I believe the logic for selecting the worst sample is flawed, not fatally though. If the first bin of the average histogram is the furthest from eta it doesn't makes sense to drop the first sample from the subset. I believe I have a fix and if I can figure out github pull requests I'll make one.
I believe the logic for selecting the worst sample is flawed, not fatally though. If the first bin of the average histogram is the furthest from eta it doesn't makes sense to drop the first sample from the subset. I believe I have a fix and if I can figure out github pull requests I'll make one.