Dear all,
i’m in trouble to get correct erros using Divide on scaled histograms.
i get bin errors as they were in the un-scaled case.
Attached a simple macro to reproduce this.
running scale_errors.C you will get 10 bin errors for un-scaled case
as they are from histogram (bin) and calculated from binomial stat (calc).
Then same thing for scaled histograms with factor 100
As you can see errors from divided histograms are as before while, of course, those calculated with formula are lower.
Any hint ?
Many thanks,
Max
scale_errors.C (1.03 KB)