I am new to RooFit but I have some experience with C++ and ROOT. Mine is a RooFit specific question. This is my first time posting to the forum but I have benefited immensely from all the past posts I have found in the forum.
Main goal: I have a single data ntuple with 5 observables in them. I have three Monte Carlo simulations ntuples with the same 5 observables, each MC representing ‘background 1’, ‘background 2’ , ‘signal’. I would like to perform a simultaneous 1D fit of the three MC across all 5 observables, using binned extended maximum likelihood.
What I have done: I have tried to follow the tutorial example here: https://root.cern.ch/root/html/tutorials/roofit/rf501_simultaneouspdf.C.html. For my MC ntuples, I have created histograms of each type (background 1, background 2, signal) and each type its own observable, and I converted them into RooHistPdfs (so I have about 15 RooHistPdf total for the MCs). For my data ntuple, I have made RooDataHist objects for each of the observables, so I have 5 RooDataHist objects, each corresponding to an observable.
My main issue is making ‘combData’ (line 92) in the example. In ‘combData’, instead of the observable ‘x’, I passed a RooArgSet with my 5 observables. When I ran it, my memory consumption spiked and it never progressed past line 92. What I think is happening is that it is trying to make a 5D histogram which I think doesn’t work. I am thinking I should use RooCategory somehow, but I am not sure how.
I am not sure how to proceed. I have looked up other forum posts but I couldn’t piece anything together. Any guidance would be great. Thanks!