Quick and dirty guide for 3mm reduction.  In this example, the source is rscl,
the gain calibrator is 0104-408, the bandpass calibrator is 2255-282, and the
flux calibrator is uranus.  Contact Tony.Wong@csiro.au to get the data
for this exercise.

                            *****************
			    * PRELIMINARIES *
			    *****************

1. atlod - Read RPFITS file into Miriad

Task: atlod
	in=2002-06-21_1746.C1048
	out=atfile.1
	options=mmrelax,noauto

Sometimes this doesn't work, and you have to repeat the task with 
options=mmrelax,noauto,noif.

2. attsys - Applies paddle-derived Tsys values to data
(Only necessary if you disabled Tsys during the observation.  The paddle values
have been written to the data file but not applied as yet.)

Task: attsys
	vis=atfile.1
	out=tcal.1
	options=auto  [don't apply it if it's already been applied]

3. elevcor - Applies a gain/elevation correction for 3mm
(This is because the antenna gain changes significantly with elevation on
some antennas in a rather predictable way.)

Task: elevcor
	vis=tcal.1
	out=elevcal.1

4. uvedit - Applies a baseline (relative antennas position) correction 
(This is needed because the baselines at 3mm do not seem to be sufficiently
well-determined.  Hopefully can be avoided in the future.)

Task: uvedit
	vis=elevcal.1
	out=uvedit.1
	dantpos=2,0.0038,0.0001,-0.0056,4,-0.0025,0.0067,-0.0079 [e.g.]

NOTE: Skip this step during the first reduction pass.

Repeat steps 1-4 for each RPFITS file.

5. List sources and select out those you want to use:
(Repeat for each source and calibrator.  Ignore pointing & paddle scans.)

Task: uvindex
	vis=elevcal.1  [can repeat for other files]

Task: uvcat
	vis="elevcal.*"  [or uvedit.*]
	select='source(rscl)'   [repeat for all calibrators]
	options=unflagged
	out=rscl

6. Plot UV coverage on source

Task: uvplt
	vis=rscl
	axis=uc,vc
	device=/xw
	options=nobase,equal
	select='pol(xx)'

7. uvsplit - Split data by frequencies

Task: uvsplit
	vis=rscl  [repeat for all calibrators]

8. puthd - Specify rest frequency for spectral line (in GHz)

Task: puthd
	in=rscl.88638/restfreq  
	value=88.6318473  


                            ****************
			    * BANDPASS CAL *
			    ****************

All of these steps must be repeated for the 2nd frequency if used.

9. Examine your bandpass calibrator vs. time and vs. frequency:

Task: uvplt
	vis=2255-282.88638
	axis=time,phase  [or time,amp]
	device=/xw
	size=2,6
	yrange=-180,180  [phase only]

Task: uvspec
	vis=2255-282.88638
	axis=chan,phase  [or chan,amp]
	interval=1000	 [average all data in time]
	nxy=1,3
	yrange=-180,180  [phase only]

Note the number of edge channels on each end of the spectrum for flagging.

10. Flag edge channels:

Task: uvflag
	vis=2255-282.88638
	edge=20,16
	flagval=f

11. Calibrate the bandpass:

Task: mfcal
	vis=2255-282.88638
	interval=1000	 [average all data in time]
	refant=3	 [choose a centrally located antenna]

12. Examine the bandpass solution:

Task: gpplt
	vis=2255-282.88638
	yaxis=phase   [or amp]
	device=/xw
	nxy=1,3
	options=bandpass
	yrange=-180,180  [or 0,2 for amp]

13. If OK, copy the gains to the other sources:

Task: gpcopy
	vis=2255-282.88638
	out=rscl.88638  [repeat for gain and flux cal]
	options=nocal   [do not copy gains vs. time]


                            *****************
			    * GAINS VS TIME *
			    *****************

All of these steps must be repeated for the 2nd frequency if used.

14. Examine phases and amplitudes on gaincal:

Task: uvplt
	vis=0104-408.88638
	axis=time,phase  [or time,amp]
	select='pol(xx,yy)'
	yrange=-180,180  [for phase]
	size=2,6
	device=/xw

15. Do the calibration and plot the gains:

Task: mfcal
	vis=0104-408.88638
	interval=5,2  [average every 5 minutes of cal data]
	refant=3      [choose a centrally located antenna]
	options=nopassol  [don't do passband]

Task: gpplt
	vis=0104-408.88638
	yaxis=phase   [or amp]
	nxy=1,3
	device=/xw
	options=wrap  [for phase, don't unwrap]
	yrange=-180,180  [for phase]

16. Examine gaincal again with gains applied:

Task: uvplt
	vis=0104-408.88638
	axis=time,phase  [or time,amp]
	select='pol(xx,yy)'
	yrange=-180,180  [for phase]
	size=2,6
	device=/xw


                            *****************
			    * ABSOLUTE FLUX *
			    *****************

All of these steps must be repeated for the 2nd frequency if used.

17. Examine amplitudes on flux cal:

Task: uvplt
	vis=uranus.88638
	axis=time,amp
	select='pol(xx,yy)'
	device=/xw

18. Rescale the amplitude gains using planet:

Task: plboot
	vis=uranus.88638,2255-282.88638,0104-408.88638 [planet,passcal,gaincal]

19. Check the fluxes for our calibrators:

Task: uvflux
	vis=2255-282.88638,0104-408.88638


                            ******************
			    * SOURCE SPECTRA *
			    ******************

20. Apply the gains from the gain calibrator:

Task: gpcopy
	vis=0104-408.88638
	out=rscl.88638
	options=nopass  [don't overwrite bandpass gains]

Task: uvcat
	vis=rscl.88638
	out=rscl.88638.cal

21. Plot the visibility spectrum for each baseline averaged over time:

Task: uvspec
	vis=rscl.88638.cal
	axis=vel,amp     [or vel,phase]
	interval=1000
	nxy=1,3
	device=/xw
	yrange=-180,180  [for phase]
	select='pol(xx,yy)'


                            *****************
			    * MAKE THE MAPS *
			    *****************

22. Make the dirty map and a postscript plot:

Task: invert
	vis=rscl.88638.cal
	map=rscl.88638.map
	beam=rscl.88638.beam
	cell=1       [1" pixels]
	imsize=128
	sup=0        [natural weighting]
	slop=0.5     [don't throw out visibilities with flagged channels]
	line=vel,55,-45,1,1
	select='pol(xx,yy)'

Task: cgdisp
	in=rscl.88638.map
	type=c
	device=dmaps.ps/ps
	labtyp=arcsec
	nxy=4,3

23. Do a light CLEAN and plot:

Task: clean
	map=rscl.88638.map
	beam=rscl.88638.beam
	out=rscl.88638.clean
	niters=100
	region='arc,box(-10,-10,10,10)'

Task: restor
	map=rscl.88638.map
	beam=rscl.88638.beam
	model=rscl.88638.clean
	out=rscl.88638.cm

Task: cgdisp
	in=rscl.88638.cm
	type=c
	device=cmaps.ps/ps
	labtyp=arcsec
	nxy=4,3