Fitting a histogram with one Breit-Wigner signal and a non-resonant background I noticed the fit parameters shown on TPaveStats are completely different, non-physical, than those ones shown in the ROOT command line. I am using version 6.36.00.
The histogram was created with a weight as in:
in your code, you do hm4rec2OSpipitt05acc->Fit(...) twice. Are you sure the parameters in the statbox and in the terminal come from the same iteration?
No, I run once the attached code mykshort.txt (cut and paste in ROOT command line).
As far as I know the information from statbox comes from the last iteration, isn’t it?