i590 interactive visual analytics
play

I590 Interactive Visual Analytics Week 5 | Sep 21, 2016 Data - PowerPoint PPT Presentation

I590 Interactive Visual Analytics Week 5 | Sep 21, 2016 Data Abstraction Visual Encoding Khairi Reda | redak@iu.edu School of Informa5cs & Compu5ng, IUPUI Project 1 Due Tuesday, Sep 27 at 8:59 PM Turn in as online web pages: Video with


  1. I590 Interactive Visual Analytics Week 5 | Sep 21, 2016 Data Abstraction Visual Encoding Khairi Reda | redak@iu.edu School of Informa5cs & Compu5ng, IUPUI

  2. Project 1 Due Tuesday, Sep 27 at 8:59 PM Turn in as online web pages: Video with Interac>ve, live Design narra>on visualiza>on Document

  3. Last 2 weeks

  4. C. Ware Visual Thinking for Design

  5. Cones, Rods, Color vision HyperPhysics, Georgia State University Wandell, “Foundations of Vision”

  6. Simultaneous contrast Via Colin Ware

  7. Simultaneous contrast Wong 2010 Via Miriah Meyer

  8. POPOUT

  9. Last 2 weeks Visual percep>on Our visual system sees differences, not absolute values, and is aGracted to edges. We can easily see objects that are different in color and shape, or that are in mo5on (popouts) Use color and shape sparingly to make the important informa5on pop out

  10. This week • Data abstrac>on • Datasets: Tables, Networks, Fields, Geometry. • Variable types: Categorical, Ordinal, and Quan5ta5ve variables • Visual encoding • Marks and channels • Perceptual accuracy of channels

  11. Data abstraction Transla5on of domain-specific terms into words that are as generic as possible type vs. seman>cs

  12. Terminology Data Type fundamental units Dataset Type collections of units

  13. Data Types • Items: are individual units of data. For example: rows in a table, points on a 5meline, nodes, … • AKribute: a property rela5ng to items, links, or posi5ons • Link: linkage between (typically) two items • Posi>on: designa5on of a point or a sub-space in a larger spa5al context • Grids: par55oning of space into typically uniform cells

  14. Dataset Types

  15. Flat Tables Key: Order ID Via Miriah Meyer

  16. Multidimensional Tables Alex Lex

  17. Visualizing Tables Example Parallel Coordinates

  18. Dataset Types

  19. Network / Graph items = people links = friendship

  20. Network / Graph Wikipedia Node-link diagram Matrix Mike Bostock

  21. Hierarchy/ Tree items = states links = winning paths

  22. Hierarchy/ Tree items = species links = evolu5onary ancestry http://orig00.deviantart.net/7534/f/2008/105/2/0/2027d31d532a91b7324f30af187e0819.png

  23. Hierarchy/ Tree items = zip codes links = containment http://orig00.deviantart.net/7534/f/2008/105/2/0/2027d31d532a91b7324f30af187e0819.png

  24. Dataset Types

  25. Grid attributes position population temperature density

  26. Fields Scalar fields aKribute = one value examples: temperature, eleva5on Vector fields aKribute = vector (direc>on) wind map

  27. Fields Scalar + Vector fields

  28. Fields Scalar fields aKribute = one (real) value examples: one MRI scan slice (5ssue density)

  29. 3D Fields Based on a slide by Miriah Meyer

  30. 3D Fields Based on a slide by Miriah Meyer grid = MRI scan slices aGributes = 5ssue density

  31. Visualizing 3D Fields Volume Rendering opacity tissue density http://www.fovia.com/gallery/medical/mg_bs.jpg http://www.dirkreiners.com/images/Research/tooth_tf.jpg

  32. Grid Types Uniform Grids Geometry & topology can be computed Rec>linear Grid Nonuniform sampling Structured Grid Unstructured Grid More flexibility, store posi5on and connec5on Wikipedia Slide by Alex Lex

  33. Dataset Types

  34. Geometry World Atlas

  35. Geometry http://farbackoutdoors.com/1593-2/

  36. Geometry

  37. Dataset Types

  38. Sets

  39. Clusters Dynamic (5me varying) Sta5c clusters clusters

  40. Lists LineUp Baseball salary vs. performance Ben Fry

  41. Attribute/Variable Types ✦ Categorical (Nominal, Qualita>ve) A finite set of categories No implicit ordering between categories ✦ Ordered •Ordinal Implicit ordering between categories/levels, but no clear magnitude difference. Can compare and determine greater/less than •Quan>ta>ve Meaningful magnitude Can do arithme5c

  42. Quan>ta>ve Data Interval vs. Ordinal ✦ Interval •Zero does not indicate an absence of detectable measurement •We can determine distance between measurement, but not propor5ons •Example: temperature, dates ✦ Ra>o •The posi5on of zero indicates there is nothing of the measured en5ty •Can determine ra5o and propor5ons •Example: weight, age

  43. Derived Data f(x) + g(x)

  44. Quiz What aGribute/variable type (Categorical, Ordinal, Interval, or Ra5o) best fit the following measurements? •Speed •Facebook reac5ons (Like, Angry, Sad, etc…) •Car configura5ons (Compact, Mid-Sedan, SUV) •Product Name •IQ scores •College Majors •50-meter race 5me Based on a slide by Alex Lex

  45. Another way to think about Ordered aKributes • Sequen5al temperature eleva5on (above/ • Diverging below sea level) • Cyclic 5me

  46. Colormap categorical sequential discrete vs. vs. vs. ordered diverging continuous Match colormap to data type & task

  47. Color design tools Helpful to match variable/aGribute type to colormap Color Brewer HCL Picker

  48. Project 1 Q&A

Recommend


More recommend