gconesab [Tue, 25 Nov 2014 09:39:26 +0000 (10:39 +0100)]
Place the rejection of dead channels in the proper location, currently it can produce many channels rejection if a dead channel is declared because the index to check the next SDigit is not updated
zconesa [Mon, 24 Nov 2014 21:13:13 +0000 (22:13 +0100)]
Updates: (1)added the check on the z coordinate of MC vertex, (2) requirement to have only 2 Lc daughters to reject decays in three prongs in the CF, (3) update the definition of TMVA cuts (A.Alici, C.Zampolli)
ppillot [Mon, 24 Nov 2014 16:44:34 +0000 (17:44 +0100)]
new version to compute trk eff from output of AliAnalysisTaskMuonTrackingEff
- Compute efficiency at DE, chamber, station and spectro levels vs various variables
- 2 options for error calculation: using frequentist or bayesian method
- improve error propagation at station and spectro levels
- integrate efficiency and error over run (if run-by-run weights provided)
ppillot [Tue, 28 Jan 2014 11:34:54 +0000 (12:34 +0100)]
new version to measure efficiency at pad/MANU/BP/DE/Chamber levels
- added possibility to use trackCuts
- added possibility to select a centrality range
- added possibility to set Align and RecoParam specific storages
- removed histograms at DE level
- added an axis phi in histograms at chamber level
- added counters to record clusters at pad/MANU/BP/DE/Chamber levels
- added extra histograms
- added possibility to display clusters in mchview in terminate (create a separate file)
- improved searching for missing clusters
- corrected treatment of overlaps
wiechula [Tue, 11 Nov 2014 14:21:23 +0000 (15:21 +0100)]
o add Reset function to CalPad and CalROC o Add functionality to AliTPCdataQA - ResetData function - Setter for minium charge in cluster - Setter for requiring a neighbouring pad
mivanov [Tue, 4 Nov 2014 06:39:28 +0000 (07:39 +0100)]
ATO-97, ATO-78 - implemented convolution of calibration parameters. Used in order to correlate Pulser (electronic) and Krypton (combined: gas gain + electronic) gain calibration
AliTPCCalROC.cxx - handling of edge effects for the median and LTM filters (http://en.wikipedia.org/wiki/Median_filter#Edge_preservation_properties) + mirroring values at the boundaries
AliTPCCalROC.h - check the ranges
test/UnitTest.C - check invariants of median and LTM filters - assert in case of failures
Filters:
1.) Bool_t MedianFilter(Int_t deltaRow, Int_t deltaPad); - Replace the pad calibration with median in neibrhood
2.) Bool_t LTMFilter(Int_t deltaRow, Int_t deltaPad, Float_t fraction, Int_t type) - Replace the pad calibration with LTM statistic in neiborhood
mivanov [Sat, 8 Nov 2014 16:46:26 +0000 (17:46 +0100)]
ATO-98 1.) distinguish between the refX radius and extrapolation radius 2.) Adding test to measure the speadd of fast parabolic fit and helix fit (CPU is the same, time spent in the evaluation of distortion not in the fit)
mivanov [Thu, 6 Nov 2014 07:07:13 +0000 (08:07 +0100)]
ATO-98 bfield(omega tau) used for the diestortion calculation used as an argument of fits (before taken from the OCDB). Modification made to enable fitting of residuals for differnt B field in the same AliRoot session.
mfasel [Tue, 18 Nov 2014 15:27:28 +0000 (16:27 +0100)]
Changes relative to restructuring of the JSON core
====================================================
- Use renamed JSON types in the macro
- Call JSON trigger maker add macro in the Jet preparation macro
mfasel [Tue, 18 Nov 2014 13:25:54 +0000 (14:25 +0100)]
Refactoring of the package
============================================
1) Move core json functionality to PWG/Tools
2) Simplify naming convention
3) Build EMCAL JSON Configuration on top of core JSON
dstocco [Wed, 19 Nov 2014 18:35:45 +0000 (19:35 +0100)]
Set ill-formed small numbers to 0
The LHgrid files contains some very small values, which should be 0.
If the value exponent is a negative number with three digits
the exponential notation is dropped.
This seems to be harmless, except when running pre-compiled code on grid.
mivanov [Thu, 20 Nov 2014 20:13:53 +0000 (21:13 +0100)]
ATO-91 Update in TPC-QA trending: MIPattach now part of dEdx descision in status bar. Status bar descisions written as friend tree to trending.root (creates trending.backup.root before). Plot ranges mostly variable now. (P.Reichelt)