Validation of Reconstruction code

• List of Validation contacts for DPG and POG on Monte Carlo

• List of Validation contacts for Data certification

• Changes to the reconstruction software

• Documented for new features of the reconstruction

Legend:

• : Changes are expected, to be set according to software or condition changes prior to validation
• : Validation done, result is "Not Validated". Report the reasons below
• : Validation done, result is "Ok, Validated"
• : Validation still in progress or results need to be followed up before final green/red light
• : Validation not (yet) done
• : means not covered by anyone in the collaboration

• The global comparison is available at

Data in 5.0

Release

Comparison

Tk

Ecal

Hcal

DT

CSC

RPC

Tracking

Electron

Photon

Muon

Jet

MET

BTag

Tau

data in pre2

data in pre3

data in pre4

data in pre5

data in pre6

data in pre7

data in 5.0

data in 5.0.1

Details

• 29-Jan-2012: Many histograms on SiStrip detector and Tracking POG are empty due to a change in L1GtUtils which affected GenericTriggerEventFlag class (used for SiStrip and Tracking monitoring). The problem is fixed in the tag (V00-03-00) for CommonTools/TriggerUtils

Problem should be fixed in CMSSW_5_0_0_patch1 * 1-Feb-2012: (J-R, Laura) 500_patch1 give green to tracker. * 6-Feb-2012: (Kevin) 500_patch1 OK for tracking

MC in 5.0

Release

Comparison

Tk

Ecal

Hcal

DT

CSC

RPC

Tracking

Electron

Photon

Muon

Jet

MET

BTag

Tau

mc in pre2

mc in pre3

mc in pre4

mc in pre5

mc in pre6

mc in pre7

,

mc in 5.0

,

mc in 5.0.1

mc in 5.0.1 good Geant (v3)

mc in 5.1.2

Details

pre2

• 07-Nov-2011 : Electron pre2 : No RECO samples available.

pre3

• 07-Nov-2011 : Electron pre3 : The comparison was done with the 442 release.
• 04-Nov-2011 : RPC : Efficiency, noise and digi strip profile discrepancy in pre3. In the new globalTag from this release (START50_V3), some noisy strips were not masked and caused large noise rate.
• 24-Jan-2011 : Tracking : Discrepancy in fake rate around eta ~ 1.5 when comparing 500pre3 v/s 440 in case of the QCD_Pt_3000_3500 sample. This seems to get fixed in pre6.

pre4

• 05-Nov-2011 : Hcal pre4 : An absence of HBHE/HO SimHits due to a bug in the SimG4CMS/Calo package, fixed in pre6

pre6

• 12-Jan-2012 : Muons in pre6 and pre7 marked as BAD because of some wrong DT timing correction applied. Fixed for 5_0_0

final 5.0

• 22-Feb-2012 : Btag: Efficiency vs Discriminant plots for JP and JBP algorithms show lower efficiencies for 500 than for 500pre7 but the performance is not affected.Scale factor between FastSim and FullSim : known problem but not yet well understood.
• 20-Jan-2012 : RPC GlobalTag problem is fixed. Realistic noise level is simulated in the RPC digitization.
• 12-Jan-2012: Photons: New energy corrections are not 100% correct. Fix is under way.
• 19-Dec-2011: Ecal: there are inconsistencies at the level of super-clusters, from the new SC calibration.

5.0.1

• 2 Feb 2012: C. Charlot: We do see a changes between the two releases in many distributions that could be summarized as coming from changes in pout and distribution related to showering in the tracking material as well as changes in cluster shape towards narrower showers. To confirm if the G4 version has been changed between the two releases.

MC+PU in 5.0

Release

Comparison

Tk

Ecal

Hcal

DT

CSC

RPC

Tracking

Electron

Photon

Muon

Jet

MET

BTag

Tau

mc in 5.0 with 15-35 PU

Details

template

Release

Comparison

Tk

Ecal

Hcal

DT

CSC

RPC

Tracking

Electron

Photon

Muon

Jet

MET

BTag

Tau

template