3He(e,e'd) Asymmetries

1.) Analyzed all (vast majority) of data.

I have now on my computer rootfiles of 961 production runs. Now I can determine asymmetries for all kinematics.

Rotation of the results

Here is a short pdf with the description of the coordinate system and used rotation. I also included a check, that confirms that at least elastic asymmetries can be rotated. The main problem with these rotation is the additional error on the results due to the uncertainty of the angle:
Asymmetry1.pdf

Here are ROTATED results from last week:

01.)

02.)

03.)

04.)

New Results for (BB, HRSL)=(-75, 12.5)deg:

05.)

06.)

New Results for (BB, HRSL)=(-75, 14.5)deg:

07.)

08.)

New Results for (BB, HRSL)=(-82, 14.5)deg:

09.)

These data have same kinematics settings as (75, 14.5)deg data. The only difference between the two is the \theta_{pq} angle. Can this cause observed difference in the results.

10.)

11.)

3GeV Longitudinal Results:

12.)

13.)

Elastic Proton asymmetry:

I was trying to determine the polarization of protons inside the 3He. Here is a short description of the procedure:
Asymmetry2.pdf

We plot data as function of the theta* at fixed values of Q**2. The only free parameter is multiplication factor, that depends on the polarization of the protons:

14.)

15.)

16.)

ZLIB problem

I am still having problems with the crashing of the analysis, due to the ZLIB problem. Until recently I though that runs, that first crash but later automatically recover, are good. However, I realized that in some cases I get strange results even if only run is bad. Please see figures below.

17.)

18.)

Left figure shows results that I got when I considered run #3229, that crashed in the middle of the analysis (but eventually came successful to the end). You can see how point at p_miss=0.14GeV is moved to higher value. This happens because of this bad analysis of the run #3229. Later I reanalyzed this run, this time without any errors and got completely different results. This brings important message, that this problem can not be swiped under the rug

Therefore I went into the analyzer code and realized that by default level 1 compression is being used, which considers Lempel-Ziv-Markov compression algorithm. I set the compression level to zero and restarted the analysis on FARM. Until now I was able to analyze approx. 50 runs. None of them reported errors. Without compression analysis seems to be working better. Unfortunately this kind of analysis requires approx. six times more disk space (800MB -> 5GB)

Checked my algorithm for the rotation of the asymmetries

I have also checked algorithm, that I used to rotate asymmetries from LAB coordinate-system to the "q-CS". I believe that my code was correct. I also believe that asymmetries that I showed you last week should be the same at low p_miss, with the same sign. When looking at the elastic proton asymmetries at the same Q**2, one can see that these asymmetries (A_perp, A_par) at given Q^2 should be approx. of the same size. I will show you more detailed results next week.

Last modified: 10/05/11