cvetan [Wed, 20 May 2009 09:01:57 +0000 (09:01 +0000)]
Possibility to include ARVersion.h witihin the user code (avoid the gmake circular dep). Closes the savannah bug#48269 and allows to implement the code related to bug#50049
hristov [Wed, 20 May 2009 07:51:56 +0000 (07:51 +0000)]
Fix for bug #45751: Low V0 efficiency at low momentum and large radii. Ported from v4-16-Release, rev. 30636: The check on the return value is temporarily disabled.
cvetan [Tue, 19 May 2009 16:35:27 +0000 (16:35 +0000)]
Introducing the fixes needed in order to read the old RCU firmware format (v<3) with the new altro decoder. The fixes are done in a way so that it would be easy to remove this addition whenever the backward compatibility will not be needed anymore. Removal of some obsolete code related to the dummy altro trailer. It is not used by the detectors for long time as they implemented their real hardware mapping.
richterm [Tue, 19 May 2009 12:07:15 +0000 (12:07 +0000)]
add the CA tracker to the HLT TPC simulation macro. The algorithm
is chosen via an option string. Available options for the
moment: "CA" (default), "CM"
bugfix esd_hlt_tracks.C:
use the esd_track_propagator_setup as it is done from the esd_track macro
agheata [Tue, 19 May 2009 09:30:03 +0000 (09:30 +0000)]
Added support for copying the outputs of the analysis in an AliEn folder from a
PROOF job. One has just to supply a valid path to an Alien location to the method:
AliAnalysisManager->SetSpecialOutputLocation(path); where:
path = "alien:///alice/cern.ch/user/.../myOutput/" has to be prefixed by alien://
and must correspond to the user token that started the proof job. The common
output files (AOD) will be copied to this location as AliAOD_n.root from each
slave (n) while the rest of the outputs (histograms) will be first merged then
copied to the same location.
masera [Mon, 18 May 2009 16:56:01 +0000 (16:56 +0000)]
Fixing the problem that was causing the dip (savannah bug 48080):
the azimuth of the track momentum was used instead of the azimuth of the
track position. The protection introduced by Ruben and Martin (Rev 32258
and 32424) is not removed, because it makes sense (A. Dainese)
cvetan [Mon, 18 May 2009 14:18:52 +0000 (14:18 +0000)]
Another round of fixes in order to use the event specie in the QA. The procedure is the following: 1. The QA data maker gets the event specie as defined by AliRecoParam class. 2. If the detector's reco-param object has different event specie the previous value is overwritten and QA uses the specie defined by the detector. If the event specie from AliRecoParam and the detector's event specie are both set to default, the QA will not fill any histograms. Therefore the detectors should properly implement and put in OCDB at least one reco-param object (there should be also one and only one default object) with event specie set to low-mult, high-mul, cosmic or calibration specie. The changes were prepared and tested by Yves.
agheata [Mon, 18 May 2009 06:39:25 +0000 (06:39 +0000)]
Train configuration can be now saved and replayed. If 'saveTrain' flag is set,
the train will generate a directory name and run in this directory.
A configuration file 'ConfigTrain.C' will be generated. One can replay at any
time the train via:
root[1] AnalysisTrainNew(ana_mode, plugin_mode, "train_default_<date>/ConfigTrain.C")
cvetan [Fri, 15 May 2009 19:05:05 +0000 (19:05 +0000)]
Bug fix: the object counter should be never reset inside the raw-data root-ification since the referenced objects are reused and this can lead to raw-data payloads mixing in case the data is read in an order different to the writing one. The fix should be propagated to v4-16-Release together with the new raw-data event format.
abercuci [Fri, 15 May 2009 14:56:32 +0000 (14:56 +0000)]
- add LUT correction for y coordinate of clusters (fitted)
(AliTRDcluster)
- use corrected radial coordinate for dq/dl calculation (AliTRDseedV1)
- first definition of a charge resolution plot (AliTRDresolution)
- chi2 plot for each degree of freedom (no of tracklets) (AliTRDcheckDET)
- miscellaneous
cvetan [Fri, 15 May 2009 14:01:03 +0000 (14:01 +0000)]
New flat raw-data event format. The details can be found at: http://indico.cern.ch/getFile.py/access?contribId=2&resId=0&materialId=slides&confId=58946. The raw-data tags are still to be revised.
vulpescu [Fri, 15 May 2009 13:32:22 +0000 (13:32 +0000)]
Main changes:
- modify the logic of exporting files to the FXS: GLOBAl and REGIONAL files
are exported whenever at least one of the four configuration files (GLOBAL,
REGIONAL, LOCAL, LUT) change the version in the detector DB
- if DA is disabled, only export files to FXS (if it is the case), do not read
the input data from the DAQ
- add lots of Doxygen comments
(Bogdan V.)
decaro [Thu, 14 May 2009 16:01:23 +0000 (16:01 +0000)]
AliESDtrack::fTOFsignalDz corresponds to Z_track in the matched TOF pad RF. As a consequence, updated AliTOFGeometry::IsInsideThePad and AliTOFtracker(V1)::MatchTracks methods. As result, changed sign to the AliESDtrack::fTOFsignalDz content
abercuci [Thu, 14 May 2009 15:24:21 +0000 (15:24 +0000)]
- cache a larger number of elements from the Kalman covariance matrix
- calculate (optionaly) also the error on momentum per layer (from pt
and tgl covariance submatrix)
hristov [Thu, 14 May 2009 12:06:54 +0000 (12:06 +0000)]
Updated fix for bug #50143: corrected calcualtion of the number of bins with entries in the fit region. Loglikelihood fit since the number of entries is small
richterm [Thu, 14 May 2009 11:09:15 +0000 (11:09 +0000)]
avoid warning for the default HLT simulation
- changed default output data type of BlockFilter component from
'Void' to 'Any'
- bugfix in AliHLTComponent: handle component output type
kAliHLTAnyDataType correctly
richterm [Thu, 14 May 2009 10:36:05 +0000 (10:36 +0000)]
Modifications of the activation scheme for module agents and
the printout at component library loading
In order to activate only the agents of the library being loaded,
a check for library name and module id is applied. Works for
component libraries with agents following the naming scheme
libAliHLT<MOD>.so
Info messages have been adjusted to the needs in order to reduce
output. The full message is now only printed if the library
contains detailed build information from the HLT build system.