belikov [Wed, 15 Oct 2008 14:17:58 +0000 (14:17 +0000)]
The track parameters constrained to the primary vertex can now be obtained via the last argument of the RelateToVertex(...) and RelateToVertexTPC(...) functions
richterm [Wed, 15 Oct 2008 14:06:37 +0000 (14:06 +0000)]
changes to the default HLT TPC simulation configuration: send bare ESD instead of embedding it to tree, use CF with AliAltroDecoder input if rawreader is available (raw data simulated)
richterm [Wed, 15 Oct 2008 13:32:55 +0000 (13:32 +0000)]
make the buffer size check a bit less strict in order to workaround a bug/feature in the generation of the simulated raw.root files: if size of raw DDL file is not multiple of 4, additional bytes are added
ivana [Wed, 15 Oct 2008 13:04:21 +0000 (13:04 +0000)]
Bug fix in AliMUONTrack:
- Clusters potentially removable to improve the quality of the track
were not properly flagged
- Make sure there are always at least 2 clusters in stations 4 & 5 together
(Philippe P.)
cvetan [Tue, 14 Oct 2008 14:24:09 +0000 (14:24 +0000)]
1. Correct inheritance of GetEntryList in AliXMLCollection + bug fix. 2. Extension of the collection:// URI, which now can handle also entry lists. The two available formats are: a) xml collection file with event-lists (produced by the tag analysis framework); b) TEntryList stored in a root file (TEntryList object should follow the defintion of an entry-list for a TChain). The formats are distiguished by the filename extension - .xml ot .root
richterm [Tue, 14 Oct 2008 13:32:32 +0000 (13:32 +0000)]
bugfix: Data was simulated in the wrong order: bunches for the higher timebins
first. But real data is the other way round: bunches are written in the order
of ascending timebins. A new consistency check was added to AliAltroDecoder
(AliAltroBunch) in revision 29090. Now the channels are checked for overlapping
bunches, the time of a bunch must be smaller than time of the previous one
minus its length.
richterm [Sat, 11 Oct 2008 19:36:44 +0000 (19:36 +0000)]
enhanced version of the selector component for increased performance of selective Altro redaout: component now can calculate average/RMS for channels and do the selection without a component preparing the list of channels
richterm [Fri, 10 Oct 2008 12:21:33 +0000 (12:21 +0000)]
applying a quick hack to avoid usage of the pad class when only the list of hardware addresses is going to be produced; ideally something like this hack can be part of AliHLTTPCPad together with some better handling of the pad objects in order to avoid large arrays
richterm [Fri, 10 Oct 2008 06:53:09 +0000 (06:53 +0000)]
correcting some bugs, partly introduced in rev 29090
- last bunch of last channel was likely to be always skipped due to wrong size check
- putting back the function documentation which somehow got lost
- doing check for minimal buffer size right in decoder initialization
- making GetRCUTrailerSize working if decoder was not initialized or failed
richterm [Fri, 10 Oct 2008 00:18:30 +0000 (00:18 +0000)]
bugfixes: using correct format specifier (unsigned) for event no in file name; the SOR/EOR prefixes had been wrongly added in front of the directory, now added to the directory
jthaeder [Thu, 9 Oct 2008 20:36:03 +0000 (20:36 +0000)]
* AliHLTRootFilePublisher
Implementation of the RootFilePublisher( was a only dummy before )
* AliHLTFilePublisher
New member added in order to switch from raw to root files
* rootfilepublisher.C
Added example/test macro for RootFilePublisher and RootFileWriter
laphecet [Wed, 8 Oct 2008 15:11:02 +0000 (15:11 +0000)]
This version includes the suppression of the 0.5 channel shift used previously to calculate mean values (both for pedestal and calibration) and some bug corrections