Create PSF solutions - global solution and prediction for all fibers

Once when we have analyzed all of the defocused and focused data, we create a solution at all possible points that we have analyzed.  For fibers that we have not analyzed (590 out of 600 fibers) I create ``fake'' data, created from interpolating all of the available parameters. For parameters that can not be interpolated (e.g., focal ratio degradation) I set those values at random values that are consistent with the values for the 10 fibers that we have analyzed.