On the advice of Rob, I spent the last two weeks continuing on with studying
GEn.

I spent a couple days checking the rest of the tracking code, which I'm
confident is now working and usable.  Rob and I also spent some time on
producing a new set of optics coefficients for the replays.  These improvements
will allow for more accurate analysis since we do not have any underlying tails
in our distributions that come from poor fits.  Rob had replayed the H2 runs
at this point as the start of Pass 2.

We spent a good part of last week figuring out how to make our raw asymmetries
agree (Rob had 4%, I had a lower 3% number).  The major change that put our
numbers in agreement was placing stricter cuts on neutron acceptance.  I had
also not worked in the possibility that the vetos have a 30ns deadtime where
they are blind.  This has been added to the code.

The rest of the time was working on a suggestion from Gregg so that we can
try and understand the background and proton leakage rates in the data.
Gregg's suggestion was to, for every event, keep the BigBite data and take
the neutron arm data from the previous event.  This should give some
description of what the background looks like.

The following analysis was done on all available H2 runs 4425, 4426, 4427,
4496, 4497 that are replayed using the new code.

Pages 1-3 are the p_miss_par, p_miss_perp, and W plots as they are from the
data.

Pages 4-6 are these same plots except representing the background

Pages 7-9 are the background subtracted plots.

In H2, we are interested in the proton leakage rate which should be contained
in the ratio between the uncharged and charged plots after background
subtraction.  These are plots 10-12.  The full distribution has been overlayed
for convenience.

What is notable here is that the proton leakage rate is roughly around 2 or 3%
for the areas of interest, which is much lower than the 8% to 12% numbers we
have been using in the past.  This number is also consistent with what was
expected.

This tells us that there are significant numbers of particles that are flagged
as neutral that are purely from background.  This is very likely to be true in
the 3He data as well, so we must be careful in determining what the dilution
factor is.  This can be done using similar techniques that I have applied here.

Since I believe this method to be incredibly useful, I have been rewriting my
asymmetry code to compensate for it.  I have yet to work in a method for
effectively combining clusters in the regions of interest, so the results here
may be skewed a bit.

Also, I am working into my code the ability that is similar to Rob's to find
asymmetries as a function of p_miss_par, p_miss_perp, and W.  

I will continue on with this type of analysis and try and come up with some 
background dilution numbers for the 3He data.  After doing this we should be 
able to come up with a much more accurate number for GEn.

Seamus