[Frost] Target polarization issues

[Michael Dugger]

Hi,

I think I might know what is causing the disagreement between Natalie and 
I as to the target polarization direction.

Let me define the circular data in run sets as such:

Set 1: 62076-62296
Set 2: 62297-62373
Set 3: 62374-62489
Set 4: 62490-62604
Set 5: 62609-62704

Table of target polarization signs:
Set (Sign from table) (Sign I use)
1	+		+
2	-		-
3	+		+
4	+		-
5	-		+

>From the table you can see that I flip the sign of the target polarization 
for set 4 and 5

In Natalie's presentation today she showed T for each individual set. By 
using an individual set, Natalie could not compensate for the detector 
efficiency (if I understand how she obtained her plots). The detector 
efficiency in terms of azimuthal angle is ragged and causing a lot of 
problems when trying to determine T using a single target polarization.

As an example, I show the azimuthal distribution of pi+ n events at Egamma 
= 775 MeV from set 4 at
http://www.public.asu.edu/~dugger/s4.gif

Where top left panel is cosThetaCm = -0.7 and bottom right panel is
cosThetaCM = +0.7, and the curve represents a fit to the function
A*(1 + B*sin(phi + C)), where A, B, and C are fit parameters.

As can be seen on the plots, the sin(phi) dependence is more a measurement 
related to the phi acceptance than a measurement of the T asymmetry.

For the signs used in my analysis, the target polarization for set 3 is 
opposite to that of set 4, while in the table the target polarization 
signs are identical for set 3 and 4.

When I use only events from set 3 and 4 using my signs, I am able to 
add the events in the right proportions to get an unpolarized sample. I 
then can take the difference (appropriately weighted) between set 3 and 4 
in terms of azimuthal yields and divide by the unpolarized sample to 
obtain the efficiency corrected azimuthal plots for T with Egamma = 775 
MeV at
http://www.public.asu.edu/~dugger/s34.gif

where the curves are fits to the function
A*sin(phi + 60.0)

The corresponding plot for T is at
http://www.public.asu.edu/~dugger/t775.gif

If I had the wrong relative sign for the set 4, then I would not expect to 
get a result that looks as reasonable  as shown.

I'm going to recheck everything and make sure that I have not tricked 
myself into plotting the wrong sets or any other dumb mistake I might have 
made.

Take care,
Michael