[d2n-analysis-talk] Trigger Efficiency Update

[Brad Sawatzky]

On Mon, 19 Jul 2010, David Flay wrote:

> Despite this result being exactly 1 for the T3 efficiency, I think such
> behavior of the cuts is correct -- in that the T4 is S1.and.GC ||
> S2.and.GC and !(S1.and.S2) -- so these T4 events could be knock-on
> electrons, and would show up in the 1-photopeak.  So a cut on the GC
> greater than 1.5 p.e.'s would remove a lot of those (~30%, actually). 
> Similarly, the GC TDC cuts show that most these events aren't in time with
> good electron events, removing all but 4 events.  Finally, the PR cuts
> further support this, as those cuts are designed to remove the knock-on
> electrons (the pion region of the 2D energy plot for the PR).
> 
> I just find it surprising that I have this behavior for ~100 runs (which
> seems like a lot), spread out over all the runs I used from p = 0.60 to p
> = 1.70.

Your numbers are close enough (within a tenth of a percent) that I think
you can note the behavior as curious, but that uncertainty won't impact
your final cross sections at all (the error will be dominated by other
effects).

> On second thought, this run has low statistics to start with (~300k).
> If I look at a larger run (20676, ~1.4M events, p = 0.60 GeV) I get an
> efficiency of 99.96%.  So maybe this is just a statistical issue...

True for run 20157.  Don't forget that when you're getting down to only
a few hits the statistical uncertainty calculation needs to use Poisson
stats, not Gaussian (so the error on 0 hits in T4 isn't sqrt(0)).

As I say though, you've basically established that any uncertainty on
this value will have a negligible impact on the final number.

-- Brad

-- 
Brad Sawatzky, PhD <brads at jlab.org>  -<>-  Jefferson Lab / Hall C / C111
Ph: 757-269-5947  -<>-  Fax: 757-269-5235  -<>- Pager: brads-page at jlab.org
The most exciting phrase to hear in science, the one that heralds new
  discoveries, is not "Eureka!" but "That's funny..."   -- Isaac Asimov