Computational Methods in Physics PHYS 3437

1. Computational Methods in Physics PHYS 3437 Dr Rob Thacker Dept of Astronomy & Physics (MM-301C) thacker@ap.smu.ca

2. Today’s Lecture • Visualization • Useful results to know about perception of information • Help you to gain some of idea of “why that looks bad” • How not to visualize • Beginning using Opendx (more next lecture) • Sources for today’s lecture: • http://www.research.ibm.com/dx/proceedings/pravda/truevis.htm • Spatial information & colour maps • http://www2.sims.berkeley.edu/courses/is247/f05/schedule.html • Lectures by Marti Hearst • http://www.cs.unb.ca/acrl/training/visual/macphee-intro_to_visual/Intro_to_Visual.ppt • Introduction to Opendx by Chris MacPhee

3. Visualization as computing • “Visualization is a method of computing. It transforms the symbolic into the geometric, enabling researchers to observe their simulations and computations. Visualization offers a method for seeing the unseen. It enriches the process of scientific discovery and fosters profound and unexpected insights. In many fields it is revolutionizing the way scientists do science.” - Visualization in Scientific Computing, ACM SIGGRAPH, 1987

4. Why use Visualization? “A picture says more than a thousand words.” “A picture says more than a thousand numbers.” “The purpose of [scientific] computing is insight, not numbers.” - Dr. Richard Hamming, Naval Postgraduate School, California “... half of the human brain is devoted directly or indirectly to vision ...” - Prof. Mriganka Sur, Brain and Cognative Sciences, MIT

5. Visualization vs. Analysis? • Visualization is best applied to data mining and data discovery • Visualization tools are helpful for exploring hunches and presenting results • Example: scatterplots • Visualization is the WRONG primary tool when the goal is to find a good model in a complex situation • May provide hints but won’t provide concrete answers • Model building requires insight into the problem at hand

6. Value of visualization for the cynical! “Conclude your technical presentation and roll the [videotape]. Audiences love razzle-dazzle color graphics, and this material often helps deflect attention from the substantive technical issues.” David Bailey, NERSC

7. Simulation of Hurricane Earl (1998)

8. Preattentive Processing • A limited set of visual properties are processed preattentively • without need for focusing attention • Note, this is critical in web-site design • 4 seconds before user decides they don’t understand the page… • Important for design of visualizations • what can be perceived immediately • Does the viewer get the information without having to consciously process the image? • what properties are good discriminators?

9. Pre-attentive Processing • < 200 - 250ms qualifies as pre-attentive • eye movements take at least 200ms • yet certain processing can be done very quickly, implying low-level processing in parallel • If a decision takes a fixed amount of time regardless of the amount of information presented, it is considered to be preattentive

10. Example: Color Selection We can instantly see the red dot – we have preemptively processed the different hue.

11. Example: Shape Selection Viewer can rapidly and accurately determine whether the target (red circle) is present or absent. Difference detected in form (curvature)

12. Example: Conjunction of Features Finding the target (red circle) requires that we sequentially search through the image. We can’t rapidly (or accurately) determine the presence of the target when there are two or more features that differentiate it from the remaining information (distractors).

13. Example: Emergent Features Despite being constructed from similar components to the distractors, the unique feature of the target (open sides) allows us to process its presence preattentively.

14. Example: Emergent Features We cannot detect the target preattentively as it has no unique feature relative to the distractors.

15. Asymmetric and Graded Preattentive Properties • Some properties are asymmetric • a sloped line among vertical lines is preattentive • a vertical line among sloped ones is not • Some properties have a gradation • some more easily discriminated among than others

16. SUBJECT PUNCHED QUICKLY OXIDIZED TCEJBUS DEHCNUP YLKCIUQ DEZIDIXO CERTAIN QUICKLY PUNCHED METHODS NIATREC YLKCIUQ DEHCNUP SDOHTEM SCIENCE ENGLISH RECORDS COLUMNS ECNEICS HSILGNE SDROCER SNMULOC GOVERNS PRECISE EXAMPLE MERCURY SNREVOG ESICERP ELPMAXE YRUCREM CERTAIN QUICKLY PUNCHED METHODS NIATREC YLKCIUQ DEHCNUP SDOHTEM GOVERNS PRECISE EXAMPLE MERCURY SNREVOG ESICERP ELPMAXE YRUCREM SCIENCE ENGLISH RECORDS COLUMNS ECNEICS HSILGNE SDROCER SNMULOC SUBJECT PUNCHED QUICKLY OXIDIZED TCEJBUS DEHCNUP YLKCIUQ DEZIDIXO CERTAIN QUICKLY PUNCHED METHODS NIATREC YLKCIUQ DEHCNUP SDOHTEM SCIENCE ENGLISH RECORDS COLUMNS ECNEICS HSILGNE SDROCER SNMULOC Text is NOT Preattentive!

17. Preattentive Visual Properties(Healey 97) length Triesman & Gormican [1988] width Julesz [1985] size Triesman & Gelade [1980] curvature Triesman & Gormican [1988] number Julesz [1985]; Trick & Pylyshyn [1994] terminators Julesz & Bergen [1983] intersection Julesz & Bergen [1983] closure Enns [1986]; Triesman & Souther [1985] colour (hue) Nagy & Sanchez [1990, 1992]; D'Zmura [1991] Kawai et al. [1995]; Bauer et al. [1996] intensity Beck et al. [1983]; Triesman & Gormican [1988] flicker Julesz [1971] direction of motion Nakayama & Silverman [1986]; Driver & McLeod [1992] binocular lustre Wolfe & Franzel [1988] stereoscopic depth Nakayama & Silverman [1986] 3-D depth cues Enns [1990] lighting direction Enns [1990]

18. Globus & Raible, 1994 Humour: 14 ways to say nothing with visualization • Never include a color legend • Avoid annotation • Never mention error characteristics • When in doubt smooth • Avoid providing any performance data • Cunningly use stop-frame techniques • Never learn anything about the underlying data or discipline

19. Humour: 14 ways to say nothing with visualization - 2 (8) Never provide contrasting visualizations to other data (9) Always ensure you develop your own new tool and disregard others as out-dated and out-moded (10) Don’t cite references for data (11) Claim generality but only ever show results from one data set (12) Use viewing angle to hide unwanted information (13) If something can’t be hidden by choosing an angle, use shadows (14) “This is easily extended to 3d!”

20. Representing different types of data • Nominal data • Data is put into categories with no implicit ordering • e.g. red and blue cars • Should be represented by distinguishably different objects without any perceived ordering • Ordinal data • Data is put into categories that have an implied ordering structure • e.g. A,B,C.. grades in a class • Should be represented by distinguishably different objects with a perceived ordering • Interval data • Data is not categorized and is instead described by a numeric system • e.g. temperatures, most scientific data • Equal steps in data value should appear as steps of equal perceived magnitude in the representation

21. HSV colour space • The HSV (Hue, Saturation, Value) model, defines a color space in terms of three constituent components: • Hue, the color type (e.g. red, blue, or yellow): • Ranges from 0-360 (but normalized to 0-100% in some applications) • Saturation, the "vibrancy" of the color: • Ranges from 0-100% • Also sometimes called the "purity" by analogy to the colorimetric quantities excitation purity and colorimetric purity • The lower the saturation of a color, the more "grayness" is present and the more faded the color will appear • Value, the brightness of the color: • Ranges from 0-100% Much more strongly related to the human perception of colour than RGB

22. Colour maps • The most common (default) colormap is the “rainbow” map (shown below) • maps the lowest value in the variable to blue, the highest value to red, and interpolates in color space (red, green, blue) to produce a color scale. • Produces several well-documented artifacts • You will percieve 5 layers in the visualization • Yellow regions are perceived as more significant due to their brighter colour

23. All plots use the same data, but different colourmaps give the appearance of less or more information

24. Perception of magnitude • Often need to visualize a single variable at many places (scalar field) • In many cases, the interpretation of the data depends on having the visual picture accurately represent the structure in the data • In order to accurately represent detailed information the “visual dimension” chosen should appear continuous to the user • Rules out the rainbow colourmap immediately • Perceived magnitude obeys a power relationship with physical luminance over a very large range of gray scales • Explains why grayscale colormaps are commonly used in medical imaging • Another method which displays this behavior is color saturation, the progression of a color from vivid to pastel Value increases monotonically, while saturation becomes more pastel

25. Perception of spatial frequency • The value component in a color (the brightness/darkness component) is critical for carrying information about high spatial frequency variations in the data • If the colour map does not contain a monotonic value variation, fine resolution information will not be seen • The saturation and hue components in color are critical for carrying information about low spatial frequency variations in the data • A colour map which only varies in luminance (e.g., a grayscale image) cannot adequately communicate information about gradual changes in the spatial structure of the data

26. Low frequency information Two colours allow you to pick out the larger systems High frequency information Two colour map over emphasizes large scale features, and some detail around these features is lost

27. Segmented maps Better low freq info? Losing high freq info? But not apples to apples comparison

28. Key Questions to Ask about a Viz • Is it for analysis or presentation? • What does it teach/show/elucidate? • What is the key contribution? • What are some compelling, useful examples? • Could it have been done more simply? • Have there been usability studies done? What do they show?

29. Are we just limited to 3d? • A visualization can use x, y, and z to represent the spatial dimensions of an object • color can be mapped onto a surface representing a fourth • the surface can be deformed according to a fifth • isocontour lines can represent a sixth • coloring them can represent a seventh • glyphs on the surface can represent a few more, not to mention animation, transparency, and stereo • This great flexibility, however, can open a Pandora's box of problems for the user, • can easily give rise to visual representations which do not adequately represent the structure in the data or which introduce misleading visual artifacts

30. Public domain viz tools • While there are a host of public domain tools, the two most popular are probably • VTK – The Visualization toolkit • http://www.vtk.org/ • Mid-level library that requires you construct scripts (Tcl-Tk) to run your visualization • Very powerful, allows you to wrap visualization code in with your own C++ • Drawback – fairly steep learning curve • Opendx • Freely available, packaged visualization program (as opposed to library) • Quick to get going with, so we’ll use it in this course

31. About Opendx • Began as an IBM product: “Visualization Data Explorer” • IBM released it Open Source and it was renamed Opendx • Note, they held back some patented routines, but most of the nuts and bolts are there • Approach to visualization is to create a network of functions that link together within a visual program editor (“VPE”) • Takes a while to get used to, but once you are familiar with it things are very easy • There is a large body of additional modules made available by other users • Great resource!

32. Each module has a specific action Many of the modules have hidden features as well Why this format? Related to the concept of a rendering pipeline A simple visual program example

33. Windows: If you are running the windows version you’ll need an X-server running Type startx at the Cygwin prompt to do this Linux: type dx at the command prompt Getting started with OpenDX Main dx panel http://www.opendx.org

34. Steps in creating a dx program • While there are many approaches the easiest way to begin with is • Import data into dx • Click on the “Import data” button • You will need to describe the precise format though • Write the visual program using the VPE • Click on “Edit Visual Programs” button

35. Opendx example • If time…

36. Summary • The HSV colour space is much more closely related to human perception than RGB • Some information can be processed preattentively and successful visualizations can exploit this • The standard rainbow colour map has two significant artifacts for visualization • 5 layers are explicitly represented • Yellow tends to dominate visually • Describing high frequency information is best achieved using value and saturation based colour maps • Low frequency information is elucidated well using hue based maps • Opendx is very powerful, but free, tool that originated out of the IBM Data Explorer project

37. Next lecture • Visualization & data representation • More on Opendx • 3d visualization methods • Isosurfaces • Volume rendering