I’m new to RooStats and have some questions regarding HistFactory.
Here’s my use case: For my analysis I estimate the background template from data and the signal template from MC. Then I sum the two templates and fit it to my data, which returns the signal yield and the background yield. I do this with RooFit, but now I would like to repeat this with HistFactory, in order to do a profile likelihood (and include nuisance parameter).
I read the manual/example in twiki.cern.ch/twiki/pub/RooStat … lihood.pdf. On page 2, the first section of the example I see this:
SigXsecOverSM is the POI, but I’m not totally sure what the other norm factor corresponds to: "<NormFactor Name=“NEle_bbAtautau120” … ". Is this the efficiency or some MC scaling factor with the input histogram for the signal normalized to Lumi*crosssection ?
Also, I’m estimating my background from data, so NormalizeByTheory=False in that case, but I’m not sure how to normalize the input histogram for the background. I’ve tried several things, but none seem to work for me (and I think the problem is the normalization).
Any help is appreciated