REFERENCES

For details on how polarimetry data was accumulated, refer to [ PART ONE -- https://drupal.star.bnl.gov/STAR/blog/dilks/all-polarimetry-lut ]
For luminosity totals (obtained from Jamie's tracker), refer to [ https://drupal.star.bnl.gov/STAR/blog/dilks/run-12-and-13-integrated-luminosity-totals ]
-----------------------------------------------------------------------------------------------------------------------------------------------
ANALYSIS DETAILS

The luminosity-weighted averaging and error propagation (including systematics) follow the [ ANALYSIS NOTE ]

I worked out details by hand in order to program a calculation of overall lumi-weighted polarization + error; this is essentially a summary (and extension for DSA) of the analysis note: [ DETAILS ON POLARIMETRY AVERAGING AND UNCERTAINTY ]
(here's the [ OLD VERSION ], handwritten and for one beam only)

-----------------------------------------------------------------------------------------------------------------------------------------------
SIDE NOTES

- if luminosity wasn't provided for a fill, then the luminosity-weighted average for that fill is substituted with the beam current weighted average (this rarely happened);
in luminosity weighted averages (or quadrature sums which involve luminosity factors), these fills are by default dropped, since the luminosity is recorded in the LUT as zero

- for A_LL, using lumi-weighted average polarization for each fill (polarization in magenta, error in blue)

-----------------------------------------------------------------------------------------------------------------------------------------------
PLOTS

The following 4 pairs of plots show polarizations (left) and uncertainties (right) vs. run index ("good" FMS runs) for the blue beam, yellow beam, and for the product

Click to enlarge the images

Here is the legend for both single beam and double beam polarizations and uncertainties:

................
Run 12 Blue ----------------------------------------------------

.
Run 12 Yellow  ---------------------------------------------------

.
Run 12 Product  ---------------------------------------------------

.
Run 13 Blue ---------------------------------------------------

.
Run 13 Yellow  ---------------------------------------------------

.
Run 13 Prodcut ---------------------------------------------------

.
===========================================================================================================
===========================================================================================================

The above plots were for "good" FMS runs, which pass first-level QA on trigger rates. Only a subset of these runs make it through the full analysis to A_LL. Some runs have spikes in kinematic distributions for pions and are consequently omitted. Other omitted runs include those with bad relative luminosity measurements or calibration.

Here is a comparison of good FMS runs (left list) and runs which made it all the way through the analysis (right list):
[ RUN 12 ]
[ RUN 13 ]

===========================================================================================================

OVERALL POLARIMETRY AND LUMINOSITY FOR RUNS WHICH COMPOSE A_LL ANALYSIS

.
.
.
.