In AliMUONTriggerChamberEfficiency:

When calculating efficiencies starting from the division of two histograms, a TGraphAsymmErrors is produced via TGraphAsymmErrors(const TH1* pass, const TH1* total, Option_t *option); following the guidelines of Page 27 of http://pdg.lbl.gov/2009/reviews/rpp2009-rev-statistics.pdf, the Clopper-Pearson 68% method is now used to determine errors. Explicitly specifying this option also prevents a segfault that occurs with the "tagged" v5-27-05-build1 of ROOT (rev.35249).
(F.Bossu, D.Berzano, D.Stocco, L.Bianchi)

diff --git a/MUON/AliMUONTriggerChamberEfficiency.cxx b/MUON/AliMUONTriggerChamberEfficiency.cxx
index c0796d0ce6432651d2d849b311ea3a83e0ec1bf8..9c1063d2be41d0c3ace9fef663359360161c6cea 100755 (executable)
--- a/MUON/AliMUONTriggerChamberEfficiency.cxx
+++ b/MUON/AliMUONTriggerChamberEfficiency.cxx
@@ -413,7 +413,7 @@ TGraphAsymmErrors* AliMUONTriggerChamberEfficiency::GetEfficiencyGraph(TH1* hist
   //
 
   TGraphAsymmErrors* auxGraph = 0x0;
-  if ( histoDen ) auxGraph = new TGraphAsymmErrors(histoNum,histoDen);
+  if ( histoDen ) auxGraph = new TGraphAsymmErrors(histoNum,histoDen,"cp");
   else auxGraph = new TGraphAsymmErrors(histoNum);
 
   Int_t npoints = histoNum->GetNbinsX();