Update for Nov 12, 2008

0) I looked at the k factors in the rate dependence calculations.  These are 
factors in the probability equation.  Seamus has them set to 5 (as the number 
of veto bars).  Last week, I attempted to have them float as a free parameter, 
and got very small numbers with large errors:
   k1 = 0.000013 +/- 1.08
   k2 = 0.000002 +/- 1.38

1) What matter is how does this affect GEn, and which is a better fit?
Ignoring the error for the moment, 
1a) Floating k gives me A_phys = -0.2693
    k1=k2=5 gives me A_phys = -0.2679

1b) looked at the quality of the plotting the difference between the 
neutral/charged ratio from the data and the ratio calculated from the fit.  
Presented as:
k5veto.pdf and ksmallveto.pdf

I fit them to a constant. For k1=k2=5, I get a constant of -0.0006432 
+/- 0.0003814.  For floating k's, I get a constant of -0.000996 +/- 0.000381, 
with comparable chi-squared values.

Since results of using the different k values is small (0.5%) compared to the 
statistical uncertainty (5.0%), and Seamus's interpretation of k is reasonable,
 I will proceed using k1=k2=5.

2) Resurrected my GEn extraction code. If you recall, I am plotting the 
"asymmetry" as calculated from a lambda and the kinematic factors for each 
event which makes the cut. I than take the central value and sigma from the 
distribution, and step lambda.  I compare the resulting curve with the value 
of A_phys from my analysis.

For the plot presented as Lambda_10run.pdf, I used 10 runs.  For the 
asymmetry, I use my current A_phys, but with Seamus' error (still calculating 
my error).  I am currently running the program with all production runs from 
kin.4.  I expect a much smaller error band (about 1/3 of what we see here).

Plans for the future:

1) Present empty cell information (including T7 event analysis)
2) Nitrogen dilution
3) Error propogation

/amk

Question for the group: meeting Nov 26?  Try for Nov 25? Move to morning?