I have a model where the normalization (coefficient) and the PDF itself depends on the same nuisance parameter. It looks like Roofit make sure this does not occur by throwing this error.
ERROR: coefficient nSig and PDF SigPDF have one or more dependents in common
I was wondering if there’s a way around this. There are nuisance parameters in our signal model that will change both the shape and normalization of the signal. We like to have the nuisance parameter vary and update the signal shape and normalization at the same time in the fit.
Here’s an older post that stated this Error.