richterm [Sun, 8 Nov 2009 12:07:59 +0000 (12:07 +0000)]
allow names with the '-' character in items. This is a temporary solution like in r35411.
It has beendeactivated in r36097, but it turns out to be necessary again for the correct
evaluation of the trigger menu until the final concept is in place
See https://savannah.cern.ch/bugs/index.php?58205
richterm [Sat, 7 Nov 2009 23:32:33 +0000 (23:32 +0000)]
removing all code regarding the solenoidBz OCDB entry formerly used
to propagate the field value from DCS to the components in the online
system. The framework now initializes the global AliMagF instance,
which is used by the components fot the field setup
richterm [Sat, 7 Nov 2009 22:41:00 +0000 (22:41 +0000)]
adding documentation of the field conventions
using the AliGRPManager again for verbose output
changing the default polarities (corresponding to positive fields):
l3: positive -> 0
dp: negative -> 1
richterm [Sat, 7 Nov 2009 03:03:05 +0000 (03:03 +0000)]
reading the CTP configuration in the initialization of the HLT reconstruction and
passing the trigger mask via the AliHLTTrigData structure to the components
mtadel [Fri, 6 Nov 2009 20:32:19 +0000 (20:32 +0000)]
Almost final version of scanning tools.
* AliEveEventManager
Emit new signal NewEventDataLoaded() after the event-data is
connected but before the macros are executed.
* AliEveTrackCounter
Connect to AliEveEventManager::NewEventDataLoaded() to do the reset.
Introduce 'Bool_t fAcitve' -- now that track and tracklet label is
misused to store scanning results it is better that track-counting
is not enabled by default.
* AliEveTrackCounterEditor
Add Activate / Deactivate buttons and handlers.
Remove histogram plotting code -- now in show_scan_results.C
* visscan_init.C
Remove code for registration of tracks/tracklets to
track-counter. This is now done in esd_tracks / esd_spd_tracklets.
* make_scan_results.C
Create the scan-result tree from individual objects stored during
scanning.
* show_scan_results.C
Create canvases and plot basic stuff from the scan-result tree.
* esd_spd_tracklets.C
Put tracklets into two containers -- Good/Bad.
Register the tracklets to track-counter if it is active.
* esd_tracks.C
In esd_tracks_by_category() register the containers to track-counter
if it is active.
zampolli [Fri, 6 Nov 2009 19:23:48 +0000 (19:23 +0000)]
Fixing behaviour wrt DCS DPs related to pressure sensors:
- if the fit is there --> ok
- no fit, but run lasting less than 60 s (according to AliDCSSensorArray) and TGraph there for the
sensor --> ok
- in any other case, don't increase the count of the DCS DPs --> making GRP fail
Changing also kFitFraction from 0.7 to -1 (Haavard), but it will not have much effect with all the other changes...
sgorbuno [Fri, 6 Nov 2009 16:36:31 +0000 (16:36 +0000)]
- There was impossible to set solenoidBz value from the configuration string. Now parameters from the configuartion string are set after the OCDB parameters.
- The order of filling of ESD tracks is changed: TPC, TPCOut, ITSOut, ITS.
The ITSin parameters are stored in the ESDtrack itself, therefore they are lost when the UpdateParameters(..) is called next time.
To keep the ITSin parameters in the ESD, one should set these parameters last.
laphecet [Fri, 6 Nov 2009 16:36:26 +0000 (16:36 +0000)]
Adding counter for TOKEN_LOST errors in AliMUONRawStreamTrackerHP.
Adding symmetric methods to AliMUONRawStreamTriggerHP, as found in AliMUONRawStreamTrackerHP for counting errors for whole events.
richterm [Fri, 6 Nov 2009 14:47:49 +0000 (14:47 +0000)]
adding functionality to downscale the publishing of TObjects in PushBack
The command line paramameter -pushback-period=<period> specifies the time
interval for the publishing. Especially important for histogramming
components which do not need to ship the data for ever event.
morsch [Fri, 6 Nov 2009 11:32:36 +0000 (11:32 +0000)]
The ANALYSIS one is a modification of the AliXMLCollection which adds
support for empty event lists (with the current code it crashes for an
empty evlist).
The STEER one is to add two new functions to the ESDInputHandler, one
to calculate the cut summary for the given list, the other to count
"empty" files, that is files with no events selected.
(A. Kisiel)
ivana [Fri, 6 Nov 2009 10:54:35 +0000 (10:54 +0000)]
In MUONChamberMaterialBudget.C
- Use the name of the mother volumes instead of the position to select
material belonging to a given chamber
- The resolution can be changed by changing the sementation level:
resolution = 1 cm / segmentationLevel
(Philippe P.)
richterm [Fri, 6 Nov 2009 01:22:19 +0000 (01:22 +0000)]
Adding common handling for the magnetic field settings tothe component interface
Components will use the AliMagF instance in the future. This instance is set up
by the framework
1) AliRoot: AliReconstruction, AliSimulation or explicitely in the macro
2) online environment: AliMagF initialized via the GRPManager from the ExternalInterface
This change will soon make all private code regarding setup of AliMagF obsolete in the
HLT code. Furthermore the SolenoidBz OCDB entry will be removed.
richterm [Thu, 5 Nov 2009 20:43:45 +0000 (20:43 +0000)]
bugfixes by Theodor and Raphaelle
- suppress empty histograms if no tracks were found (fix according to the current shuttle behavior)
- print the object size and not the pointer size
- correct value of min entries for vdrift calculation
mtadel [Thu, 5 Nov 2009 18:08:05 +0000 (18:08 +0000)]
* show_scan_results.C
New macro - build tree from scan results on individual events.
* esd_spd_tracklets.C
From Domenico - add cuts to mark "bad" tracklets at load time.
* AliEveTrackCounter
Store scan results by writing ESD track container and AliMultiplicity
into scan_results.root.
Display the results by running show_scan_results.C and then
calling TTree::Draw() commands on the created tree.
Probably the drawing function should also be a macro.
laphecet [Thu, 5 Nov 2009 17:40:25 +0000 (17:40 +0000)]
Consolidation of the classes used to get the calibration data stores from OCDB and ASCII files/string.
This is particularly important for AmoreMCH.
Note that AliMUONTrackerIO has been improved as well (e.g. we no longer have multiple temporary TString objects)
While doing so, two bugs were corrected :
- Bug fix : corrected the interpretation of the HV switches (important both for QA and reco)
- Bug fix : corrected the display of the config in mchview which was wrong (again, important for AmoreMCH)
cvetan [Thu, 5 Nov 2009 17:16:36 +0000 (17:16 +0000)]
Possibility to use a rec.C file outside the aliroot distribution folder. To be used as soon as we switch to aliroot RPM installation at P2. In this way one allows changes to the reco macro without touching the aliroot installtion done by DAQ. The macro location and name are defined by ONLINERECO_MACRO env. In case the env is not there, the standard rec.C from MONITOR is used.
martinez [Thu, 5 Nov 2009 14:28:00 +0000 (14:28 +0000)]
Improved algorithm for the extrapolation of the tracks in AliCheckMuonDetEltResponse. Adding efficiency computation per detection element. Adding special setting conditions for a cosmic runs. Adding histograms of Tracks detected and total as output of AliAnalysisTaskMuonTrackingEff (Matthieu)
sgorbuno [Thu, 5 Nov 2009 14:18:16 +0000 (14:18 +0000)]
quality check for the maximum Pt error removed. The Pt error depends on the magnetic field value and doesn't show the error of the reconstructed curvature.