Hi all,
I may be asking about a fairly standard procedure in RooFit, but just haven’t found it in the manual, so apologies in advance if I just missed it.
I would like to provide my own events for the normalization of my PDF. For example, suppose I have some data I am trying to fit that I know came from a Gaussian distribution, but convolved with some acceptance function. If I can parametrize the acceptance, I can write this into my PDF and RooFit will handle the normalization for me, easy peasy.
But I can also choose to attack this problem by running a separate set of flat events through the acceptance function (or MC simulation) and use these events for normalization. At that point my PDF is just a Gaussian and the acceptance is handled by the normalization events.
Has anyone done this before with RooFit? It’s a handy procedure that allows one to do away with binning and formal acceptance functions and I’ve been successful with it in the past…but I’ve written my own interfaces to Minuit to handle this. I’d like to try this with RooFit if possible. Thanks in advance!
Matt