shahoian [Thu, 17 Jun 2010 11:58:25 +0000 (11:58 +0000)]
Modified trackleting: AliMultiplicity stores info if the clusters
used by the tracklets or "singles" are used in tracks.
The trackleting step is moved from the vertexer to the end of
reconstruction loop.
agheata [Thu, 17 Jun 2010 10:17:41 +0000 (10:17 +0000)]
1. SetDefaultOutputs(kTRUE) used by default. This will automatically archive the analysis outputs. In this case, calling SetOutputFiles and SetOutputArchive is now forbidden and will produce a fatal error. To use arbitrary output files and archives, one has to manually reset the SetDefaultOutputs flag.
2. Specific SE's are not allowed anymore via the plugin.
3. To specify the number of replicas for the output archives, one should use SetNumberOfReplicas(n) where the default is 2 and the maximum allowed is 4.
4. stdout and stderr files are not registered in the catalog in case a job enters in DONE state.
mfloris [Thu, 17 Jun 2010 09:43:28 +0000 (09:43 +0000)]
AliCollisionNormalization:
- Adding member for energy and pointer to stat histogram (all events)
- Some more printout in ComputeNint
- Using energy in GetRelativeFractions
AliCollisionNormalizationTask:
- Using the bin0 callback
- Explicitely setting ntracklet = 0 in MC if the event is not selected
runProofNormalization.C:
- Changes to run on the new CAF
cblume [Wed, 16 Jun 2010 12:44:50 +0000 (12:44 +0000)]
Update by Jochen concerning HC header treatment, including:
- error output generated in error functions with debug level
(configurable at compile time)
- set error flags to steer further reading (run-time configurable for
every error type)
cvetan [Wed, 16 Jun 2010 11:30:25 +0000 (11:30 +0000)]
Adding a bit that specifies if the online (beam-beam,beam-gas) flags are filled (correctly) in the ESD. To be used in AliTriggerAnalysis in the replay of the online trigger.
marian [Tue, 15 Jun 2010 17:35:57 +0000 (17:35 +0000)]
Gap between paderow used for seeding as parameter in AliTPCRecoParam
(Marian)
Some theoretical consideration:
The track seeding efficiency depends on the cluster finder efficiency - ceff.
ceff is dEdx dependent, in mean is about 0.8 for MIP is about 0.75
The probability to find a seed=
pseed = ceff^3
Probability to find a track in nseedings
ptrack = (1-(1-pseed)^nseeding)
nseedin = npadrows/gap.
Until now we were seeding in outer sectors with gap 6 and 10 pad-rows (r dependent).
For tracks crossing the full TPC the track finding efficiency is
( you can draw it using TF1 f1("f1","(1-(1-0.75**3)**((160.-64.)/x))",1,10))
99.5% once the gap 10 pad rows is used.
For short tracks = kink decays e.g 30 cm from the end of TPC it is about 65% for gap 10 pad-rows, 85% for 5 gap 5 and 95% for gap 3 pad-rows.
(see function
TF1 f1("f1","(1-(1-0.75**3)**((20.)/x))",1,10))
masera [Tue, 15 Jun 2010 17:24:47 +0000 (17:24 +0000)]
- AliITSv11GeometrySPD.h : changed signature of method CreateConeModule
- AliITSv11GeometrySPD.cxx : mods in CreateConeModule to prolong cooling tubes where they were missing and to add freon
inside tubes; CreateCones has been modfied to add new volumes
- AliITSv11GeometrySupport.cxx : mods in SDDCone to add cables inside the holes of the SDD cone;
mods in SPDCableTraysSideA and SPDCableTraysSideC to add HV cbles in the trays and freon in the cooling tubes
- AliITSv11Hybrid.cxx : mods in CreateMaterials to define gaseous Freon and the average material for SPD HV cables.
belikov [Mon, 14 Jun 2010 20:05:49 +0000 (20:05 +0000)]
Bug fix with matrix inversion in SetProductionVertex(); sign of GetDistanceToVertexXY() fixed; GetDecayLengthXY() and GetErrDecayLengthXY() methods added (S. Gorbunov)
rpreghen [Mon, 14 Jun 2010 12:52:31 +0000 (12:52 +0000)]
updates on TOF software:
- added new TOF raw data decoder (enhanced decoder)
- enhanced decoder enabled as default in reconstruction
- TOF raw data policy updated
- all calibration code in AliTOFcalib class
- updates on TOF-T0 classes
- DA for physics runs split for physics and calib events
- DA-physics computes readout+decode efficiency
- PreProcessor put readout+decode efficiency in OCDB
- readout-info object added (Reference)
- changed online T0-fill measurement in PreProcessor
- added online resolution measurement in PreProcessor
- added ResponseParams class and file for TExp correction
cvetan [Mon, 14 Jun 2010 12:27:20 +0000 (12:27 +0000)]
Major fixes in the digits and raw-data reco. Now both are equivalent. To be propagated to the release. The trigger inputs bit mask is filled in the friend only in case of raw data. Will be fixed when we switch to the new trigger framework. It is not clear now where to put this data.
gconesab [Fri, 11 Jun 2010 13:16:02 +0000 (13:16 +0000)]
Cleaned usage of TClonesArrays in the loader
In the AliEMCALRecParam::Print, cosmetic changes.
Macro AliEMCALSetRecParam.C, update with last parameters for real data reconstruction
odjuvsla [Thu, 10 Jun 2010 15:23:09 +0000 (15:23 +0000)]
- added different sorting algorithms for the digits
- pre-defined compare functions for sorting by energy and by position
- also possible to set the compare function to anything you like
- cleaned up some unnecessary code
- added some comments
esicking [Thu, 10 Jun 2010 13:22:39 +0000 (13:22 +0000)]
Added macros for production using Geant4. 3 different physics lists can be used: QGSP_BERT_EMV, CHISP and QGSP_BERT_CHIPS. All 3 PL can be used combined with optical processes, too. sim.C and rec.C can be used for local testing. Check simrun.C for the usage.
cvetan [Thu, 10 Jun 2010 13:03:27 +0000 (13:03 +0000)]
Major fixes to the raw-data simulation. The ADC samples and signal width are added to the raw-data stream. The trigger flags are stored in the digits tree user-info and then propagated to the raw data and esd. The digit class is updated by removal of unneeded members. Some code clean-up as well. We need one more step before in the reconstructor in order to declare raw-data and digits sim/reco are completely equivalent. This step is to move the code from ConvertDigits to Reconstruct method.