The run 15 pp, pAu and pAl data are organized as in the previous presentations as html tables with rows for >100 AN Data points.
  Tables are updated for 

2)  Cone .08 Rad, >= 2 Photons in cone.

3) Cone .08 Rad, ==2 Photons in cone

Web pages contain:
Table with >100 rows followed by  An vs Pt plots. 
Appended are plots showing the relative luminosity from fits as a function of pT for each energy set.
Finally we see the chi^2 distributions for fits to the phi distribution (click on table col 7 entries to see data and fits).  

Pt plotted with uncertainty in Pt from col 1 and An from  col 7 of table. Horizontal errors from col 1 and vertical error bars representing statistical errors from col 7.

Systematic vertical errors (again from col 7) are shown as yellow bar.  Included as systematic errors are Polarization error (col 6); Mass Fit Bck Error (col 3 sErrM) and non-beam error (col 3 sErrB) in quadrature. Mass Fit Bck Error dominates in most cases.  

Run by Run Polarization (intercept +slope time) with corresponding errors are shown when clicking col 6 (upper plot).  Also shown is the quadrature combination of those nominal errors up to a particular run index (lower plot) as a function of run index (black). The errors will be dominated by scale errors shown in yellow.

Weighting of per/run polarization is done by events counted with a pt and Energy bin for each row in the table, not luminosity per run.  Slight variation in polarization with pT reflects the reduced trigger prescale factors for the last runs in the fill where polarization may have decayed.