Attendees: Ashot, Chao G, Chao P, Dipangkar, Eugene, Mahbub, Maxime, Weizhi, Xinzhan, Xuefei.

Xuefei:
Repeated analysis of Ingo Sick about uncertainty in radius when Q^4 term is included in the fit to extract radius, He can reproduce Ingo's results but they are for the completely unconstrained fit. If you include constraints such as therm should be positive it reduces uncertainty or if the dipole charge distribution is assumed the uncertainty is comparable to the linear fit.

Also set-up a suite of fitting tools that can be used with the PRad data.
They have been cross checked against each other, using set of simulated data.



Weizhi: 
Cosmic rejection power compared for all the different methods using 2 2 GeV coamic runs (76k events, for 1800 < E < 2500 MeV, > 5 deg.)
The best method misses 3 out of 151 or a ratio of 2%.

Adding a 1 sigma energy cut can do better but will also reject good ep events.

In the 2 to 5 degree range, the best method misses 9 out of 1580 

<TASK> Need to show both rejection power and efficiency and its uncertainty.

Also discussed GEM efficiency for ep and ee convoluted with matching condition.

Ratio of ep/ee from simulation to data. Compared for bin-by-bin vs integrated Moller method, now for all angles. There is large deviation from 1 at large angles.
<TASK> check radiative corrections from different material. Will try without wondow to see effect. Try also varying the inelastic contribution and subtraction.

Xinzhan:
GEM efficiency with hycal cluster size and energy cuts
Not much dependent on energy cuts (below 2sigma)
<TASK> training NN with cosmic simulation
<TASK> comparison with Weizhi's efficiency

General:
Needs to think about bullets and plots to give to Volker for CLAS collaboration meeting