phille [Mon, 12 Feb 2007 09:04:45 +0000 (09:04 +0000)]
Bug fix: When moving from C style pointer based strings to the C++ string
class the memebr initialization list was not updated accordingly.
The member fDirectory and fFilename was initialized to zero wich
is fine for a C pointerbased string but when used to initialize a string
with the statement fDirctory(0) it is intepret as an attemt to initialize
a string object with a C character string starting at 0.
The error cuase a runtime error of aliroot, but is now fixed.
gustavo [Fri, 9 Feb 2007 16:00:39 +0000 (16:00 +0000)]
Calibration data member fCalibData initialized in Init() calling to the AliEMCALLoader::CalibData() method, pedestals and gains always data recovered from CDB
gustavo [Fri, 9 Feb 2007 15:57:21 +0000 (15:57 +0000)]
New calibration data member with AliEMCALCalibData instance. Initialized in Init() calling to the AliEMCALLoader::CalibData() method, pedestals and gains always data recovered from CDB
gustavo [Fri, 9 Feb 2007 15:34:23 +0000 (15:34 +0000)]
Reference calibration histogram file, one histogram per channel with accumulated amplitudes. This is a version for shuttle testing, moved from EMCAL/data to EMCAL/ShuttleInput
phille [Fri, 9 Feb 2007 15:28:28 +0000 (15:28 +0000)]
HLT component to write data to files. The directory where to write the files
has to be specified as component argumet with the -directory <directory> option.
At least one data type has to be specified as component argumet with the
-datatype <datatype>. Any number of valid datatype can be specified.
Example of usage.
The command abowe will write both celle energies and untouched raw RCU data to separate files in the
directory /home/perthi/HLT/phos/data/. Currently one file will be written per event.
phille [Fri, 9 Feb 2007 14:07:02 +0000 (14:07 +0000)]
Bug fix:
1) The path to the peakfinder vector is now taken programatically from
the environment variable ALICE_ROOT, and PF_VECTOR_DIR constant given in
AliHLTPHOSCommonDefs.h. The name of the file is constructed from the constants
PF_DEFAULT_N_SAMPLE, PF_DEFAULT_STARTINDEX, DEFAULT_TAU, DEFAULT_FS also
defined in ALIHLTPHOSCommonDefs.h .
The default PF vector found this way can be overiiden by the user (programtically).
2) Error messages are writen to the HLT
log file of the component rather than to stdout.
ivana [Fri, 9 Feb 2007 12:24:55 +0000 (12:24 +0000)]
Removing use of gAlice from reconstruction;
the MUONSegmentation is now created via MUONSegFactory in
MUONReconstructor and passed to AliMUONEventRecoCombi and
AliMUONDetElement objects.
cvetan [Thu, 8 Feb 2007 13:59:31 +0000 (13:59 +0000)]
The folder which should be used to store the guid files is added as an option to the alimdc API. In case that the option is empty, the guid files will be stored in the same folder as raw-data files.
ivana [Thu, 8 Feb 2007 12:25:59 +0000 (12:25 +0000)]
- treating the closing of the popup window properly
- add an iterator (Next button) for DetElemId
- add more infos while drawing (pad size & motif size)
(Christian)
phille [Wed, 7 Feb 2007 14:55:34 +0000 (14:55 +0000)]
Bug fix, path to peakfinder vectors was hardcoded into the code for debugging purphoses
but was not removed when the code was commited. the error message reads
AliHLTPHOSRawAnalyzerPeakFinderComponent, ERROR: could not open PF vector file
and can be ignored unless one uses the PeakFinderComponent within the HLT framework.
The vectors can also be set with the SetAVector and SetTVector int AliHLTRawAnalyzerPeakFinder class.
The error message is commented out in the current version.
masera [Wed, 7 Feb 2007 12:08:54 +0000 (12:08 +0000)]
Bug fix (L.Gaudichet): SPD chip used instead of sensor. In the present configuration the chip is 150 um thick - it was 200 um - and the sensor is 200 um thick. This bug was seen as a bias in the reconstructed vertex position.
richterm [Wed, 7 Feb 2007 10:24:18 +0000 (10:24 +0000)]
build system: configure checks added
- for version of AliExternalTrackParam needed by AliHLTTPCTrack conversion to ESD format
- for AliCaloRawStream.h needed by PHOS
ivana [Tue, 6 Feb 2007 17:18:07 +0000 (17:18 +0000)]
Commenting out StdoutToAliDebug which is, with the current AliLog
implementation not unpluggeable, even with LOG_NO_DEBUG compilation option,
and is a big performance eater...
(Laurent)