Hi, All,
I am trying to take an input PDF over energy, E, and “smear” it by taking a conditional RooProdPDF with a Gaussian with energy-dependant resolution. The product is conditional since the gaussian depends on a dummy variable used to parameterize the resolution. I then try to project out the the dummy variable using RooAbsPDF::createProjection(). When I do this, I do in fact get a PDF with resolution varying as I want; however, the scale of the PDF is on the order of 1e-243! I think this is causing problems when I try to fit later on to a data set, not to mention that I don’t think this behavior is normal. I’ve attached my code and a plot showing the behavior, and I could use any advice. The plot shows the resolution vs. energy, the unsmeared PDF, the smeared PDF, and the attempt to fit the smeared pdf to a dataset.
Thanks,
Simon
resSmearTest6_4.C (5.74 KB)