An embedding test sample was produced for 3.85GeV FXT data and made available on August 31, 2020. Looking through the online QA Xainglei sent out along with some QA I ran on the minimc files, there are a few concerns.

First, the number of hits for reconstructed tracks in the embedding sample has a very significant drop as a function of eta (or rapidity) heading from -1 to 0.

The distribution for the embedded tracks was flat in this region though.

After pointing this out, Xianglei asked if we had observed this in the production data when vertex cuts had been applied. When using the standard event selection cuts ( 198<Vz<202cm and Vr<1.5cm about beam spot centered at [0,-2]), neither the global tracks nor primary tracks show this drop in this eta region.

If anything, the nHits distribution is wider in this eta region for global tracks but there is not systematic drop in either the global or primary tracks in produced data. The fact that the embedding reconstruction shows this drop is concerning.

Additionally, I observed that the number of hits for some tracks extends above 45, all the way up to 72. iTPC sector 20 was installed for this run but it was not used in the production. Looking at the last fit padrow (which should have a maximum of 45 since there are only the original inner TPC sectors) there are clearly tracks at all embedded pT values which have the last fit padrow extending to 72. At Xainglei's suggestion, I've added plots of the last fit padrow vs phi and vs eta, which both confirm that these tracks are getting hits from iTPC sector 20. Additionally, a plot vs the last sector clearly shows sector 20 has a number of tracks where the last hit is past padrow 45.

 
Perhaps I am misunderstanding what some of the variables in the minimc files are, but if I am not these two issues both cause concern about this embedding test sample.

Sept. 3 Update:
Xianglei recommended looking at the number of hits in sector 20 to see if there were only a maximum of 20 hits. In the plot of nHits vs eta for reconstructed tracks where the last hit is in sector 20, there is a clear hard cutoff at 45 hits. As Xianglei suggested, there is a change in the number of padrows with the iTPC which only had sector 20 installed. Even though the sector was not used in reconstruction (as evident by the cap of 45 hits), the new numbering scheme has been implemented, which accounts for my previous concerns.