Communication & Documentation

1. Have you accessed the new DIII-D Data Analysis web page (http://web.gat.com/diii-d/data/)?

• Yes: 13
• No, I didn’t need to: 6
• No, I didn’t know where to find it: 0
• No, I didn’t know it existed: 6
• No for some other reason : 1

2. If so, how easy was it to find what you were looking for?

• Found it right away: 0
• Found it very quickly (less than a minute): 4
• Found it quickly (a minute or two): 3
• Found it eventually (more than a few minutes of searching): 2
• I could not find what I was looking for. Specify.: 1
• N/A: 2

3. Have you accessed the ReviewPlus manual?

• Yes: 7
• No, I didn’t need to: 6
• No, I didn’t know where to find the manual: 3
• No, I didn’t know the manual existed: 9
• No for some other reason : 1

4. If so, how easy was it to find what you were looking for?

• Found it very quickly (less than a minute): 1
• Found it quickly (a minute or two): 2
• Found it eventually (more than a few minutes of searching): 2
• I could not find what I was looking for: 0
• I could not find what I was looking for. Specify.: 2
• N/A: 0

5. How do you prefer to look for documentation?

• Webpages: 18
• Printed documentation: 13
• UNIX man pages: 5
• Other (specify): 3
  1. depends - Web is good for quickie answers, but paper is better for in-depth info.
  2. ask where to find it
  3. Usually webpages are fine, but there are times when I find printed documentation useful

6. Would you benefit from better documentation for any of the following? Check all that apply.
For each item checked, please indicate how the documentation could be made better.

• Access Grid (AG) : 2
  1. Many people (including me) still don't quite know what we can use this for and how to get started. I think Gheni will fix this.
• D3DRDB summary quantities: 4
  1. Could use better documentation for accessing data (building queries) and building dbs. This is bigger than the summaries.
  2. I need to know more about this and I'm not sure how to proceed
  3. should have a search widget
• MDSplus signals: 7
  1. simple list to search through
  2. improve search for description strings
• PTDATA pointnames : 7
  1. Google type search instead of SQL, also listed by diagnostic
  2. A Web page of ALL active pointnames and pseudopointnames with link to info about them such as digitizer and clocks. The clock link should be to the relevant clock page that has all pointnamess digitized listed.
  3. better descriptions
  4. Clearer documentation without so many errors; for example, multiple errors in CER pointname list
• Data Analysis Monitor (DAM): 0
• DBPlot: 2
  1. Not sure what it is. Like plotn? Need getting started info
• EFITLoader: 2
• EFITTools: 1
• EFITViewer: 3
• entry_display : 3
  1. I have always found entry_display very nonintuitive. Labels on the GUI seem to have little relation to what I think I want to do.
• GADAT/GADAT2 : 1
• General data analysis documentation : 4
• General MDSplus documentation : 3
  1. I am using a some old mdsplus, tdi presentation as a reference. Something more compact or concise would be useful
• General PTDATA documentation : 4
  1. List of error codes and their meaning
• General relational database documentation : 4
  1. Could use better documentation for accessing data (building queries) and building dbs.
  2. more examples of SIMPLE searches; syntax is awkward, so examples help
• General UNIX documentation : 2
  1. KEEP A GOOD HUMAN ON CALL AT ALL TIMES
• PreTRANSP: 3
  1. Is this TRANSP? I couldn't find a list of signal names
• ReviewPlus: 8
  1. how about a plethora of revplus example files to plot standards, e.g. cerneur(R,time)
  2. On-line help: does it exist? Is it reasonably functional?
  3. improve search for description strings
• IDL : 4
  1. Sometimes found it hard to locate the compiled routines
  2. maybe a Web page of ALL non-IDL released routines with summary of action
  3. I find the acrobat based on line manual difficult to "browse"
• GA Plot Objects : 4
• Other (specify): 2
  1. I would probably use documentation on all of these eventually, but I can't specify something specific that I see missing right now
  2. home x server software

7. Would you benefit from the addition of any of the following communications tools with regards to communication with the Data Analysis Applications Group?
Check all that apply.

• Shared desktop displays: 1
• Instant messaging: 2
• Newsgroups: 3
• Web-based forum: 2
• Desktop videoconferencing: 1
• Other (specify): 2
  1. attempting to use vrvs/ag now
  2. All of these have some value, but use existing technologies. Don't reinvent the wheel!

USER NEEDS & DEMOGRAPHICS

8. In a typical week, do you work from home?

• No: 8
• Yes, less than 8 hours a week: 12
• Yes, 8 hours a week or more: 6

9. When working from home, which operating systems do you use regularly on your home computer? Check all that apply.

• Linux: 7
• Mac OS 9 or older: 3
• Mac OS X : 8
• Win32 (Windows NT/2000/XP): 7
• Other (Specify): 1
  1. Windows 98SE

10. When working from home, which web browsers do you use regularly? Check all that apply.

• Firebird : 1
• Mozilla: 6
• Safari : 7
• Internet Explorer: 10
• Netscape: 7
• Other (specify): 0

11. When working from home, do you run any data analysis applications directly on your home computer? Check all that apply.

• DBPlot: 0
• EFITLoader: 0
• EFITTools: 4
• EFITViewer: 6
• entry_display: 3
• ReviewPlus: 6
• GAProfiles: 1
• PreTRANSP: 0
• SAV (Signal Analysis and Visualization): 0
• Your own code: 7

12. When working from home, what kind of Internet connection do you use?

• None: 0
• Dial-up modem : 1
• Broadband (Cable modem or DSL): 17
• Other (specify): 0

13. In a typical month, do you work offsite (at another institution, not at home)?

• No, I do not work offsite in a typical month: 21
• Yes, I work offsite a few days a month (1-5 days a month): 2
• Yes, I work offsite several days a month (6-19 days a month): 0
• Yes, I work offsite most of the time (20+ days a month): 2

Optional - Institutions:

• ASDEX-U, JET, JT-60U
• The real answer is: I work offsite, but not in a typical month.

14. When working offsite, which computing platforms do you use regularly on offsite computer(s)? Check all that apply.

• HP-UX : 1
• Linux : 2
• Mac OS 9 or older: 1
• MAC OS X : 2
• OSF/1 (a.k.a. TRU64) : 1
• Solaris: 1
• VMS: 0
• WIN32 (Windows NT/2000/XP): 1
• Other (specify): 0

15. When working offsite, which web browsers do you use regularly? Check all that apply.

• Firebird : 0
• Internet Explorer: 2
• Mozilla: 0
• Netscape: 4
• Safari : 2
• Other (specify): 0

16. When working offsite, do you run any data analysis applications directly on offsite computer(s)? Check all that apply.

• PreTRANSP: 0
• DBPlot: 0
• EFITLoader: 0
• EFITTools: 0
• EFITViewer: 1
• entry_display: 0
• ReviewPlus: 0
• GAProfiles: 1
• SAV (Signal Analysis and Visualization): 0
• Your own code: 3

17. Please describe your job. Check all that apply.

• Experiment: 22
• Simulation: 7
• Theory: 2
• Engineering: 6
• Tokamak Operations: 7
• Software Programming: 10
• Computer Systems administration: 3
• Management: 5
• Other (specify): 1 See All Answer(s)

REMOTE PARTICIPATION AND COLLABORATION

18. Would you benefit from remote participation software that enabled you to attend meetings remotely?
An example would be software that allowed you to join the pre-ops meeting from your office or from home.

• Yes: 18
• No: 7

19. What type of remote collaboration/sharing software do you use? Check all that apply.

• Access Grid: 2
• VNC: 3
• VRVS: 4
• None: 12
• Other (specify): 2
  1. D3D-TV
  2. Real Player (one way, but it works)

20. A tiled display wall system was tested in the control room physics operator area during last (2003) experimental campaign.
New units will be installed soon. In your opinion, what kind of information should be displayed on the tiled display screen? (max 1000 characters)

• Number of response: 14
  1. I'm not sure how much it can handle. Let me look at it. The 'overhead' JET display on the web would be a good start, but I think it could have a bit more than this.
  2. Physics things not engineering things Ip, ne, Wdot, bdot, Te, Ti, Impurity signal, Beams, photodiode Also should be some 'traces of the day'
  3. (no opinion)
  4. Ip, pinj, drsep, filterscope, EFIT contours
  5. Simple one line summary of the last n shots similar to what is done at JET
  6. text with plan for the next shot discharge shape some data waveforms of general interest or which show basic discharge results (some real time display)
  7. Experiment status, plots of general interest (current, heating power, etc.). Some space should be set aside for the session leader to use.
  8. One tile with shot number and general shot info like in the summary page. Remainder of tiles should replicate scope data and allow the session leader to put up other important signals.
  9. The equilibrium shape. A GUI is needed for the Session Leader to decide what info is most important to display; this will change daily. A lot of info that will routinely be displayed is already on the 4 LCD monitors. We will need to gain some experience with the new tiled wall to decide what it will display routinely, so a general purpose GUI to start with would probably be best.
  10. Data from last plasma shot, including: Shot #,Purpose of experiment,session leader,Ip, Pinj, Prad_tot, Densv2 Other time traces selected by sesson leader: e.g. ECHPWR, ICHPWR, gasa_cal, phd05 Movie of EFIT01 separatrix, boundary and divertor legs, vs time. Video of all active remote "control rooms"
  11. information that rapidly lets individuals understand what happened on the previous shot, so that on-the-fly experimental decisions can be made. Session leaders' decisions about what the next shot is supposed to accomplish. Shot anomaly information, e.g. "BEAMS FAILED" or "GasPUFF didn't work"as soon as information is ready to display. Again, the goal is to maximize the number of useful shots in a day by letting people respond to diagnostic changes, pit runs, etc as quickly as possible so you don't lose the next shot.
  12. Forget it; it was a waste of space
  13. Real time signals similar to what's on the display above the operator. I'm not sure how useful having the session leader's or someone else's analysis displayed would be.
  14. All C02 chords

DATA ANALYSIS AND VISUALIZATION

21. Which local data analysis and visualization applications do you use regularly? Check all that apply.

• DBPlot: 0
• EFITTools: 10
• EFITViewer: 21
• GAProfiles: 17
• ProfileViewer: 3
• REVIEW: 11
• ReviewPlus: 18
• SAV (Signal Analysis and Visualization): 0
• PreTRANSP: 3
• Other (specify): 3
  1. own idl routines
  2. CERVIEW CERFIT
  3. Tom Osborne's python profile analysis routines

22. Which commercial data analysis or visualization software packages do you use regularly? Check all that apply.

• Excel: 14
• IDL : 19
• JMP : 0
• MATLAB : 4
• Maple: 0
• Mathematica: 1
• Other (specify): 3
  1. PYTHON !
  2. Kaleidagraph LabView
  3. PV-Wave

23. What types of visualizations do you find useful? Check all that apply.

• Simple f(x) line plots: 22
• XY scatter plots: 22
• Profile plots: 20
• Images (2D B&W or grayscale): 12
• Images (2D with color): 17
• 3D plots (B&W or grayscale) : 7
• 3D plots (with color) : 11
• Histogram: 7
• Bar chart: 3
• 2D animation: 5
• 3D animation: 4
• Other (specify): 1
  1. Not a good general question: depends on the data I'm trying to visualize.

PROGRAMMING

24. For which platforms (if any) do you write or maintain software? Check all that apply.

• HP-UX : 18
• Linux : 13
• Mac OS 9 or older: 0
• MAC OS X : 6
• OSF/1 (a.k.a. TRU64) : 5
• Solaris: 2
• VMS: 4
• WIN32 (Windows NT/2000/XP): 3
• Other (specify): 1
  1. IDL, LabVIEW

25. Which programming languages do you use?

• C: 6
• C++: 5
• C#: 0
• Fortran: 15
• IDL: 19
• Java: 1
• Matlab: 4
• Pascal: 0
• Perl: 1
• Python: 2
• SQL: 3
• Visual Basic: 1
• Other (specify): 3
  1. LabVIEW
  2. Occassionally BASIS (LLNL)
  3. LABVIEW

26. Do any of the programs you write or maintain use any graphics software?
If so, please indicate any graphics software below. Please check all that apply.

• DirectX: 0
• DISSPLA: 4
• GA Plot Objects: 3
• IDL direct draw (e.g. plot, oplot): 16
• IDL object graphics (e.g. IDLgrSurface, IDLgrVolume): 1
• IDL widgets: 8
• Kylix/Delphi: 0
• Matlab: 4
• Motif: 0
• NCAR graphics: 2
• OpenGL: 0
• Other (specify): 2
  1. AVS
  2. pgplot

COMMENTS

27. Please take a moment to list any data analysis needs that could be solved with a new Data Analysis Application or with enhancements to existing Data Analysis Applications. (max 1000 characters)

• Number of response: 8
  1. I only really use ReviewPlus, and it is plagued by bugs. Please address!!
  2. It would be helpful if EFITviewer defaults to searching MDSPlus for the most recent shot.
  3. 1. A profile viewer that allows viewing (without fitting) all the profiles from a diagnostic (TS or CER or ECE) using a slider bar to move through time. Basically like the old 4D program did. 2. Modify GAprofiles so that fitting of the profile data can be turned off and allow the use to just look at the data without trying to fit it for every time slice looked at.
  4. Current project to make it easy to use CONT to analyze the Alfven gap structure is important to me.
  5. reviewplus still needs major work to make it a good application. Presently it is an essential application but it is also a bottleneck. There are many bugs and it is still difficult to do simple things. Many of the bugs are difficult to reproduce, indicating some sort of design problem. The old 4d application is still much better for control room use because it is easier to use and its performance is faster. Ctas requests for reviewplus are going untouched. My opinion is that the task of fixing reviewplus should be a much higher priority than starting construction of new tools that don't have as broad usage as reviewplus. You can quote me. John Ferron
  6. I find REVIEWPLUS frustrating to use, with a long learning curve that I still haven't ridden into competency. A multiple 'UNDO' button would speed my learning the intricacies without having to start over each time I make a mistake - and I make plenty, often without knowing exactly what I did wrong. I also don't use REVIEWPLUS because changing pointnames and other parameters I often use is relatively slow compared to the command line input capability of REVIEW. Adding a real, fully functional command line input capabiltiy would really enhance the speed experience for me. The faux REVIEW-style language command line currently in place has a LOT of troubles and is for most things not usable without great effort.
  7. GAProfiles is very awkward to use. For example, the user has to pick the equilibrium for each kinetic profile, when usually you want the same one carried forward. The spline control box has many unexplained numbers which appear; some documentation on their interpretation would be very useful.
  8. automation of some mundane LIBEAM analysis tasks.

28. Do you have any suggestions as to how to further increase the data analysis throughput rate at DIII-D?
Please take time to write down any specific ideas below. (max 1000 characters)

• Number of response: 3
  1. provide tools to more easily manage automated processing of data by individual users
  2. I am probably woefully lacking in knowing how to use what is available now. I, and I assume many physicists, prefer to get up to speed on these things by having working examples to ponder, not reading a manual with generic variable names or looking at cryptic symbology to for example learn how to read things out of mdsplus. So I can't comment usefully on how to improve things.
  3. WQe will see what happens, but usually the main problem that I have is slow access of PTData and MDSplus data during run times. We need to do a better job of monitoring and diagnosing data bottlenecks.