Developers Meet with Collaborators – Kelley 2114.
- Thursday, Feb 27, 12:30-2pm: Kirsten, Dani, Justin, Jenny, Kendra? (others welcome) Work session for VISTAS web site and Dani’s paper. Kirsten, Justin, Jenny gave Dani feedback on her paper (which Dani has submitted – need your paper to post!); Kirsten and Dani reviewed the web site and gave the following recommendations:
– web site recommendations ?????? - Thursday, Feb 27, 2 – 3:30: John, Patrick, with Vir (Justin, Dani, Jenny, Kirsten, Judy (late due to Bend weather) – Kendra ???).
1. review VISTAS running as backend to ENVISION. a major bug (seemingly in MFC – Windows event system – full message queue preventing screen update????) is preventing this. Vir really needs help from John’s crew to solve this problem. resolution???
2. collect and annotate new VISTAS visualizations of recent ENVISION data. Alexis was not at the meeting, and so this has not yet been done. Vir emailed Alexis with cc to Judy; Judy chimed in with why this is important (need visualization examples for June 15 VisWeek Poster submission).
3. Additional desiderata from John, as reported by Vir
– height exaggeration – need a slider (needs to be put into the ENVISION VISTAS plug-in and should be coded by John’s people – Patrick?!!)
– step back and forth with results ???
– have VISTAS run in Post Run Results, as well as interactively ????
– record movies to show later (has been implemented, but putting key frames is a little jerky – would be nice to smooth it out. Mike gave Vir some .net files to produce an AVI file (captures screen pixels and produces AVI).
– provide landmarks (Google geolocation API gets these on line) ideally, VISTAS would support different layers….
– roads important (use TIGER data set???) see data sets sent by Kirsten and John
– need county lines - Thursday, Feb 27, 3:30 – 5: Christoph, Kendra, Jenny, Vir, Judy, Justin, Dani, Mike, Kirsten).
1. Kendra showed status of her thesis project (due May 5) and visualization mockup of Primet McRae (Jerilyn’s thesis data) using WebGL and JavaScript; feedback and possible help with HJA data. need url for kendra ???
Christoph feedback:
– now hard coding, but will use not matlab, but raw ASCII (original SODAR) data ???
– will one be able to “move” the stations, interactively, for different data sets, e.g., by giving LAT/LONG, or clicking on a location; she’s been working on markers for the two stations using (LAT/LONG, UTMs so this should work for the different stations….)
– can we change contour distance (now at 50 m, e.g., 10, 30…)
– don’t use red/green contrast (for color blind persons -10% males, .01% females)
– want arrows anchored with the same y-axis, each with an anchor point
– would be nice to be able to give (interactively) which parameter to visualize (e.g., wind speed or direction, reflectivity, standard deviation, temp for terrain from PRISM data set. speed & reflectivity shown as arrows. her min requirements are just speed & direction. will see if these can be toggled
– would be great if he could view all 3 (speed, direction, reflectivity (or st dev) on single canvas. one way to do this would be to create a color-coded central pole as the tower, with each 10m bin colored with different color representing relative turbulence or dopler effect ???
– be sure to include context of the vis, i.e., date, time of day (don’t need compass direction). don’t need compass direction
– a limitation is that the color scale is only linear; need to adjust the scale and choose color. Dani is working on color intensity ideas…so rather than a color scheme, could perhaps use this….Nik says this is fairly common with color ramps…
– for v2.0 – be able to compute simple states and derive new variables, e.g., change in speed/height, wind direction/height.
– he would like the UI to be more intuitive, with camera control. Kendra now using a standard set of camera control…. she might be able to add in keyboard controls….
– I have an additional page of notes – does anyone else have notes ???? would someone volunteer to combine notes into one summary ????
– KENDRA – will your thesis blog be online ‘forever’ as part of the VISTAS record or should we abstract major findings/specs from it when you leave ???? ditto for Justin, Dani.2. Vir demo’d and explained his thesis project (3D line interval convolution) and ask for mesh wind data.
– for now he just shows a cylinder, xy plane, with 3D flow, and color codes speed.
– which of Christoph’s data should he use? the fog data is sparse and only includes changes over time. instead, Christoph has some continuous data (see his u-tube videos ?? where ???)
– Vir will aim for visualizing 1 night, for which Christoph’s group has done a full analysis. Christoph will send file access and documentation.
– Christoph noted a major challenge of his big data: he gets relatively easy $$$ to do experiments but gets so much data, that he can only use/analyze a small amount. the rest just sites on a shelf. - Thursday, 5pm: adjourn downtown to cash-only Squirrels (100 SW 2nd St) for beer and/or dinner; Kirsten is leading a group afterwards for dessert at del Alma (http://delalmarestaurant.com)
- Friday, Feb 28, 9-10:30: Bob, Allen, with Nik (Judy, Justin, Dani, Kirsten, Mike). Review recent VISTAS changes for usability and functionality in light of upcoming revisions, and (if possible) collect and annotate new VISTAS visualizations of recent VELMA runs for our record.
1. Nik demo’d recent changes to VISTAS : Project management – create, rename, delete, save, reload. still need to tie filename extension to program for auto-launch on double click (association with filetype will be done in the installer – Justin will look into this)
2. Nik reported that VISTAS seems to run OK with Bob’s new data….
3. We asked Bob to run VISTAS himself, and watched…. his major questions:
– to zoom
– loading data
– camera setting
4. Bob had a March 11 demo to BLM scheduled, and wanted to produce a movie for that…. he left with a new VISTAS, and we decided on a strategy for producing the video (Bob later reported that he did not have time to run the VISTAS visualizations, so will do the presentation without the movie).
5. we decided to write up a ‘typical’ scenario to post (and use for our poster). perhaps eventually to make a video.
6. I have a long list of ‘todo’s that came from this session; will integrate my notes with Justin’s…and post them here….
HCI consultation. Developers Meet with Margaret Burnett – Kelley 2114.
- 10:30-12: Judy, Dani, and Justin working meeting at HGI or Kelley, to refine their senior project specs and Justin’s notes on Thursday Collaborators’ sessions (for Margaret).
- 12:30-1:30: Lunch (Nearly Normals?)
- 2-5: 2-4? Meet with Margaret Burnett.
Objectives:
– Running better how to devise user studies, given our user community? do we develop use cases but of what granularity?
– how to better run sessions with our users to get at functionality and UI issues
– how to balance functionality and UI priorities
– Identify low-hanging fruit UI changes or egregious UI problems –
running sessions for users to get at this
– better understanding how scientist users differ from non-scientists, and how new developers differ from scientists. We most ‘care’ about our users; they are the ambassadors to VISTAS stakeholders.
can we get at this by talking about Margaret’s gender difference studiesAgenda: -
- (5 min) Quick Statement of Meeting Goals & Objectives, Introductions (Judy and Mike).
- (15 min) VISTAS demonstration (Nik, then Vir)
Ask Margaret for her initial UI observations. - (30 min) Rather than reporting on our recent meetings with collaborators (yesterday and today), ask Margaret how we might find out from our users: Collaborators’ Commentary on the functionality and usability
- (90 min) short break, then Discussion/Work Session – mock up proposed user sessions….and granularity of user tasks we give them?
– how much do we (the developers) get involved with the user as they are working.
– how do we get at a half dozen or so typical user scenarios that would get users to expert VISTAS user level.
– what do we want to produce about those scenarios to maintain and develop users expertise.
– how do we test that ‘documentation’. - (15 min) recap Action Items
Resources for Margaret prior to meeting by Jan. 31 (3-4 pages max) :
- blurbs on VISTAS project goals & objectives (1 p. max) and our visualization niche (1 para max). include statement of research objectives
- short profile of our user community – differentiate primary and secondary users (developers and their end users
- profile of VISTAS research and development team
- other, at her request