I have added a cut on core location to autoFit5. It now requires that the core of the shower be within 2 km of at least one site in the array.
Showers which are affected by this cut:
11-05-03 New. Energy quite high. Off edge of the array.This cut more showers than I intended. Also, looking at the area included in a 2km cutoff, I think a larger number would be better. Next up: 3km.
With a 3km cutoff, only 5 showers are excluded:
11-05-03 New. Energy quite high. Off edge of the array.This seems like a better set, as it agrees with showers visually identified as being outside the array.
Oddly, one additional shower passed in both of these trials (06-15-07). Not sure why.
I looked at the showers with -Log(likelihood)>200 and most of them had afterpulses or sites with multiple spread-out hits. 07-09-05 was one of these but is a very nice (large) shower. I don't think it would be helpful to put a cut on this number.
Distribution of core locations by energy bin: