I have a question regarding the return value of ROOT::Fit::FitResult::MinFcnValue().
I’m performing the fit with ROOT::Fit::Fitter::LikelihoodFit() to binned data, but it is not clear to me if what I get with MinFcnValue() is the value of the likelihood itself or the negative log likelihood.
I would really appreciate the clarification of an expert.
Sorry if the answer is already somewhere. I’ve tried to look for the answer elsewhere in the forum but I am still confused. Thanks!
It is the minimum of the negative log likelihood (NLL). You will almost never see the normal likelihood in software like this, it’s just much better to use the NLL because it’s approximately parabolic and a big sum instead of a product. That’s better for numerics.