Skip to content

Latest commit

 

History

History
23 lines (17 loc) · 2.07 KB

README.md

File metadata and controls

23 lines (17 loc) · 2.07 KB

Tracking Resolution

This a repository compiling all the tracking pT resolution work. Please, use CMSSW 10_2_7 for the tests, which is the version that works so far.

To produce the histograms, there are a few steps needed. Inside the TrackingResolution/TrackingResolution/ folder you should

  • Compile the modules scram b -j 8
  • Amputate the tracks
    • this is performed by the command cmsRun python/ClusterSurgeon.py outputFile=OUTPUT_FILE_NAME.root (the input files for this step are ZMM RelVal samples, so far)
  • Run the reRECO file: cmsRun python/reRECO_new.py inputFiles=file:OUTPUT_FILE_NAME.root outputFileName=reRECO_OUTPUT_FILE_NAME.root
  • Run the DQM file: cmsRun test/Tracker_DataMCValidation_cfg.py inputFiles=reRECO_OUTPUT_FILE_NAME.root outputFile=DQM_reRECO_OUTPUT_FILE_NAME.root
  • Run the Harvest file: cmsRun test/Tracker_DataMCValidation_Harvest_cfg.py inputFiles=DQM_reRECO_OUTPUT_FILE_NAME.root
  • The output file with the histograms will be named something similar to DQM_V0001_R000000001__Global__CMSSW_X_Y_Z__RECO.root

Tests are being made to run the procedure in CMSSW_12_0_0_pre3. Even though it is possible to execute the tracks shortening, re-reconstruction and validation using the RelValZMM file from CMSSW_10_2_5, when trying to use files from CMSSW_12_0_0_pre3, it accuses that some modules are missing when they are defined into the task to be executed.

A way to more easily add modules into the re-recontruction step is work in progress so far.

Necessary checks

  • Why the number of reconstructed tracks (before selection) is extremely higher than the number of good tracks to reconstruct? See where and how this tracks appear in the detector
  • Number of hits in reconstructed tracks (after selection) is a bit different from the number of hits provided in the dataset to be reconstructed. Compare reco hits with hits from dataset

Improvements

  • Instead of clonning the relevant paths to reconstruct tracks, pass only the hits to the InitialStep and use the same other modules for reconstruction, Faster and more reliable (?)