Gradients and 2D Visualization of Multiple Variables

CS137 Assignment #2

1. Dates

2. Goals

1. Gain an understanding of the sources of experimental and computational data.
2. Learn about scientific data and identify relationships among variables in scientific data.
3. Learn to create a visualization with streamlines and texture fields showing variations in the relationships within variables in a dataset.
4. Explore and experiment with color and shape gradients, layering, metaphorical reading of visual elements, and legend design.

3. Readings

• For Part A:
  • Color and Information, ch. 3 of Envisioning Information by Edward Tufte
  • Visual Thinking for Design by Colin Ware, chapters:
    • 2: What We Can Easily See
    • 3: Structuring Two-Dimensional Space
  • Loose, Artistic Textures for Visualization by David H. Laidlaw. IEEE Computer Graphics and Applications, 21(2):6-9, March/April 2001.
  • Patterns in Nature by Peter Stevens, pp.53–71. pp.53–67 are in the linked PDF; see the book in the Cave room for the additional pages.
• For Part B:
  • World of Science Encyclopedia, Volume 14, pp.119-128. Motion Of Fluids [2]
  • Comparing 2D vector field visualization methods: A user study, by David H. Laidlaw, Michael Kirby, Cullen Jackson, J.Scott Davidson, Timothy Miller, Marco DaSilva, William Warren, and Michael Tarr. TVCG, 11(1):59-70, January 2005.
  • Strategies for the Visualization of Multiple 2D Vector Fields, by Timothy Urness, Victoria Interrante, et al. IEEE CG&A, 26(4):74-82, July 2006.

4. Assignment

4.1 Part A: Gradient Design

A gradient is defined for this assignment as any gradually changing visual element or visual quality: a color blend or fade, a morphing texture, a thickening line or accumulation of lines, a gradual change in orientation of a line, form, or pattern. After re-reading the text by Stevens from Assignment 1 on the variables involved in fluid flow, begin the design process for a 2D visualization of fluid flow by generating a resource library of as many different kinds of gradients as you can imagine. A list of some possible qualities or characteristics follows.

Consider the legibility of each gradient; its change should be easily perceivable. Ideally, there should only be one varying visual aspect per gradient, but you can experiment also with marks that contain two or more separately varying qualities. This is fine as long as the two varying qualities can be adjusted independently of each other. Consider also its legibility in a complex "layered" situation, in the company of other gradients. Not only should the gradient be legible in "mixed company";, but also it should not interfere with the legibility of its neighbors. You can highlight the "combinability" of your gradients by showing them in tandem usage with others. Finally, consider the "intuitive reading" of your various gradients: the possible interpretation a viewer might have of the coded meaning shown by the change that takes place in the gradient: intensification of force; increase/decrease of velocity; temperature etc.

There are three sets of compositions that you have to make for this assignment:

1. Make a "library" chart with at least 15 different visual gradients, showing the full range of the variation of each. Each gradient should be depicted by itself, not in conjunction with anything else.
2. Consider each gradient's legibility in a complex layered situation with other gradients. Each gradient should be legible when layered and should not interfere with the legibility of its neighbors. Make a new chart with 10 combinations of two or more gradients from your library. Try to mix gradients that have easily discernible difference in terms of their associative character.
3. Finally, make five gradient compositions, each with five or more layered gradients, adjusted for co-legibility, as described above. Imagine a 2D field with forces playing across it, each represented by a different, clearly legible gradient.

4.2 Part B: 2D Graphing of Multiple Variables

Design a 2D visualization for fluid flow around a cylinder, with six variables, based on the image in the handout (see David's slides for bigger versions). Begin by assessing the data, and consulting your gradient library for possible visual representation for each variable in the data. Ideally, the gradient you choose for a specific variable should have an intuitive reading connected with your understanding of the variable: e.g. velocity should look like velocity. In addition, your choices should take layering into account: the gradients should not interfere with each other in combination. Visual simplicity of the whole will enhance individual value legibility and legibility of data interconnection.

Design a clear, well-organized legend or key for each of your visualizations, detailing each gradient's change in relation to the changing value of each data variable. Label the axes of the keys with numeric data values. You'll need to make up the numbers. In general, zero will be clear. Make your own choice about non-zero values and try to also fill in the units (e.g. meters/second for velocity).

5. Questions (answer briefly)

Questions are due, emailed to the TA, by 11am on the date listed above. Most of these are very short questions intended to help guide you through the assignment. They should not take long to answer. Most of the time, a very brief, one sentence or less, answer is sufficient.

1. How well are data values discernable in your visualization?
2. How well are relationships among data values visualized easily by your design?
3. What are the factors behind your choice of visual characteristics that show gradient?
4. What are additional factors that need to be considered when designing visualization for the bat wings that are different when you think about this problem in 3D in the Cave as opposed to in 2D on paper?