Optimal way to compare and validate simulation and experimental data

Hi All,

I currently have two histograms, both TH1F, 75 bins, one experimental data, the other data from a Geant4 simulation. The histograms represent frequency of photoelectrons detected in a photomultiplier tube; counts vs number of photoelectrons.

I am curious to learn of any built in ROOT methods that will allow me to compare these two datasets, and determine if the simulation is a decent model of the experiment. All and any suggestions are greatly appreciated.

Thank you in advance!


ROOT Version: 6.22/06
Platform: MacOS Big Sur
Compiler: gcc



Thanks so much! With regards to the chi-squared test, would you happen to know or have any suggestion as to how to determine the degrees of freedom for my data?

From my understanding, given that I have 75 bins, each representing a number of photoelectrons, that the degrees of freedom would be: 75-1 = 74. Is this wildly off?

Thanks again for all of your help!

You can use the "P" TH1::Chi2Test option to print “NDF”.
I guess: “NDF” = “number of histogram bins” - 1 - “number of bins which are 0 in both histograms simultaneously”.

@Wile_E_Coyote ,

Thank you so much. I misread the reference for the chi squared test you provided. I thought that the user would have to supply the degrees of freedom.

Thank you again for your time and help.