Likelyhood scan with TFeldmanCousin

Dear experts,

I’ve been using TFeldmanCousin class to calculate trusted intervals for the number of signal events. This appeared to be simple, and the following lines handle this task:

f = ROOT.TFeldmanCousins()
ul = f.CalculateUpperLimit(int_val_scaled_all, int_val_scaled_bkg)
ll = f.GetLowerLimit()

Although, now I’m trying to extract a likelihood curve from it to convolute it with a Gaussian, which will account the systematic uncertainties. And I cannot find anything useful within the implemented methods for v6.26:
https://root.cern.ch/doc/v626/classTFeldmanCousins.html

Could you please advise how it can be done? How may I extract the likelihood curve from a TFeldmanCousin?

I’ve found this discussion, where a similar question is reviewed. But I still cannot come up with the method to implement it technically.

D.

Dear @mc.dmytro ,

Let me ask @moneta to help you on this one.
Cheers,
Vincenzo