• Riccardo told me to use CMSSW_1_8_0_pre9 for 'NoDrift' check
  • No entiry problem of 'LinearDrift' in CMSSW_1_8_0_pre9
    • checked if Martijn's modification affect it <-- No effect.


  • tried to read cosmic data in CMSSW_1_8_0_pre9
    • 'NoDrift' ran well but chi2 distribution is very different
    • 'LinearDrift' ran well but no entry is filled into histogram


Martijn sent e-mail about commiting the change of 'NoDrift' and testing it in CMSSW_1_8_x.


gave a talk in DTDPG meeting http://indico.cern.ch/conferenceDisplay.py?confId=29424


comments from the lepton meeting

  • definition of phi? <-- recognized by event display
  • why NoDrift can't find the events in positive phi direction?
    • probably cosmic muon should be reconstructed in downward. It is strange that LinearDrift reconstruct it in +phi direction
  • check the condition of trigger of GREN
  • check if I compared the chi2 for the same track in LinearDrift and NoDrift <-- Yes, it was same chi2
  • check individual DT info such as time, etc
  • what is green line in the event display?


  • compared old parameter and new parameter in NoDrift algorithm. That is, parameters are changed like following in RecoLocalMuon/DTRecHit/data/DTNoDriftAlgo_CosmicData.cfi
    • fixedDrift to 0.0,
    • hitResolution to 0.6 cm (= 2cm / sqrt(12))


  • read the slide about NoDrift algorithm presented by Martijn
  • Martijn also suggested to change some parameters of NoDrift? algorithm


check the following things tomorrow.

  • check how many tracks are seen with "NoDrift" and not seen with "LinearDrift" in the event display (I'm going to scan about 100 events).
  • check 'track chi2' for the events to make different track between "LinearDrift" and "NoDrift".


Made some plots comparing 'LinearDrift' and 'NoDrift' with the following variable
  • number of muon tracks
  • number of DT Hits/Segments on track
  • eta of track
  • phi of track
  • chi2 of track


Added more variable to cosmic data analysis code
  • 'track chi2', 'eta', 'phi'


Martijn explained why 'LinearDrift' gave the different result with and without the change in DTSegment though he expected it doesn't affect in 'LinearDrift' alorithm. And I confirmed it.

------------- e-mail from martijn -------------
There was a problem with initialization of the propagators used when there
is no magnetic field (B=0)... This may explain why LinearDrift gave
different results with and without the 'NoDrift' fix. Could you try to
compare again 'LinearDrift' with and without the change in DTSegment,
after you apply the following fix ??

in your CMSSW_1_7_5/src directory:

 project CMSSW
 cvs co -r V01-06-03
 eval `scramv1 runtime -csh`
 scramv1 b




Discussed cosmic data analysis with Prof. Suyong Choi

  • check how many tracks are seen with "NoDrift" and not seen with "LinearDrift" in the event display.
  • check track chi2 for the events to make different track between "LinearDrift" and "NoDrift".
  • check eta and phi distribution of track for both "NoDrift" and "LinearDrift"
    (might get some clue why "LinearDrift" doesn't give good muon reconstruction. and it should be symmetric)


Martijn sent me the modified code to fix the slow process problem with NoDrift? algorithm


ran Martijn's cfg file with NoDrift?/LinearDrift option for comparison. In case of NoDrift?, it is very slow and it seem like that # of events setting doesn't work


read & ran the code Martjn gave and looked at the output histogram but have no idea what to do with cosmic data yet.


Martijn gave me another cfg file using different analyzer, PhysicsObjectsMonitor?. I asked him how to judge cosmic data with event display
Dear Hyunkwan,

How are things going with analyzing the cosmics data?

I have made a new cfg file:

This one uses a different analyzer, PhysicsObjectsMonitor. This analyzer
is a DQM analyzer, which produces root histograms in a slightly
different way (they are called 'Monitor Elements') It produces
a root file PhysicsObjectsMonitor.root which contains some useful
information about the reconstructed muons. You can run it as before:

 cmsRun GREN_DQMAnalyzer.cfg


 iguana GREN_DQMAnalyzer.cfg

The code for the PhysicsObjectsMonitor module is in

in your CMSSW_1_7_5/src directory you can do:

project CMSSW
addpkg DQM/PhysicsObjectsMonitoring

if you would like to see the code and change it (for example to add a new
histogram...) --> after you make a change to the code you have to
re-compile by doing

 scramv1 b

(in the CMSSW_1_7_5/src or the
CMSSW_1_7_5/src/DQM/PhysicsObjectsMonitoring directory...)

Have fun... Let me know if you have any questions,



  • Martijn sent me the cfg file to read cosmic data (GREN). it work well in my directory
mkdir YourDirectory
cd YourDirectory
scramv1 project CMSSW CMSSW_1_7_5
cd CMSSW_1_7_5/src
eval `scramv1 runtime -csh`
cp /afs/cern.ch/user/m/mulders/public/GREN/GRENAnalyzer.cfg .
cmsRun GRENAnalyzer.cfg
Inside GRENAnalyzer.cfg you can switch to the "NoDrift" algorithm by commenting out the following 3 lines:
    include "RecoLocalMuon/DTRecHit/data/dt1DRecHits_LinearDrift_CosmicData.cfi"   
    include "RecoLocalMuon/DTSegment/data/dt2DSegments_CombPatternReco2D_LinearDrift_CosmicData.cfi"
    include "RecoLocalMuon/DTSegment/data/dt4DSegments_CombPatternReco4D_LinearDrift_CosmicData.cfi"

and un-commenting the following 3 lines:

#    include "RecoLocalMuon/DTRecHit/data/dt1DRecHits_NoDrift_CosmicData.cfi"
#    include "RecoLocalMuon/DTSegment/data/dt2DSegments_CombPatternReco2D_NoDrift_CosmicData.cfi"
#    include "RecoLocalMuon/DTSegment/data/dt4DSegments_CombPatternReco4D_NoDrift_CosmicData.cfi"

  • there was a problem to run iguana in Xshell but it was solved by installing Xming
  • But I don't know what point I have to check in event display. It seems like that I have to learn about event display.
Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r3 - 2011-12-22 - GunmoRyu
This site is powered by the TWiki collaboration platform Powered by Perl This site is powered by the TWiki collaboration platformCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding KoreaCmsWiki? Send feedback