Allow opening of the file from two places in case at least one of them is read-only. Should not affect any normal offline processing, but unavoidable for online threaded alieve+aliroot reconstruction.
Enabled reading from the MC particles in the AOD, activate with SetReadAODMC(1) (all) or (2) (charged). if activated reconstructed tracks are ignored when FillMomentumArray is called
1. in AliMixedEvent, the method GetMagneticField() returns -999. Is it possible to have, instead, the magnetic field of the first
event used to build the mixed event?
2. in AliMixedEvent, there are 2 methods to get the primary vertex: one is GetPrimaryVertex and returns a null pointer, the second
is GetPrimaryVtx and returns the mean vertex. Would it be possible to remove the second one and let the first one return the mean
vertex?
Yes to 1 and 2
1. Leading time was not decoded properly (there are coarse and fine gains in the HPTDC which I did not not about). 2. HPTDC were written in raw data from 7 to 0 within a ring whereas for everything else it was from 0 to 7. This was Yannick's mistake but he did not want to change the programming of his FPGAs so I had to change it in aliroot. 3. All the ADCs in Float_t instead of integers. (Brigitter). Changes in the ESD classes were agreed by Andreas, Rene confirmed that the schema evolution will handle them properly.
AliCaloTrackAODReader: When requested, keep the selected Tracks and CaloClusters in the output AOD.
AliAnaPartCorrMaker: Do not abort when there is no input analysis, just print that it is doing no analysis.
Added access to magnetic field value in readers. Corrected bugs in AliMCAnalysisUtils when tagging electrons. General updates in AliAnaElectron (Jenn)
Few minor coding violations corrected.
Change in the monitoring policy. Use the DATE monitoring feature which takes events from all GDCs associated to a given run. The run is transmitted via env variable. The GRP preprocessor is changed so that still to give the GDCs, but this time the full list used in the run. This could be useful in the future.
Moving a some structs and classes used for handling "AliESDCaloCluster like" clusters inside HLT.
This was done to avoid dependencies between libAliHLTGlobal and libAliHLTPHOS when we merge
PHOS clusters into the global HLT ESD.
Several classes in the HLT/PHOS had to be changed due to name changes of the structs/classes
- AliMUONQAChecker prevent the method AliMUONQAChecker::CheckRaws to return a NULL pointer if a
TrackerData object is not present in the first specie (kDefault). Now it returns
NULL only if there no TrackerData object at all.
- AliMUONCalibrationData : add a bypass to force the usage of some external calibration data (for AMORE)
- AliMUONTrackerDataWrapper : remove an uneeded method
(Guillaume)
- AliMUONQADataMakerRec : introducing the beautifier for presenting a nicer Raw Data QA to the shifter.
This is just a first approach, as most of that should probably be done in the Checker on the longer term (to be discussed...)
- AliMUONDigitCalibrator & AliMUONTrackerDataMaker : do not force the number of sigma to cut on the charge. Should be taken from the recoparam.
This was impacting the QA only (forcing it to use a 3 sigma cut, where anywhere else we use 4 sigma cut).
bugfix https://savannah.cern.ch/bugs/?55067
Updating AliHLTMessage to TMessage r27689
HLT encountered a problem when streaming the AliESDEvent. The problem occurs if the
TClonesArray members contain more then one entry. In this case, members of the individual
objects have been mixed.
The problem turned out to be solely on the serialization part, and did not effect the
restoration part. This has been proven by using either TMessage and AliHLTMessage on the
sender, and only AliHLTMessage on the receiver side. Merging the recent changes in
TMessage solved the issue.
1)AliCaloPID: Posibility to recalculate PID bayesian in EMCAL
2)AliAnalysisTaskEMCALPi0CalibSelection: Corrected number of supermodules and towers
Few coding violations corrected.
Tender wagon providing access to ESD event and CDB. The tender calls an arbitrary number of user algorithms that add or correct information in ESD based on CDB info that was not available during pass1 reconstruction. The base class for user-defined ESD additions and corrections is AliTenderSuppy. This has access to CDB and the current ESD event and has to implement the methods Init() (specific CDB storages, initialization) and ProcessEvent() (algorithm that adds to or modifies the current ESD event.
ivana [Tue, 1 Sep 2009 09:46:08 +0000 (09:46 +0000)]
In reconstruction:
Added protection in case the bending momentum of the track is so low that it turns around in the magnetic field.
--> Most of the time, stop the tracking of that track and remove it.
--> In any case, its chi2 is set to a maximum value (1.e10) so it can be easily detected and removed afterward.
(Philippe P.)
marian [Mon, 31 Aug 2009 16:00:42 +0000 (16:00 +0000)]
AliTPCcalibDB.cxx.diff Change magnetic field return value from kG to T
AliTPCCalibViewerGUItime.cxx.diff Change from tree to chain. Allow wildcard in static AliTPCCalibViewerGUItime::ShowGUI("file*.root")
AliTPCCalibViewerGUItime.h.diff Change from tree to chain. Allow wildcard in static AliTPCCalibViewerGUItime::ShowGUI("file*.root")
CalibEnv.C.diff Add Nan protection in noise calculations
TPCRAWda.cxx.diff Add network timeout as suggested by Sylvain
cvetan [Mon, 31 Aug 2009 15:41:12 +0000 (15:41 +0000)]
Changes in order to be sure that the raw-reader always detects the EOR events. In prinicple even this is not enough because it can happend that a detector is busy and the run ends w/o EOR events. Will have to put a SOR notifier based on DIM.
agheata [Mon, 31 Aug 2009 10:12:49 +0000 (10:12 +0000)]
Added possibility to lock/unlock CDB manager. User is provided with a key everytime it locks. The key has to be used in the SetLock() method whenever the state of the lock need to be changed. (A.G.)
cblume [Mon, 31 Aug 2009 07:55:05 +0000 (07:55 +0000)]
1. more than 30 timebin data reading is available now in the fast reader
2. rawstreamer provide the number of timebin information from HC header
and put it in a AliTRDdigitsParam, so later you can get this nubmer of
timebin information through digit manager like:
digitsparam = (AliTRDdigitsParam *) digitsManager->GetDigitsParam();
digitsparam->GetNTimeBins();
so, Theodor also can use this function for HLT.
3. enable swapped optical fiber checking
4. bug fix in getting the chamber number directly through
rawreaderpointer->GetDet() fuction(based on Tom's bug report) but it
doens't give an effect in normal reconstruction.