NLL in RooFit

Dear experts,

we extract NLL from the best fit point like that:

RooFitResult* r = model.fitTo(data,Range(“R1,R2”),Save());
cout << "-log(L) at minimum = " << r->minNll() << endl;

I was told by CMS combine tool experts the followig:

“Since we assume we always use likelihood ratios, there are constant terms that are neglected in the minimisation procedures both in combine and roofit, so we don’t actually know the “real” nll value, nor does roofit.”

Is it correct to what concern RooFit ?

Thanks, Sasha.

Hi @anikiten,

I am adding @jonas, our export for RooFit related questions.

Thank you !
Cheers, Sasha.