Update 08.22.2018 -- Run 9 pp: purity, zT systematic

Previously, I tried averaging the purity between 2 adjacent bins of zTtrk:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-08202018-run-9-pp-purity-testing-zttrk-cuts

However, we've decided to estimate the systematic error due our choice of zTtrk as so: the purity for zTtrk = (0.2, 0.3) will be considered our default, and the average of the purity over zTtrk = (0.2, 0.5) will be an estimate of the systematic error.

And in numbers:

eTtrg [GeV] B (0.2, 0.3) error (stat.) B (0.2, 0.5) error (stat.) %-diff.
9 - 11 0.650 0.022 0.619 0.016 4.8
11 - 15 0.554 0.030 0.526 0.023 5.1
15 - 20 0.350 0.065 0.378 0.063 8.0

Here the %-difference is given by:

diff = |B(0.2, 0.3) - B(0.2, 0.5)| / B(0.2, 0.3)