Both vector and scalar averages are formed. It also prints out the rms scatter around the vector mean, and the RMS variation in the amplitude.
Definitions in Detail: UvFlux prints out six quantities.
Theoretical RMS -- This is the rms error, resulting from receiver thermal noise, that would be expected in the real or imaginary part of each visibility.
Vector Average -- This is simply the normal average (mean) of the real and imaginary parts of the visibility data. For a point source, the real part should give the point source flux density, and the imaginary part should be noise.
RMS Scatter -- This is the RMS scatter of real and imaginary parts of the actual visibilities around their mean value. If the data do represent a point source, and it is well calibrated, this should be the same number as the ``Theoretical RMS''.
Average Amplitude -- This gives the mean value of the visibilty amplitude (the so-call amplitude scalar average). If the S/N ratio is much less than 1, this average amplitude will be dominated by thermal noise bias, and should be 1.25 times larger than the ``Theoretical RMS''. If the S/N ratio is much greater than 1, then this should be a good estimate of the point-source's flux density. This will be unaffected by phase calibration errors (unlike the ``Vector Average'', which decorrelates).
RMS Amp Scatter -- This gives the RMS scsatter of the visibility amplitudes around the mean visibility amplitude. If the S/N ratio is much less than 1, then this should be 0.65 times the ``Theoretical RMS'. If the S/N ratio is much greater than 1, this should be the same as ``Theoretical RMS''.
Number Corrs -- This gives the number of correlations, N, used in forming each mean. NOTE that the three RMS values printed are RMS difference between the correlations and some mean -- it is not an error-of-the-mean. To convert an RMS to an error-in-the-mean, divide by sqrt(N).
nocal Do not apply any antenna gain calibration corrections. By default these are applied if they exist. nopol Do not apply polarisation leakage corrections. By default these are applied if they exist. nopass Do not apply bandpass corrections. By default these are applied if they exist. uvpol Print out fractional linear polarisation and polarisation position angle (provided Stokes I,Q,U are requested).