This dataset is one day (out of four) of a deep HI survey of
NGC2403.  The data have been published by Fraternali et al. (2002,
AJ, 123, 3124) and a pdf file of this paper is available in this
directory, Fraternali02.pdf.

The following is a brief summary of my data reduction:

Note, Gustaaf's fillm used douvcomp 1, so weights for individual IFs
are lost.

Initial flagging and calibration was carried out on the channel 0 data.

Editing:

Some editing of obvious bad times and the odd pixel is needed
(tvflg).  I ran tvflg in two timeranges in order to display the
whole dataset.  Inspection of the data in VPLOT also shows some bad
integrations associated with antenna 22 (Lpol) and these I edited
using uvflg.  Both pols were flagged simultaneously in both the
tvflg and uvflg editing.

Setjy/calibration:

There are three primary flux calibrators, none of which has any uv
restrictions for this particular configuration.  Calib is straightforward,
clcal was run with default settings.

Imaging:

Even the channel 0 image is very pretty!  I used a cell size of 2",
2048x2048, robust 0 weighting.  Without using any clean boxes there
is a negative bowl.

Copy CLv2, FGv1 to the line data:

Bandpass calibration:

This will be a useful exercise for the students: two of the primary
calibrators contain Galactic absorption, so they should work out which
one they should use.  Another teaching note: the Galactic absorption is
quite strong for one of the calibrators, and exhibits the Gibbs phenomenon
when bpass is run without Hanning smoothing.

If you need it, the observe file used for these observations is in this
directory, AS649.OBS.

Split out data; use docal 2, douvcomp 1 as for fillm.

Imaging:

This can take a very long time if you want to image the entire cube.

Decent sensitivity in channels 2-121, line emission in channels
~35-92.  Use channels 2-30 and 97-121 for the continuum, uvlin with
order 1.

Image channels 31-96 averaging 2 channels together.  Cell 4",
512x512 to make it faster, niter 100000, flux 0.0011 (3-sigma in
line-free dirty image).

Note Galactic HI absorption at velocities ~0km/s.

CJC