Below are plots of the purity calculated from identified pi0 triggers from the run 9 dijet embedding sample.  As the sample was generated using only hard 2-to-2 QCD processes, there should be no prompt photons.  Hence the purity should be roughly flat.  I checked for both "inclusive" and "exclusive" gamma-rich triggers, which are defined the same way as in this post:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-01302018-run-9-purity

The bad run list, hot tower list, trigger requirements, and track selection are identical to those used in this post:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-02012018-run-9-purity

The purities (technically R) are below, and the corresponding delta-phi and zTtrk plots are attached.  First, I checked the purity for zTtrk = (0.2, 0.3):

And second for zTtrk = (0.3, 0.6):

Note: these plots are shown for partial statistics.  When running our clustering algorithm over the embedding files, some of the jobs for the pTparton = (15, 25) and pTparton (25, 35) GeV/c bins crashed and haven't been regenerated yet.