.py -b

Notes:

1 Input to Refit is full RECO, not ALCARECO!