CHICOS Analysiscut2

Running analysiscut2 on old data:

I wanted to look at how well the automatic reconstructor was doing and whether it would be useful to put in more cuts on quality of fit. Three things to look at are how many sites are fit with non-zero intensity, and how far off in intensity and time the worst-fit hits in the shower are.

I had one problem in getting this info because I wanted to loop over all hits in the shower object and check the fit, but when the only hits at a site are above the charge threshold of 1000 MIP, this breaks. I removed showers with this problem from the file just for the purpose of getting some statistics, but it would be good to figure out a way around this.

Number of sites in the shower fit with >= 0.1 MIP:

Distance between time average and fit for worst fit site in the shower:

Distance between intensity average and fit for worst fit site in the shower:

Correlation between intensity/time offsets and number of fit sites:

Having a hit far off in intensity fit seems to correlate with large showers. On the other hand, a hit far off in time is more common in showers with < 4 sites fit with non-zero intensity. I would propose making a cut on showers with < 4 hit sites AND one hit more than 5 us away from the fit time. This would remove 8 showers from the best candidate list:

• 06-18-03 Previously failed intensity cut. Only fits to the trigger. Passed here because I changed the check on intensity to loop over all sites instead of just all matches - since the trigger fits, it counts as one non-zero match.
• 08-17-03 This shower looks ok. Barely fails this cut.
• 06-01-04 Previously noted as a bad fit.
• 11-16-04 Could be a good shower, just has some outlying points (removed in sample reconstruction).
• 11-29-04 Previously noted as a bad fit. (Though sample is ok.)
• 02-03-05 One site 6 us off. Otherwise not bad.
• 03-11-05 One site 7 us off. Otherwise not bad.
• 06-23-06 Only fit the trigger site.
• 03-05-07 (Wrong image.)

This doesn't actually seem like a very helpful cut. It might be good to cut out all showers with < 3 sites fit AND all showers with a site off by > 5 us.