Meaning of FCN value

Hi all,

I was wondering about the meaning of the FCN value returned by a fit.
When I fit a distribution with

corresponds the FCN value given by Minuit correspond to log(L)? Or is it just an “arbitrary” number?

And what is the meaning when I fit with


Thanks for any help on this.




In the first case the FCN is the log - likelihood built using Poisson probabilities of each bin and it is stripped by the constant term (not depending on the function parameters)

In the second case the FCN is the sum square of the residuals of each bin, normalized by the bin error

Best Regards