[asac] Illinois AIPS++ workshop report

[Richard Crutcher]

At our April telecon discussion of data handling I promised a report on
the Illinois BIMA users AIPS++ workshop.

The workshop was one of a continuing series of training sessions we are
conducting. Details of our work with Illinois BIMA astronomers using
AIPS++ may be found at:

http://monet.astro.uiuc.edu/aips++/forBIMA/

This workshop involved a short demonstration of using AIPS++ followed
by Illinois astronomers using AIPS++ with their own data sets. People
generally worked in groups of 2-3; there were 5 different groups, with
4 BIMA and 1 ACA data sets. The astronomers were a mix of faculty,
postdocs, and graduate students who are experienced in using MIRIAD
and/or (classic)AIPS; none have been involved in AIPS++ development.
Both spectral line and continuum data sets were represented. A detailed
handout of the workshop was given to all participates in advance. This
gives a detailed, step by step cookbook of what to do, and may be of
interest to those who want to see the details of using AIPS++. It
may be found on the above web site, or specifically at:

http://monet.astro.uiuc.edu/aips++/forBIMA/handson.txt

Briefly, the workshop involved logging in and launching AIPS++,
importing each group's own (already calibrated) visibility data into
AIPS++, creating AIPS++ tools, using the imager tool to make a dirty
map and beam, using Clark and multi-resolution CLEAN to produce
deconvolved images, and using AIPS++ to visualize images.

The 4 groups with BIMA data completed the tutorial. The group with
Australia Compact Array data had problems, but our lack of detailed
information about ACA data precluded an analysis of the nature
of the problem.

After the workshop we conducted a survey of the astronomer participates,
which follows. Averaged results are reported where numerical ratings were
requested.

-----------------------------------------------------------------

Follow-up Survey of AIPS++ tutorial:

1. Have you tried using AIPS++ again since the workshop?
	0 Yes, 5 No.

   Do you think you will try using AIPS++ again to explore the
   multiscale clean algorithm on your data?
        4 Yes, 1 No (but will have the student do it)

   Briefly, why or why not?
	Yes, because "have extended structure; explore it better with
	multiscale."

	Yes, because "multiscale clean is an interesting way to get a
	good model; reasonable way to model extended emission more
	accurately."

	Yes, "mostly to make sure that the regular CLEAN is working right,
	that I understand the extended emission well."

	Yes, because "my data has an extended source + a point-like
	component."

****************************************************************************
Please indicate on the following scale how easy or hard you found various
aspects of aips++ to be.
   Very Easy -------------- 5
   Somewhat Easy ---------- 4
   Neutral ---------------- 3
   Somewhat Hard ---------- 2
   Very Hard -------------- 1

2. Setting up to run aips++: 4.0

3. Starting up aips++: 4.4

4. Locating tools/functions in the GUI: 2.2

5. The concept of the tool-based approach: 2.8

6. Entering inputs into the GUI: 4.6

7. Browsing datasets on disk: 4.7

8. Using the Image Viewer: 3.4

9. Creating a tool: 2.5

10. Using the imager tool to create images: 3.8

11. Obtaining statistics for an image: 4.3

****************************************************************************
Indicate on the following scale how useful you think various aspects of
aips++  are for the kinds of processing you do.
    Very Useful ----------------- 3
    Somewhat useful ------------- 2
    Not useful ------------------ 1

12. The concept of the tool-based approach: 2.3

13. The GUI interface: 2.6

14. The Glish interface: 2.5

15. The catalog browser: 3.0

16. The Image Viewer: 2.8

17. Obtaining statistics with the Viewer: 2.6

****************************************************************************
18. During the workshop, did you try any functionality that was not
    covered by the tutorial hand-out or demo? If so, what was it?

	No
	Yes, convert to FITS file
	Yes, plot of profile in frequency
	No
	No

19. How difficult was it to use this functionality?

	"Conversion to FITS file was easy to use, but hard to find;
	wouldn't have been able to locate if I had not been told
	where and how to look for it"

	"Neutral; for plotting the frequency profile, I was shown what
	to do, so I don't know how easy it will be if I have to find
	out by myself how it is done"


****************************************************************************
20. What impressed you the most about AIPS++?
	"gui interface concept"

	"easy visualization; using scripter to follow up on what you
	did with the gui, so that in subsequent reductions, you don't
	have to keep using the gui"

	"how easy it was to get set up"

	"can get a beginner going into and learning it quicker than
	if you sent them to a task-based software like miriad"

21. What impressed you the least about AIPS++?

	"obscure; not intuitive"

	"still difficult to find desired tools"

	"logic and organization of the various tools, the way they were
	grouped"

	"time it took to do some of the tasks"

	"still don't know what to do when you run into a problem - is
	it a bug? or did I do something wrong"

****************************************************************************
22. Imagine that you became as proficient using AIPS++ as you are now with
    Miriad.  If you had a choice between using Miriad and using AIPS++,
    which do you think you would choose?

	"use both - you never use one"

	"depends on whether aips++ is a complete package"

	"aips++, but only if it acquires complete functionality. I don't
	want to live in both worlds."

	"aips++, because everything is going to that + aips++ will
	probably be supported better than miriad"

	"don't know the full capability, so can't tell"

****************************************************************************
How important are the following features in making your choice?  For each
feature, please choose one of the following:
   Very important ------------- 5
   Somewhat important --------- 4
   Moderately important ------- 3
   Not very important --------- 2
   Not important at all ------- 1

23. The GUI interface: 4.4

24. The command-line interface: 4.6

25. Ease of scripting: 4.4

26. Sophistication of scripting (e.g. access to arbitrary data in
    datasets, statistics & fitting functions, importing custom data,
    plotting): 4.2

27. Access to algorithms or functionality not available elsewhere: 4.8

28. Quality of documentation: 4.8

29. Visualization tools: 4.4

30. Custom Plotting: 4.4

31. How likely do you think it is that you will try AIPS++ again in
    the next year?  Choose from the following:
       Very likely -------- 5
       Somewhat likely ---- 4
       50-50 -------------- 3
       Not very likely ---- 2
       Highly unlikely ---- 1

    Average score: 4.6

-------------------
Richard M. Crutcher
Professor & Chair, Astronomy Department
Chief Application Scientist, NCSA
University of Illinois
1002 W. Green Street
Urbana, IL  61801
Voice: 217/333-9581
Fax: 217/244-7638