-
Notifications
You must be signed in to change notification settings - Fork 75
Open
Description
We have a fraction accumulator now, but it does not work correctly for weighted samples. However, we definitely need this in high-energy physics because we often estimate the efficiency of some detector from weighted simulation samples.
There are several ways how this could be implemented
- New class
weighted_fraction<T> - Specialization
fraction<weighted_sum<T>> - Template specialization
fraction<T, bool weight_support = False>
Number 3) is my least favored. I like 2), if that's feasible.
gohil-jay and henryiii