Request for help in suggesting suitable fit for the given data

[shankar]

Dear All,

If possible can please anyone help me on what functions can I use to fit the histogram which I am attaching in this thread.
I have tried “gaus” in a selected region of x=0.07 to 0.18 (pion mass window) for the peak but the chi2 value is coming very large and it is not able to fit the peak properly,
and any suggestions on how to fit the background , or should I subtract the background.

Thank you and regards,
Shankar Nair

[Ako_b]

Hello @shankar !
As for the picture on the right, you could try to use relativistic Breit-Wigner distribution or its non-relativistic variant.
The left picture looks more complicated, and I can’t say much without context of underlying process. May be sum of Breit-Wigner distribution and polynomial background will be good enough.
Also look here

How to fit polynomail and BreitWigner function?

Hi, I try to use polynomial and breit-wigner function to fit a curve where background follows polynomial and the peak follows breit-wigner. But when I try to fit I only get a horizontal line which is much higher than data’s curve. I don’t know what’s wrong with my code. Here’s the code about fitting and the attachment is the data I use. Thanks, Zhaozhou RooRealVar mass_peak("m0","m0",60,120); RooRealVar width("width","width",0,100); RooRealVar a0("a0","",0,1000); RooRealVar x("x…

[shankar]

Hi @Ako_b ,

Thanks for the suggestions, I will try to fit it with breit wigner.

For the left side plot , the underlying process through which it is generated is the neutral pion decay channel , two photons are reconstructed in pp min bias collisions in a detector simulation.

Thanks!

[Ako_b]

@shankar

two photons are reconstructed in pp min bias collisions in a detector simulation

I am afraid that’s out of my competence. I guess, to choose appropriate model, one should know which modes contribute to signal and background, account for misreconstructed events, energy resolution effects, ISR/FSR corrections etc, which is huge work to be done.
You may wish to consult with your colleagues or supervisor who are likely to know these details or can offer a suitable empirical approach.

Besides that, I found that in this paper (look at Fig. 1b, which looks similar to peak on your left plot)

Exclusive Measurements of b -> s gamma Transition Rate and Photon Energy...

a sum of Argus PDF and Novosibirsk function was used.
Maybe it will be useful for you.

[shankar]

Thank you for your suggestions… and materials regarding fitting.