Extending the functionality of the physics selection. Instead of providing a yes/no decision, it
supports now several concurrent offline trigger selections. At present these are (defined in
AliVEvent.h --> EOfflineTriggerTypes):
1) AliVEvent::kMB: Minimum bias trigger (corresponding to the old behavior), all detectors
2) AliVEvent::kMBNoTRD: Same as kMB, but TRD not read out (this is a separate online trigger
class which became active now)
3) AliVEvent::kMUON: Muon trigger
4) AliVEvent::kHighMult: High-multiplicity trigger
For all those the (usual) offline SPD or V0 selection is performed. The available offline
triggers will be extended when needed.
To use this functionality you call for your task SetCollisionCandidates with a mask, composed
from the above flags. E.g. to select MB and MUON events:
myTask->SetCollisionCandidates(AliVEvent::kMB | AliVEvent::kMUON)
Doing that your task is only called for the requested events.
If in your UserExec you want to look at the decision of the physics selection, you can do that
in the same way as before with
UInt_t mask = fInputHandler->IsEventSelected()
and in mask the bits corresponding to the above offline triggers are set.
The output bit mask of the physics selection is stored in the AOD event header, and therefore
the AOD analysis can work in the same way as ESD analysis. I.e. you can use
myTask->SetCollisionCandidates and fInputHandler->IsEventSelected() in the exact same way.
adding the D0 trigger component to default the configuration for HLT simulation. Depending on the availability of simulated raw data the component works on ESDs and MC information (running on digits) or ITS tracks and vertex information (raw data)
make generation of monitoring trigger an optional feature
- to be switched on by argument '-monitoring'
- introduce an interval of n seconds between monitoring triggers '-monitoring=n'
- send empty block list for not triggered events and events dropped because of the interval
This change will reduce the amount of data sent to the monitoring components, first of all
by sending only sparse triggers (configurable interval) and secondly because of the empty
monitoring filter list for all but the single triggered event per interval.
Many improvements in EMCAL calibration class:
- Added possibility to remove bad channels/clusters by hand
- Added possibility to recalibrate channels on the flight
- Fill always max energy channel invariant mass for 2 clusters of the pair
- Added possiblity to group cells
- Set histograms range and binning via setters
- Publish parameters in output file
Changes in PROOF*/SETUP.C to ensure that the new par file compiled library is the one used.
removing obsolete functions for initialisation of streamer info which had been moved to AliHLTMiscImplementation in r41147, redirecting calls to AliHLTMisc
Introduction of the LHC-clock phase class + changes in the GRP pp in order to read the DCS data and fill the phase object. Next ste to do is to activate this method in the execution chain on the pp.
ivana [Wed, 21 Jul 2010 14:19:33 +0000 (14:19 +0000)]
In AliMUONSimpleClusterServer
Clearing of fPads[] maps in UseDigits() replaced with clearing of TClonesArrays within the maps. This avoids re-creating of TClonesArrays with each event and fixes the problem #69261.
- Protection in MC, not to access stack out of range (Markus K.)
- Add Print function in SignalBase
- Add ITS nSigmas in VarManager (Markus K.)
- Add PID class in pkg file. Was missing in the last commit
Corrected raw->sdigits method. Using default altro raw-data reader class. Avoid allocating and deleting of containers after each sector + other minor fixes.
bug https://savannah.cern.ch/bugs/?69885
it turned out that a wrong calibration object caused the displacement of clusters
putting back revisions 41975 and 41935
Fixes related to the latest changes in raw-data decoding + some additional histos + new set of histos for the QA shifter (expert<->non-expert) (Raphael)