[evlatests] Missing data in SY table

[Rick Perley]

     I am carefully reducing D-configuration data taken on June 27. The 
bands were L, S, C and X.  The latter two utilized the 3-bit system.

     The 3-bit system requires setting the requantizer gains, which thus 
requires application of these gains.  (This operation is not required 
for the L and S band data.)

     When I applied these gains, I quickly found I was missing a 
significant fraction of my data -- the visibilities 'disappeared'. The 
problem was quickly traced to gaps in the SY tables.   Some rough 
statistics show:

     1) There are no missing data in the SY tables at L and S bands.

     2) The fraction of missing SY table values at C-band is modest -- 
typically 10%.

     3) The fraction of missing SY table values at X-band is much higher 
-- typically 20%, with some antennas missing more than 50% (!).

     4) The opposite polarizations (R and L) have exactly the same gaps.

     5) The losses in A1/C1, A2/C2, B1/D1 and B2/D2 are different.

     I have no idea if this problem exists today.  It did not happen in 
March, when the A configuration data for this project were taken.