Beeston Barlow method in Roofit

Roofit and RooStats
1

Hi,

Are there any instructions for how to use the Beeston-Barlow method in Roofit?

If I understand correctly using the Beeston-Barlow method in RooFit, the MC stat uncertainty is taken into account in the fit but otherwise using the “standard” RooFit, only the template shape is used. I only found something in this post: Problem with Beeston Barlow light method

but seems this problem wasn’t resolved.

I am currently doing a log-likelihood fit in Roofit but the fit doesn’t converge in the last pt bin where there are less stats so I want to use the MC stat uncertainty in the fit. I tried using a global constraint based on this tutorial (https://root.cern.ch/root/html/tutorials/roofit/rf604_constraints.C) but that didn’t help either. The fit works only if I smooth the MC distribution which points towards this being a MC stat issue.

Any help in this regard would be appreciated!

Thanks,

ADG

2

I would also be interested in this, especially for the 2D case. I tried examples and code snippets from over the internet and from Wouter’s tutorials, but with no luck…