Replies: 1 comment
-
Hi Massimiliano, that is a really interesting thing to do. (I am just traveling so I'll gladly have a more detailed look in two weeks from now). I think this is the right way to go, indeed. You can use Gaussian constraints or arbitrary constraints, either created by yourself or by using the base class for TFP distributions and wrapping any available distribution here. Please just come forward with any question, I don't see any any problem with doing it but I am interested to see how it works myself. And, be picky with the zfit and hepstats code, it may be that there are things which are missing in the library (but they should, hopefully, be well readable. |
Beta Was this translation helpful? Give feedback.
-
Hi, I'm trying to implement the exercise described here using zfit and hepstats.
Combine is the tool which is used for most of the Higgs related analyses in CMS and it is based on the concept of datacard to provide the information necessary for the fit. A description of the main fields provided in a simple 1 bin counting experiment is shown in the link.
Here is my question: how can the systematic uncertainties be implemented in zfit? Quoting from the link:
I was thinking, for this simple case, that adding a
zfit.constraint.GaussianConstraint
for each systematic might be the solution, but I'm not sure.For completeness, a more detailed description of the types of systematic uncertainties is provided here.
Thanks,
Massimiliano
Beta Was this translation helpful? Give feedback.
All reactions