wiechula [Sat, 22 Feb 2014 17:03:50 +0000 (18:03 +0100)]
Fix crash that occurred after switching to git. In some circumstances (checkout a tag) the AliROOT version string is not present. Seems to be a cmake bug and can be circumvented creating an explicit branch from a tag, e.g. git checkout -b v5-05-AN-68 v5-05-AN-68 (Thanks Alina)
mkrzewic [Sun, 23 Feb 2014 16:36:41 +0000 (17:36 +0100)]
can run without a config file (with sane defaults)
input format changed: no positional parameters anymore
to see docs, just run the script without args
bhess [Wed, 19 Feb 2014 12:33:42 +0000 (13:33 +0100)]
- Added loose rejection of light nuclei - Better treatment of PID probs w.r.t. light nuclei - Replaced most probable PID in data at low pT to some manual TPC only approach to get pure dEdx templates this way
- Special treatment if most prob PID is light nucleus only if well above el expectation (in order not to mix up at ID at very high pT)
ppillot [Wed, 19 Feb 2014 16:50:51 +0000 (17:50 +0100)]
fix finding of pad neighbours; remove methods to write them in OCDB
Round slates are rotated compared to others so the pad density
increases instead of decreasing when going from left to right:
--> increased number of possible neighbours from 10 to 12 to be
full symmetric (also consider increasing density from bottom to top)
--> corrected possibility to store the neighbours at their
corresponding position in the array (holes filled with invalid pads)
shabetai [Tue, 18 Feb 2014 13:20:48 +0000 (14:20 +0100)]
Initial implementation of the DCAL Geometry (work done by Mengliang WANG for his service task). 3 versions have been implemented: - EMCAL_COMPLETE12SMV1_DCAL: 12 EMCAL SM + 6 DCAL SM - EMCAL_COMPLETE12SMV1_DCAL_DEV: 12 EMCAL SM shifted and 10 DCAL SM (this is a possible future DCAL geometry) - EMCAL_COMPLETE12SMV1_DCAL_8SM: 12 EMCAL SM and 8 DCAL SM (the current DCAL geometry)
gconesab [Wed, 19 Feb 2014 16:48:06 +0000 (17:48 +0100)]
add setters for L1 trigger threshold and Bits, in case the automatic ones are not correct, change how the L1 automatic threshold is recovered in case of old AOD format
rgrosso [Tue, 4 Feb 2014 10:00:31 +0000 (11:00 +0100)]
Check CVMFS_OCDB and deal with storages accordingly
In SetDefaultStorage check environment variable triggering cvmfs case:
CVMFS_OCDB. It is supposed to be set to the cvmfs path, containing
OCDB tags (list of all files seen in each OCDB at the moment when
the AliRoot tag was created) and two scripts, to get the year from the
run number and to extract the files valid for the given run.
The method ValidateCvmfsCase validates the $OCDB_PATH path
// and modify Uri if it is "raw://"
The "alien://" uri passed to GetStorage is converted into a
"local:///cvmfs" uri inside CreateParameter called by GetStorage, via
AlienToCvmfsUri.
E.g.: OCDB_PATH="/cvmfs/alice.cern.ch/Packages/v5-04-Rev-12"
Example of expected subtree inside OCDB_PATH:
Ideal.list.gz
MC/Residual.list.gz
Full.list.gz
2009.list.gz
data/.......
2013.list.gz
bin/getUriFromYear.sh
getOCDBFilesPerRun.sh
In addition:
Set indent to standard two spaces
Fixed indentation of comments foll. method header
Get only highest sub+version in L::QueryValidFiles
Let AliCDBLocal::GetId look in the list of valid fileIds and let
AliCDBLocal::GetEntry get from the list of valid FileIds
SetRun sets AliEn URI via SetDefaultStorageFromRun also in
CVMFS case, via getUriFromYear.sh script