[daip] Re: AIPS questions

[Eric Greisen]

Göran Sandell writes:

 >      I haven't reduced VLA data since 1983/84 and it has been a 
 > struggle getting AIPS running here at NASA Ames; Usually if something 
 > has not worked, I have assumed it has been due to problems with the 
 > AIPS installation. I did, however, figure out that the tv was working 
 > after I talked to Joe McMullin on Friday. My confusion was probably 
 > more due to me expecting the AIPS tv to behave like most other 
 > interactive graphics packages I have used and the buttons A,B, and C 
 > were
 > confusing ( I thought it referred to the buttons on my mouse, 
 > especially since I  find a passage in the manual saying that D is 
 > usually function key F6).

       Joe does not work with AIPS much (he heads the aips++ project
these days).  "Buttons A,B,C,D" are both a,b,c,d on the keyboard and
F3-F6 on those keyboards that support them.  We use the left button on
the mouse only to tell the programs whether the mouse is moving for
aips' tv (left down or clicked) or for other things.

 > 
 >      Now I can display data, zoom, change color palette, but I still 
 > haven't manage to flag any data. I find TVFLG confusing - I hardly 
 > know what I am
 > looking at, EDITA and EDITR are more self explanatory. But, I still 
 > have not managed to flag any data. Yesterday I used EDITR to display 
 > data for me, I then
 > changed timerange so that I could identify bad data and uvflg to 
 > manually flag the data. I have not tried to flag single data points, 
 > but I tried to somehow get it to give me a start and and end range - 
 > but no luck. There must be something that I am still missing, maybe I 
 > am just stupid.

     There is a lot of documentation on TVFLG in its help and explain
file and in the aips cookbook available on the web 
      http://www.aoc.nrao.edu/aips/cook.html
It displays time on the vertical axis and baseline number on the
horizontal axis.  The latter can be sorted into approximate baseline
length order.  You get some function by moving the mouse to the menu
function, selecting it with the left mouse button, and then hitting
a, b, or c to do the function or d to get help about the function.

Some functions like select an IF do not do anything immediately except
mark the IF at the bottom with an * so that you will know that it will
change on the next LOAD operation.  This allows you to do several
selections and changes and then do the LOAD which can be slow.

Flagging is easy.  Select some flag function and pink graphics lines
will show you what you are about to flag.  Follow the instructions on
the text window and data will disappear from the display.  Thus,
select time range, move the mouse (left button down) to move the first
time, hit button a, move the mouse to set the higher time, hit button
C to flag this time range, and return to the menu.  Now select the
function that shows you the current contents of the FC table - and it
will show you what you just did.  When you EXIT it will ask whether
you want to apply the FC table to the FG table kept with your data.
Answer yes and it will add records to the FG table.  Then when you run
SPLIT (or any other task with the adverb FLAGVER) these flags will be
applied to the data.

 > 
 >      The data set I am working on is from 1978. It should have had 14 
 > antennas but I only find 12 that have data at 1.48 GHz and 2 - 3 are 
 > bad at each polarization, so it is only a few antennas to work with 
 > and it is only continuum, mostly I reduce spectral line observations.
 > 
 >      I now have a reasonable calibration, not perfect, but good enough 
 > to create some first maps. However, when I did a uvplot of the phase 
 > calibrator next to the source I am interested in, I noticed that it 
 > is generally flat over the whole UV range with a few high and low 
 > points randomly scattered around, except
 > at near 0 uv distance, where I see a fair amount of high and low 
 > points. Why is that? Could it indicate shadowing (it is a low dec 
 > source 1741-038) and if so, how do I correct for it? Should I just 
 > clip the calibrated UV data and redo the calibration stage?

     Shadowed data should have been flagged when the data were loaded
by FILLM.  It is possible that your calibrator source has some small
extended structure (I do not know anything about the source - you can
look it up on NRAO's web pages for more info) or that spillover, RFI,
or other things contribute at very short spacings.

 > 
 >      A second question? How do I write out the data as FITS on a disk 
 > (not to TAPE - I have no idea if the AIPS installation would 
 > recognize my DAT drive). It would probably be easier for me to export 
 > the calibrated uv-data to MIRIAD and do the rest of the reduction in 
 > a package that I use much more frequently than AIPS.
 > 

     All AIPS tasks have pseudo-tape disk files as an option.  The
adverb OUTFILE in FITTP controls this.  Again, the introductory
chapters in the CookBook cover this as do the help files.

Eric Greisen


 >     You don't have to reply. I am sending this to you so that you can 
 > see my questions, and then I'll try to give you a call later today.
 > 
 >                         Cheers,
 > 
 >                                     Göran
 > 
 > 
 > -- 
 > Göran Sandell
 > USRA, NASA Ames Research Center
 > MS 144-2
 > Moffett Field, CA 94035, U.S.A.
 > Phone 1-650-604-3576  Fax 1-650-604-1984
 > 
 >