This is a follow-up of my previous blog post, making use of the jobs I ran for it: Compare event-by-event T0 performance. In this post, I just make some observations on the impacts of using the modification to vertex-finding in FXT that was recently introduced in PR #720, biasing (shifting) the selection of tracks to use in vertex reconstruction to originate from -2 cm below the z-axis.

#1: As shown on that previous post, the width (RMS) of the difference V_z,east - V_z,west got narrower (please see the plots and discussion there).

#2: The means of both V_y,east and V_y,west shifted downwards (perhaps the most expected observation). And while the V_y,east distribution got narrower, the V_y,west got slightly wider for some reason. Here red and magenta are the unshifted VF east and west respectively, and blue and cyan are the shifted VF east and west respectively.



#3: As in observation #1, the width of both the difference V_x,east - V_x,west and the difference V_y,east - V_y,west got narrower, and their means got closer to zero. Here red is the unshifted VF, and blue is shifted.



-Gene