TGraphAsymmErrors::Divide -> SetBinContent leads to wrong results

Hi,

I am using TGraphAsymmErrors::BayesDivide to calculate cut efficiencies.

Under certain circumstances I'd like to mask out some bins of my histogram by setting both enumerator and denominator to zero using SetBinContent.

This breaks TGraphAsymmErrors::(Bayes)Divide, because then the built-in flag bEffective is (I guess) set to true leading to error messages, wrong results etc. Unfortunately, there is no way of steering bEffective from outside the function.

Is there a way to amend this?

Thanks,
Alexander Received on Tue Feb 08 2011 - 17:57:54 CET