abercuci [Fri, 13 Aug 2010 07:11:03 +0000 (07:11 +0000)]
new version of the V0 stuff for the RefMaker (Alex W.)
A new member of AliTRDtrackInfo::AliESDinfo added which has the v0pid info.
One can access it via AliTRDtrackInfo::AliESDinfo::GetV0pid(). In
addition the structure of the infoGen had to be changed. Now the first loop
is over the v0s, the second over the ESDtrackss. In case an ESDtrack has
a valid v0pid it is set, otherwise the values stay at 0 for all species.
RefMaker only depends on ESDinfo and not on Friends. Also some
modifications for the RefMakerNN.
bug fixes:
1. TObjArrays like fTracksBarrel etc. are deleted now at the beginning
of the UserExec. Otherwise one gets v0pid values for tracks which do not
origin from a valid v0.
2. in the AliTRDv0info. Now there is a protection in case not both
tracks of a V0 are found (due to what reason ever).
hristov [Thu, 12 Aug 2010 16:29:28 +0000 (16:29 +0000)]
Changes from Constantin:
a) Makefile
The changes are: treatment of ROOT to be external/system
treatment for warnings (using -isystem)
b) Makefile
Since rootcint produces obsolete string code I propose
to add -Wno-write-strings to CXXFLAGSNO. I have talked
long time ago (6 months maybe or more) to Phillipe,
but I dont think he fixed cint.
c) Makefile
Different logic to implement whether date-config is there,
so that no annoying output from the stderr of which is
produced.
d) build/module.tpl
The changes in module.tpl use ${MUTE }to avoid the
printout of several lines when not requested with ALIVERBOSE=1.
e) JETAN
Patch for JETAN to treat FASTJET external like ROOT.
Changes from Marco:
Fix for bug #71073: Relative path for 'include' in dependency files
cblume [Thu, 12 Aug 2010 14:21:30 +0000 (14:21 +0000)]
- use static const arrays instead of const arrays to avoid the copying from static memory to the stack when entering the scope
- dentangle the usage of HLT mode and transformation skip in order to test the procedures individually
snelling [Wed, 11 Aug 2010 21:12:34 +0000 (21:12 +0000)]
All formulas now work for any case of the overlap between POI1, POI2 and RP in the correlator cos[n(psi1+psi2-2phi3)] (remember that psi always labels azimuth of POI and phi azimuth of RP). Since this correlator (relevant for Strong Parity Violation study) is evaluated vs (pt1+pt2)/2 and |pt1-pt2| two nested loops are needed and by default this calculation is switched off.
shahoian [Wed, 11 Aug 2010 14:20:25 +0000 (14:20 +0000)]
1) Added method SetRunMultFinder to allow swithing off the Mult.Finder
2) Added protection: if fRunMultFinder is TRUE but the ITS reconstruction is
not done, then the warning is printed (only once) and fRunMultFinder is set
to FALSE
cvetan [Wed, 11 Aug 2010 12:13:11 +0000 (12:13 +0000)]
1. Loading the LHC clock-phase in the reco. 2. Methods to get the mean phase B1 & B2 or the phase at given time-stamp. 3. Chnages in the VZERO code which use the phase from OCDB. For the moment we rely on the mean phase between the two beams.
jgrosseo [Wed, 11 Aug 2010 08:36:44 +0000 (08:36 +0000)]
From Sara and Jan Fiete:
underlying event analysis w.r.t leading track
although there are some new classes a lot of the analysis code has been refactored from the
previous classes
new "in content" are the correction classes AliUEHist, AliUEHistograms
gconesab [Tue, 10 Aug 2010 16:37:58 +0000 (16:37 +0000)]
Define Clear for AliEMCALDigits to delete the owned arrays when the TClonesArray Clear is called. change the call to the TClonesArray::Clear to Clear("C") to call the clear of the digits.
aszostak [Tue, 10 Aug 2010 09:38:21 +0000 (09:38 +0000)]
Cleaning up ClassDef. Should have the version == 0 for the algorithmic classes like the component classes. Only data containers that should be streamed by ROOT should have the version set > 0.
git://git.uio.no / u / mrichter / AliRoot.git / log