This is related to #152 , where we found add eps to the histogram ruins the fitting, because of the added eps=1, which could be significantly larger than efficiency when efficiency is very low. In this PR the eps is changed to mean(eff) to avoid such a problem.
This is related to #152 , where we found add eps to the histogram ruins the fitting, because of the added eps=1, which could be significantly larger than efficiency when efficiency is very low. In this PR the eps is changed to mean(eff) to avoid such a problem.