Unlike most of the other detectors, the HMS Cerenkov (cere) and SHMS Heavy Gas (hg) detector efficiencies are strongly dependant on where the particle trajectory. For cere, there was a broken mirror up to run 4400, which causes low efficiency at dp/p<-7%. For the hg, after run 4400, the mirrors were moved making for a gap in the middle of the detector near dp/p=0. The inefficiencies are also a strong function of angle, as well as dp/p. For this reason, I used clean samples of events to generate efficiency maps in 30 bins in dp/p and 20 bins in each of xptar and yptar. For cere, I required a good e-pi coincidence, as shower counter etrk>0.9. For the hg, I required a good coincidence time, and rf timing cut, and only looked at negative particles so almost all particles should fire the detector. For the hg, I divided the efficiency values into three regions of pion momentum: 3 to 3.25 GeV, 325 to 3.7 GeV, and >3.7 GeV. For P<3.0 GeV, the hg is not used for pion PID, because it too close to the pion threshold. I used a cut of 0.5 pe for the hg and 0.3 pe for the cere. I used all the runs in a given run period to get enough counts.
The first figure shows the cere efficiency versus dp/p for spring 18. The left plots has 20 curves summed over xptar for different bins in yptar, and the left plot is visa versa. It can bee seen that the inefficincy is not just a function of dp/p. Figure 2 shows the hg efficiency for fall 2018 for 3.25<ppi<3.7 GV. Notice the big dip near dp/o=0, which again is angle dependant.
I applied these efficiency arrays to the weights of the SIMC gererated events on an event by event basis.
Figure 3 shows a comparison to SIMC without efficiency corrections, and figure 4 is with corrections. Notice the improvement in HMS dp/p at negative dp/p for the HMS for run the first run (3669, spring 2018), and for the SHMS dpd/p for the second run in the plots (5409, fall 2018).