SoLID Ecal Weekly 20230629

(short meeting due to JLUO meeting earlier in the week)

Update from Jixie

Presentation on pulse deconvolution: [ post slides here]. The code adapted the idea from Hall A (DVCS) code but is not a direct copy. Similarly, won't be directly usable for other experiments' (such as NPS) decoder, though the code an be adapted to NPS. Discussions below:
The code works for 1,2,3-peak deconvolution. Can expand this to more peaks.
At the moment, the peak integral stored in the level 1 tree is the "first in time". Will change this to the peak integral that is "closest to the timing window of that detector";
Need to add multiple pulse information to level 1 tree. This requires changing the pulse from variable to array/vector. Note that tracking info will already be in vector form so analysis code should adapt to this (vector structure).
Will work on a new level1 tree with "closest in time" peak variable, and another debug version that stores all peaks as vectors.
Will write a technical note on the code/algorithm.

Preliminary electron efficiency vs. pion rejection for full-sim, classical method: [post slide here]. Comments below:
For each momentum slice, it appears there are still low energy electrons. Need to check cuts used (select events hitting center of detector setup) and variable used (need to use momentum of particle at the vertical plane in front of ECal). -- Some of these variables need to be added to full-sim and Ye will work on this.
1D cut is arbitrary at the moment, could fix 1D cut at above MIP (such as 1.5*MIP) for both preshower and shower. 2D cuts are okay though could check if the slope agree with electron distribution (when electrons are binned for a certain momentum range).
Probably best to plot both "ele eff vs. pion rej", and "ele eff vs cut" and "pion rej vs. cut".

So far Darren has worked on pencil sim but not binned in momentum. Should also use the same set of cuts as Spencer.
cuts proposed: amend notes here...