1) Add tree variables for the three BPMs and projections to target
in the EPICs coordinate system (+X beam right, +Y up)
rb.raster.fr_xbpm_tar
rb.raster.fr_ybpm_tar
rb.raster.fr_xbpmA
rb.raster.fr_ybpmA
rb.raster.fr_xbpmB
rb.raster.fr_ybpmB
rb.raster.fr_xbpmC
rb.raster.fr_ybpmC

2) Modified code so that it only modifies the BPM variables if they
are all !=0 ( beam on) . Otherwise it sets them to the previous EPICS
read. This gets rid of all zero EPICS reads,except if the first EPICs
read the beam is off.

3) Removed the reading of gpbeam from the global parameter list in the Process
  method and just have it in ReadDatabase.

4) Change the calculation of beam at target to use the projection of
   BPM_A and BPM_C instead of BPM_A and BPM_B.

5) Remove setting the raster position to the hcana beam coordinate system
   with (+X beam left). The raster is in the EPICS coordinate system.
    This makes it easier to interpret the carbon hole runs.