[Sane-analysis] Minutes of December 8 Analysis Meeting

[Narbe Kalantarians]

_*James:*_ Updated asymmetries using table of corrected half-wave plate 
status for each run and also correcting for live-times. Will take out 
runs with live-times of less than 30%. To account for background, might 
consider looking at Cerenkov out of time.
Put together a list of target polarization change (i.e. how much in 
percentage the target polarization increased or decreased) for each run. 
Might consider taking out runs with target pol. change of greater than 5%.
Oscar noticed that for the 4.7GeV perp data the early part was mostly 
from Bottom cup and the latter part was mostly from Top cup. This might 
be relevant to the puzzling change of packing fraction seen in the 
period of changing from perp to parallel using the same load.
https://hallcweb.jlab.org/experiments/sane/wiki/index.php/Raw_Asymmetries#Updates_to_Asyms_from_HWP.2C_LT:_12.2F8.2F2010
Tracking to target. Taking difference of target -y position, projected 
from BigCal to (Front)Tracker, with and without Bdl as a function of E'. 
The spread is larger at lower E', as expected. Can almost see a 2 
different deflections as E' is decreasing, possibly indicating e+/e- 
pairs. 
https://hallcweb.jlab.org/experiments/sane/wiki/index.php/Particle_Tracking_to_Front_Tracker%2C_Target#Difference_in_Y_position_at_Target_using_Bdl:_12.2F8.2F2010
Mark suggested superimposing a calcualted curve for the electrons for 
better visual. The smearing at E' < 0.5GeV is also from the neural 
network not working too well at that range. Will cut out E'<0.5GeV. Jon 
will compare with MC.


_*Luwani:*_ Looking at energy distribution for neutral pions. 
Distribution seems to be towards higher E' (>1GeV). This fro madding the 
energy for the 2 final-state photons.
https://hallcweb.jlab.org/experiments/sane/wiki/index.php/Background_from_Pi0_decay
Will make a similar plot for electrons ans sort pi0 distributions by 
helicity to get asymmetry.


_*Hovhannes:*_ Plotting azimuthal and polar (physics) angles vs. each 
other to look at the efficiencies. Using cuts of
nclust=1 and E_clust>1GeV for 5.9GeV perp data. Reversing the direction 
of the field seems to make the distribution seem more reasonable. Mark 
said that the coincidence data didn't show signs of the analysis-engine 
having it reversed. This brought the question of whether the field 
direction is correct in the MC used for training the neural network. 
James had seen some possible hint of this back in 11/17/10. Will check this.


_*Oscar:*_ Made similar plots with 5.9GeV perp data to cross-check 
Hovhannes. Sees similar distributions. Notices that the Cerenkov makes a 
significant difference in the distributions, redistributing some of the 
events. Will try binning in E'. Jon is looking through the Cerenkov part 
of the analysis-engine to see what is being done. The BigCal x&y 
distributions should also be looked at.
http://www.jlab.org/~rondon/sane/analysis/replay/phi-theta/

As always, please let me know if I missed anything or if I made any 
mistakes.

Next meeting will be December 15th.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/sane-analysis/attachments/20101209/e15085ee/attachment.html