lmilano [Mon, 14 May 2012 18:59:32 +0000 (18:59 +0000)]
- processed events + QA plots on vtx and dndeta added in Event Cuts - small bugfix in task (eta cut) - runAOD updated - YLabel of generated changen in histomanager - some new plots in MainAnalysis
miweber [Sun, 13 May 2012 11:41:35 +0000 (11:41 +0000)]
Introduced HBT and conversion cuts for standard BF analysis; Move from Vectors to AliBFBasicParticle for IO to AliBalance for triggered BF analysis (first step for DPhiCorrelations like Event Mixing)
marian [Sat, 12 May 2012 07:27:10 +0000 (07:27 +0000)]
Changes (Marian and Mikolaj):
1. To enable merging per component
2. New TPC and TRD calibration
3. More configuration parameters to enable automatic tests of CPass0/CPass1/QA
on the farm. Extensive test were done in this mode.
CPass0/mergeMakeOCDB.byComponent.jdl
new jdl which calls the new merge by component macro
CPass0/test/runCPass0User.sh
run number is now a number (e.g. 146807) not a string ("000146807")
CPass0/test/mergeMakeOCDBUser.sh
run number is now an integer
CPass0/makeOCDB.byComponent.C
run number is now an integer
CPass0/makeOCDB.C
run number is now an integer
CPass0/runCPass0.sh
add makeSyswatchCPass0.C which adds the syswatch tree to calibration output
CPass0/mergeMakeOCDB.byComponent.sh
a new scheme of merging facilitating two main changes:
- the merging of large calibration files by component, every top directory
in the file is merged in a separate aliroot session keeping the memory use down.
- files are downloaded in groups (default 100 files) and merged in parallel to
improve the cputime/walltime ratio
CPass0/mergeMakeOCDB.sh
run number is now an integer in the arguments + readability
CPass0/mergeByComponent.C
- remove hardwired file names
- split functionality of producing a list of alien files using TGrid (for use on alien)
and copying them to allow for manipulation of the list
CPass0/recCPass0.C
- add TRD pad gain use
- explicitly set the detectors to be reconstructed
CPass1/mergeMakeOCDB.byComponent.jdl
new file
CPass1/mergeByComponent.C
new file
CPass1/ConfigCalibTrain.C
- allow for setting specific storages for use in testing
CPass1/test/runCPass1User.sh
- run number is now an integer
CPass1/QAtrain.C
- switch off MUON
- remove hardwiring of raw:// as default OCDB storage
CPass1/recCPass1_OuterDet.C
- allow for setting specific storages to the local output of CPass0 for testing
CPass1/makeOCDB.C
- run number is now an integer
CPass1/runCPass1.sh
- setting of local storages from CPass0 if available (check for localOCDBaccessConfig.C file)
CPass1/mergeMakeOCDB.byComponent.sh
- like in CPass0
CPass1/recCPass1.C
- setting of local storages from CPass0 if available (check for localOCDBaccessConfig.C file)
- use TRD pad gain
- explicit switching on of detectors
CPass1/runCalibTrain.C
- allow for setting specific storages to the local output of CPass0 for testing
marian [Fri, 11 May 2012 22:12:52 +0000 (22:12 +0000)]
AliTPCPreprocessorOffline.h AliTPCPreprocessorOffline.cxx - Using new component (AliTPCCorrectionFit) to create distortion maps
AliTPCExBEffectiveSector.h AliTPCExBEffectiveSector.cxx - remove functions implemented in the AliTPCCorrectionFit
zampolli [Fri, 11 May 2012 15:47:07 +0000 (15:47 +0000)]
Updating T0 make OCDB part (Alla): always update PMT offsets and the global offset (see https://savannah.cern.ch/task/index.php?27425, comment #171 and following)
abercuci [Thu, 10 May 2012 13:01:51 +0000 (13:01 +0000)]
The position of the trd clusters in EVE is transformed with different
matrices depending on the flag fIsMisaligned. As figured out in the
past, the flag is not set correctly or this information is lost at some
point. Just a reminder:
https://indico.cern.ch/getFile.py/access?contribId=1&resId=0&materialId=slides&confId=190673
Adapted the trd_cluster.C macro to just skip the flag and take the
correct transformation. Now, in May 2012, the flag seems to be still
wrong (see slides 1-4 in):