Updates and bug fixes for the full tracker and hit reconstructor. (Indra)
The FullTracker code was crashing because it was not correctly handling the chambers ID and detection element ID. Some protection is added for fullTracker component at the
DoEvent method.
If the chi2 value is too high in case Kalman calculation that track comparison is also skipped. This is the protection to avoid the cases with Determinant=0 during Kalman
Chi2 test.
Getting rid of a cut-and-pasted piece of code in CorrectForMaterial(...) and introducing a new CorrectForMeanMaterialZA(...). The arguments of this new function allow for setting the values of all the parameters of the BB formula used inside.
- BX ids are now selected by default. For the 7 TeV, the 2b_1_1_1 is assumed in the run interval 114783 - 116684.
- Muons: If you are analizing muons and you want to keep the muon triggers (CMUSB), besides the CINT1B, you can set:
physSelTask->GetPhysicsSelection()->SetUseMuonTriggers();
- Trigger pattern: A new histogram showing the pattern of triggered events was implemented. This should allow to spot quickly
problems with the triggering detectors.
Adding class to merge calibration objects generated chunk by chunk
from which to then extract the calibration parameters.
Based on the mergeCalibObjects.C macro in
$ALICE_ROOT/ANALYSIS/CalibMacro/MergeCalibration.
(Marian (main part) + Chiara (fixes to run on alien))
Adding new task to implement central filtering (AliAnalysisTaskFilterSteer).
Removing obsolete task (AliAnalysisTaskFilterSTEER).
Adding steering macro for AliAnalysisTaskFilterSteer.
- Implementing Central task for filtering the ESD friends:
- keeping only 1% of the whole sample, randomly selected
- Adding the setting/getting of the "skip" bit (bit 23) for AliESDfriendTrack/AliESDfriend
during the filtering
- AliESDfriendTrack --> to allow to have the logical or between tracks from different tasks
when adding a "dummy" friend track (when not selected by the filter)
- AliESDfriend --> to be able to disentagle between ESDfriends added because
selected, or added as "dummy" ones when not selected
- Corresponding changes in AliAnalysisTaskFilter/AliESDHandler and in the
example filtering tasks
- Removal of AliAnalysisFilterSTEER:
- the 1% selection is now done in AliAnalysisFilterSteer
- the pt selection should be handled in separate detector-specific tasks
- macro for filling histograms for tracking diagnostics (TPCTrackHisto component)
(shorter and legible version of the very long HLT-TPC reconstruction macro)
- first trigger version for ultra-peripheral collisions (Kyrre Skjerdal)
- added macro that runs the HLT reconstruction and processes the output through the trigger component
(global trigger menu to be modified, so that the chain can be complete with the addition of the global trigger component)
- no arguments passed from OCDB entry yet
ivana [Wed, 21 Apr 2010 15:13:02 +0000 (15:13 +0000)]
In MUONRecoCheck.C:
- Add protection against 0 when computing the average momentum per bin
- Correct the phi resolution for a possible +- 2*pi shift in its computation
(Philippe P.)
Updated AliUnfolding class. The regularization schemas have changed taking into account bin
width and in a different way the error of the single bin.
Please CHECK the weights given to the regularization. They most likely need to be changed.
Added to fieldmaps the parameterized integrals of field transv.components/bz
AliMagF::GetTPCRatInt(xyz,b) gives the integral of bx/bz, by/bz and
(bx/bz)^2+(by/bz)^2 from 0 to xyz[2].
The (old) integral of components /GetTPCInt(xyz,b)/ is brought to the same
convention (before was from z=+-250 to xyz[2])
The min/max Z and R allowed is 300 cm.
GetTPCRatIntCyl and GetTPCIntCyl provide similar integrals in cyl.frame.
Reparameterized