Systematic error bar per bin

Newbie
1

Hello Rooters,

I need your help to figure out this. I have two histograms as follows.

h1->Fill(variable, weight);
h1->sumW2();
h2->Fill(variable, weight_uncertainty);

What could be the total statistics + systematics error bar per bin?

= sqrt{err_weight(bin i)^2 + err_syst(bin i)^2}
or just
= err_syst(bin i)

err_weight(bin i) = error value of bin i of variable_weight
and
err_syst(bin i) = error value of bin i of variable_weight_uncertainty

thnaks
Dil

2

Hi @Dil ,
if I understand correctly, this is less a question about ROOT and more a question about statistics. Let’s see if @moneta has some suggestions, otherwise you should ask your research group’s statistics expert.

Cheers,
Enrico

3

Hi,
The total error (statistical one and the one given by the weights uncertainties) can be approximated with
sqrt{err_weight(bin i)^2 + err_syst(bin i)^2} using your definitions or err_weight and err_syst.

Lorenzo

4

Thank you Lorenzo.