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

Weizhi:
Look at energy distribution without any event selection but with GEM matching.
For the outermost PWO layer there is a small bump between 1500 - 2000 MeV in data which is not seen in the simulation.
The bump gets smaller as we go from outermost blocks to more inner blocks, disappears for the inner layers.
It does not come from a particular block or particular run period.

Some discrepancy also seen for LG blocks but poorer resolution.

<TASK> Will add the contribution of gaps between blocks in simulation to hunt for the cause. 
<TASK> Try different clustering algorithms.

The slope between ep to ee ratio may be partly due to difference between form factors used in simulation, but some of it maybe due to the bump in the energy spectrum.

Looked at ep and ee separately, the angular dependency seems to be from the ep only.

Also studied trigger efficiency, it is flat. The correction for ep is simple but for ee it is more involved.



Xinzhan:
Used two methods to calculate offsets. Method 1 using GEM overlap area and then use ee2 events in GEM2 to get beam center and get offset of each detector. Method 2, in overlap area use ee2 events to get beam center, then get offset. (method 2 is independent of target z but can be applied only to overlapping area). 

The correction due to the new target position information has minimal effect (except for the early runs before 1100). But there is some difference between the methods.


<TASK> Use overlapping area (method 2) to get y-alignment and get the average target position (center of the distribution) and then use that to get x-alignment in the non-overlapping area (method 1). Try to understand difference between runs before and after run 1100.