Regarding the luminosity weights in RooFit and calculating the upper limit and significance using StandardHypoTestInvDemo

Dear Experts,

Probably a very stupid question. I have Monte Carlo samples of higher luminosity (5 ab-1) than data (100 fb-1). So I am trying to do some weighted fitting for MC (all the luminosity and all other possible trigger weights etc are taken into account) using RooFit and then I am giving this weighted RooDataSet and model to StandardHypoTestInvDemo to calculate upper limit and also to p0plot for significance. So, I was asked not use the luminosity weights in the fitting and take a random subsample from MC equivalent to data and do the rest (or I could do it using toys). And I was told that it’s because due to statistical fluctuations will not be same (for 100 fb-1) if I include luminosity weights in fitting. So my question is, could it be possible to include the luminosity weights in the fitting as a weight? Could the RooFit and upper limit calculator I am using take into account the fluctuations equivalent to data (just for the information I am using SumW2Error(kTRUE) in RooFit but I think (StandardHypoTestInvDemo could take that into account too))?

Any suggestions regarding that will be helpful.

Many Thanks in advance.


I guess @moneta may have some suggestions.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.