[Sane-analysis] Check of theta and phi fits in ph2.f

[Mark Jones]

I ran a simple program in which 100,000 Events were thrown uniformly between 1 and 3 GeV and with BigCal at 40 degs and z = 335cm. 
In the Bigcal coordinate system +x is large angle, +y is up and +z points toward Bigcal. 
Angles were thrown over dx/dz = 0.5 to -0.3 radians and dy/dz = +/- 0.2 radians. 
The raster position was fixed at zero. The particles were tracked through 
the target field until they travelled z=335cm. If the event was within 
the x and y dimensions of BigCal it was saved. About 80% of the events were saved.
The Theta_calc and Phi_calc were determined using the polynominals in ph2.f and the difference
with the thrown Theta and Phi were calculated.

With all this range of events for the parallel case Theta difference goes from 0 to 0.05 degrees 
while the Phi difference goes between -0.2 and 0.4 degrees. 

For the perpendicular case (-80 deg), the Theta difference goes from -0.05 to 0.35 degrees 
while the Phi difference goes between -0.2 and 1.4 degrees. 

Various plots are at 
https://hallcweb.jlab.org/experiments/sane/wiki/index.php/Mkj-angle_correction-checks


    Cheers,
           Mark