marian [Sun, 28 Mar 2010 21:22:50 +0000 (21:22 +0000)]
1. Cleaning of the code
2. Possibility to filter Laser outlyers in analysis/
Algorithm implemented before (tested with 2009 data) on the calibration level with 2010 filtered full calibration.
laphecet [Sun, 28 Mar 2010 21:22:01 +0000 (21:22 +0000)]
Modifications to get a much more explicit report on token lost readout errors for DQM. Plus a small bug fix in RawStreamTrackerHP wrt to that error reporting
gconesab [Sun, 28 Mar 2010 21:09:20 +0000 (21:09 +0000)]
AliCaloTrackReader, AliAnaCalorimterQA, AliAnaPi0: Leak with PHOS rotation matrices found and corrected, revert changes in the PHOS geometry initialization.
AliMCAnalysisUtils: Put a protection against stack=0 or AODParticle branch = 0
AliAnaPhoton: Revert wrong modification of the y position of the cluster.
Few other minor corrections regarding non streaming some pointers, deleting in the main task destructor the maker pointer.
zampolli [Sat, 27 Mar 2010 15:59:01 +0000 (15:59 +0000)]
Updating configurations.
For mostly of the detctors nothing has changed - GRP and TPC have
CavernAtmosPressure2 added, and TOF and V00 have alidcsamanda2 instead of alidcsamanda.
Saving anyway the configuration since there are some issues with the
amanda server.
gconesab [Sat, 27 Mar 2010 15:18:32 +0000 (15:18 +0000)]
Fix memory leaks in output container for each analysis, delete when not used anymore in the Maker, in AliCaloPID TFormula leaks, remove usage until it is understood.
morsch [Sat, 27 Mar 2010 10:04:00 +0000 (10:04 +0000)]
Brute force initialisation of V(,) array at the beginning of each event
solves floating point exception (big number in V(4))
Probably a MSTU(33) = 1 at the right position would do the same job.
morsch [Fri, 26 Mar 2010 08:41:11 +0000 (08:41 +0000)]
We have implemented, in AliGenMuonlib.cxx, the possibility to generate JPsi events with a realistic distribution in one variable (eg. pt)
and a flat distribution in another variable, for example y. This is useful to compute systematics error on our acceptance evaluation.
R. Arnaldi
E. Scomparin
richterm [Thu, 25 Mar 2010 23:50:41 +0000 (23:50 +0000)]
Activating some parts of the code which have been commented out by mistake. Prevents the TRD preprocessor from crashing when no TRD tracks were found (Raphaelle)
decaro [Thu, 25 Mar 2010 23:14:06 +0000 (23:14 +0000)]
Updated code to read LV.and.HV status maps during a run. Waiting numbering conversion problem at DCS-Amanda level before toupdate HW status map in OCDB.
aszostak [Thu, 25 Mar 2010 21:18:11 +0000 (21:18 +0000)]
Fixing bug in generated trigger logic code for checking class inheritance.
Fixing bug due to possible missing dictionaries when loading the global decision. Had to put back explicit checks for NULL pointers since the custom streamer is not
guaranteed to be called if the object was written to a TTree with split-mode > 0.
aszostak [Thu, 25 Mar 2010 13:57:52 +0000 (13:57 +0000)]
Fixing bugs.
A crash was being caused by incorrect checking of pointers in AliHLTMUONDecision.
AliHLTMUONRootifierComponent was incorrectly making duplicate track objects if they were not found.
The deep copy of AliHLTMUONEvent was not maintaining the cross reference pointers correctly.
Fixing documentation for AliHLTMUONMansoTrack.
Improving checking for MUON module in AliHLTMUONAgent.
kkanaki [Thu, 25 Mar 2010 11:10:31 +0000 (11:10 +0000)]
- complete modification of the creation of AliTPCseed objects from HLT tracks (now method from AliHLTTPCdEdxComponent is used)
- added MC labels on the seeds for comparison between HLT and offline results
- dEdx information present (histogram)
This component is not used in the calibration any more. At the moment it will serve more as a means of comparison with offline. After optimization and correct placement
of the process on the HLT cluster, it can be used for providing the dEdx online.
- testing macro based on AliSimulation producing a TObjArray of AliTPCseed objects per event
rgrosso [Thu, 25 Mar 2010 08:56:32 +0000 (08:56 +0000)]
Added the method AddDateToComment, adding the current date to the comment datamember.
This method is supposed to be useful if called at the time when the object
is created, so that later it can more easily be tracked, in particular
when the date of the file can be lost or when one is interested in the
date of creation, irrespective of a later copy of it