Error bars in TGraphAsymmErrors


how are the error bars in TGraphAsymmErrors computed?

I understand it is integrating a Binomial distribution…

Can you point me to any documentation that can explain it to me in some detail?

Thanks a lot


there is no error computation in TGraphAsymmErrors; the error bars are set through its constructor. The only exception is … ayesDivide and the constructor taking two TH1 which uses BayesDivide(), that function is well documented.

Cheers, Axel.