jgrosseo [Tue, 23 Nov 2010 16:26:03 +0000 (16:26 +0000)]
fixing bug in par file creation when package name was equally to a directory name (occured for EventMixing)
added par file support for EventMixing library
pulvir [Tue, 23 Nov 2010 15:56:26 +0000 (15:56 +0000)]
Added the Print() method to the cut objects, which allows the AliRsnCutSet to print some information about the cut being added, just to allow a final cross-check when configuring the analysis that the cuts are corectly configured
martinez [Mon, 22 Nov 2010 22:14:52 +0000 (22:14 +0000)]
New task to study fake tracks in simulation: need to access the recoParam used during the reconstruction; associate reco & MC tracks either by using MC labels or by comparing the position of the clusters (default); fill counters to get the detailed statistics of good tracks, fakes, events containing fakes, ... ; fill histograms to compare the chi2, number of associated clusters, pt, DCA, ... of good tracks and fakes (Philippe P.)
agheata [Mon, 22 Nov 2010 10:59:27 +0000 (10:59 +0000)]
A task can now declare the list of input branches that are requested during analysis.
The way to do this is to add in the constructor where the output slots are defined:
fBranchNames = "ESD:bre1,bre2,...,breN AOD:bra1,bra2,...,braM";
The declared branches are currently not loaded automatically, the list being only used to make sure we don't combine in a train tasks that are using the branch loading feature with others that expect the full event.
A task using this should manually request loading of needed branches during UserExec - see example below.
This commit modifies the AliPhysicsSelectionTask to use this feature, gaining a factor of 2 in speed when run standalone.
//________________________________________________________________________
AliAnalysisTaskPt::AliAnalysisTaskPt(const char *name)
: AliAnalysisTaskSE(name), fEvent(0), fOutputList(0), fHistPt(0)
{
// Constructor
// Define input and output slots here
// Input slot #0 works with a TChain
DefineInput(0, TChain::Class());
// Output slot #0 id reserved by the base class for AOD
// Output slot #1 writes into a TH1 container
DefineOutput(1, TList::Class());
fBranchNames = "ESD:AliESDRun.,AliESDHeader.,Tracks";
}
//________________________________________________________________________
void AliAnalysisTaskPt::UserExec(Option_t *)
{
// Main loop
// Called for each event
// Post output data.
fEvent = InputEvent();
if (!fEvent) {
printf("ERROR: Event not available\n");
return;
}
// Track loop to fill a pT spectrum.
am->LoadBranch("Tracks");
Int_t ntracks = fEvent->GetNumberOfTracks();
printf("There are %d tracks in this event\n", ntracks);
for (Int_t iTracks = 0; iTracks < ntracks; iTracks++) {
...
}
martinez [Mon, 22 Nov 2010 06:40:41 +0000 (06:40 +0000)]
- Remove the holes in the lists of output histograms to make the merging working with the new root version. Add a new compiled macro (MuonResolution.C) to run the task. Add the possibility to fit the residuals to extract the resolution (now the default option instead of taking the RMS). Add the possibility to select events of a given trigger class (MUON, MB, …). Add the possibility to select tracks in the geometrical acceptance (passing Rabs and eta cuts). Add mchview objects to the output file to display DE resolution and systematic shifts (Philippe P.)
cnattras [Sun, 21 Nov 2010 14:27:27 +0000 (14:27 +0000)]
Fixing some more coding violations, fix in reweighting of V0s, changed GetCorrections to use TPC+ITS tracks, changed AliAnalysisHadEtReconstructed to look at TPC+ITS tracks
gconesab [Sun, 21 Nov 2010 11:13:45 +0000 (11:13 +0000)]
AliAnaPi0: Change most of TH3 to TH2
Change fixed Vertex histogram dimension to smaller ones
SetOwner of mixing list to FALSE in AliAnaPi0 and AliAnaOmegaToPi0Gamma
AliCaloTrackReader: Add option for analysis of filtered data with calorimeter only information
Updated analysis configuration macros
cnattras [Sat, 20 Nov 2010 23:12:07 +0000 (23:12 +0000)]
implementing realistic reweighting of V0s, fixing some minor coding violations, adding hadronic corrections to AliAnalysisEtReconstructed, config macros, and ITS only and ITSTPC track cuts to AliAnalysisTaskTotEt, now requires newer aliroot version because it uses default cuts for 2010
zampolli [Sat, 20 Nov 2010 16:32:24 +0000 (16:32 +0000)]
Implementing LTU configuration retrieval from DAQ logbook (logbook_detectors table),
to be read from the GRP preprocessor, and stored in the OCDB in /GRP/CTP/LTUConfig.
The validity of the object is [current_run, current_run], even if the entry committed
in AliRoot has infinite validity.
Only runs taken in global partitions will now be considered to fill the entry (a dummy
entry should be implemented as for the other CTP entries).
gconesab [Fri, 19 Nov 2010 20:10:25 +0000 (20:10 +0000)]
Correct the setting of geometry matrices from ESDs
Fill cell position histograms only when needed.
Change information print in the copy of the Event Statistics histogram
pulvir [Fri, 19 Nov 2010 10:06:29 +0000 (10:06 +0000)]
Update of package classes.
Introduced a common base class for cuts and values, which allows to generalize the evaluation methods, and select the target "a posteriori".
Generalized the AliRsnValue computation with the addition of a generic cut support object which can be any TObject (this copes with the need
to have an AliRsnPair in some cuts related to the mother, and AliESDtrackCuts for multiplicity computation.
In processInfo adjusted some indentation and adapted also to PROOF.
Adapted some header comments and removed some useless methods.
cnattras [Thu, 18 Nov 2010 17:58:18 +0000 (17:58 +0000)]
Adding common base class for AliAnalysisEt and AliAnalysisHadEt so we are not repeating code, added config files and options to config had code, updating alien handlers and task driver macros to include base class
dainese [Thu, 18 Nov 2010 14:01:13 +0000 (14:01 +0000)]
Update by Zaida
1) the systematics uncertainties class: I added a TGrapAsymmErrors to plot the global asymmetric uncertainties and had to re-order the plotting again
2) the correction class :
-- the theoretical spectrum will now always be rebinned
-- the acc x eff statistical uncertainties propagation are temporarily neglected (commented out) until we will decide if we want them in