STAR has measured for  last  12 years beam line parameterized as

• x = x₀ + dxdz*z,
• y = y₀  + dydz*z

  I am curious what we can learn from 

• correlations between  x₀  and dxdz, 
• correlations between  y₀  and dydz, and
• time dependence of all these parameters.

I took all data base entries in vertexSeed table, exclude bad ones and plot them.

• From  correlations between  x₀  and dxdz I see 3 ragions: two of them with  |x₀ | > 0.2 cm corresponds to large dxdz (| dxdz| ~ 1e-3) and these regions can be connected  by eye with straight line, and region with low  |x₀ | < 0.2 cm which has dxdz ~ -1e3. This third region looks strange.

• correlations between  y₀  and dydz looks ~ o.k. From this plots it follows that < dydz > ~ 0.2e-3. Does this mean that STAR East sags with respect to West by 0.2 mrad ?

• Time dependence of x₀ shows that the strange region (small  x₀ and large dxdz) is associated with y2003 and y2006 runs.

• dxdz time dependence (except y2001 and y2006) can be approximated dy staight line with slope (0.32 +/- 0.02) e-3/year.  Does this imply that STAR beam is rotated around Y-axis by 0.3 mrad/year or that since  y2008 till now we have stable regime with dxdz ~ 1.3e-3 ?

• Time dependence of y₀ shows that after y2008 it reaches a saturation.

• dydz time dependence also can be approximated dy staight line with slope (-0.045 +/- 0.006) e-3/year.  Does this imply that STAR or beam is rotated around X-axis by 0.045 mrad/year ? From other side dydz dependence can be approximated by constant (0.19e-3) for region y2007-y2012, except the last point.

I would conclude that there some patterns in beam position. It would be interesting to understand the reasons for these patterns. Because X direction is RHIC bending direction it could come from CAD beam tuning. The shift (~0.2 mrad) of beam in Y direction (non bending) could be related to STAR position with respect to RHIC.