Data Presentation - Lecture 4 - Information Visualisation (4019538FNR)

Data Presentation - Lecture 4 - Information Visualisation (4019538FNR)

This lecture forms part of the course Information Visualisation given at the Vrije Universiteit Brussel.

1135dc242dcff3b90ae46fc586ff4da8?s=128

Beat Signer

March 03, 2020
Tweet

Transcript

  1. 2 December 2005 Information Visualisation Data Presentation Prof. Beat Signer

    Department of Computer Science Vrije Universiteit Brussel beatsigner.com
  2. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 2

    March 3, 2020 Information Visualisation Process Data Representation Data Data Presentation Interaction perception and visual thinking mapping
  3. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 3

    March 3, 2020 Marks and Channels  Marks are basic geometric elements to represent items or links  Channels control the appearance of marks  Vis design space described by orthogonal combination of marks and channels  Complex visual encodings can be decomposed and analysed in terms of their marks and channels  building blocks for analysing visual encodings
  4. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 4

    March 3, 2020 Marks  Basic geometric/graphical element in an image  classified according to the number of spatial dimensions
  5. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 5

    March 3, 2020 Marks …  Zero-, one- or two-dimensional marks (three-dimensional marks are not used frequently)
  6. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 6

    March 3, 2020 Mark Types  Item marks  Link marks  connection marks - pairwise relationship between two items via a line  containment marks (enclosure or nesting) - hierarchical relationships using areas
  7. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 7

    March 3, 2020 Channels
  8. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 8

    March 3, 2020 Channels …  Control appearance of mark independently of the dimensionality of the geometric primitive  Many visual channels  spatial position  shape  colour (hue, saturation and luminance)  motion (e.g. flicker, direction and velocity)  size (i.e. length, area and volume)  tilt (angle)  Size and shape channels cannot be used on all types of marks  e.g. area marks typically not size or shape coded
  9. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 9

    March 3, 2020 Channel Types  Identity channels  information about what something is  e.g. shape, hue colour channel, motion pattern  Magnitude channels  how much of something is there  e.g. size (length, area or volume), luminance or saturation colour channels, angle, …
  10. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 10

    March 3, 2020 Using Marks and Channels  Progression of chart types  one quantitative and one categorical attribute (a)  scatterplot with two quantitative attributes (b)  two quantitative and one categorical attribute via hue (c)  three quantitative (one via size) and one categorical attribute (d)  Each attribute encoded via a single channel in this examples  multiple channels might also be used redundantly
  11. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 11

    March 3, 2020 Using Marks and Channels …  Use of marks and channels guided by the principles of expressiveness and effectiveness  after identifying most important attributes ensure that they are encoded with the highest ranked channel  Expressiveness principle  visual encoding should express all of, and only, the information in the dataset attributes - ordered data should be shown in a way that our perceptual system senses as ordered  use magnitude channels - unordered data should not be shown in a way that perceptually implies an ordering that does not exist  use identity channels
  12. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 12

    March 3, 2020 Using Marks and Channels …  Effectiveness principle  importance of attribute should match the salience of the channel  most important attributes encoded with most effective channels  Attributes encoded with position will dominate the user's mental model  choice of which attributes to encode with position is the most central choice in visual encoding
  13. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 13

    March 3, 2020 Channel Effectiveness [Visualization Analysis & Design, Tamara Munzner, 2014]
  14. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 14

    March 3, 2020 Channel Effectiveness …  Obvious way to quantify effectiveness via accuracy  how close is human perceptual judgement to some objective measurement of the stimulus?  Different visual channels are perceived with different levels of accuracy  characterised by Steven's Psychophysical Power Law
  15. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 15

    March 3, 2020 Steven's Psychophysical Power Law  Responses to sensory experience of magnitude are characterisable by power laws  = perceived sensation  I = physical intensity  exponent N depends on sensory modality  most stimuli are magnified (superlinear) or compressed (sublinear) [Visualization Analysis & Design, Tamara Munzner, 2014]
  16. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 16

    March 3, 2020 Error Rates Across Channels Results by Cleveland and McGill, 1984
  17. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 17

    March 3, 2020 Channel Effectiveness …  Channel effectiveness mainly based on accuracy but also takes into account  discriminability  separability  popout  grouping
  18. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 18

    March 3, 2020 Discriminability  Quantify the number of distinguishable steps (bins) that are available within a visual channel  some channels (e.g. line width) have a very limited number of bins  small number of bins is not a problem if the number of values to be encoded is also small  number of different values that need to be shown for an attribute must not be greater than the available bins for the visual channel
  19. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 19

    March 3, 2020 Effective Line Width Use  Limited number of discriminable bins
  20. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 20

    March 3, 2020 Separability  Channels are not always completely independent from each other (interchannel interference)  ranging from fully separable channels to the inextricably combined integral channels (major interference)  Visual encoding straightforward with separable channels  encoding of different information in integral channels will fail
  21. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 21

    March 3, 2020 Popout  Many channels provide visual popout (preattentive processing) where a distinct item stands out from many others immediately  time to spot the different object does not depend on the number of distractor objects (a) vs. (b)  massively parallel processing of low-level features  popout effect slower for shapes ((c) and (d)) than for colour hue channel ((a) and (b))  channels with individual popout cannot simply be combined ((e) and (f)) - need serial search to find the red circle in (f)  Most pairs of channels do not support popout  use popout for a single channel at a time
  22. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 22

    March 3, 2020 Popout …
  23. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 23

    March 3, 2020 Popout Channels  Popout cannot only occur for colour hue and shape channels  tilt  size  shape  proximity  shadow direction
  24. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 24

    March 3, 2020 Popup Channels …
  25. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 25

    March 3, 2020 Grouping  Containment (links) is the strongest cue for grouping followed with connection coming in second  Items sharing the same level of a categorical attribute can also be perceived as a group  Proximity is the third strongest grouping approach  Similarity (hue, motion and shape)  shape and motion channel to be used with care
  26. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 26

    March 3, 2020 Relative versus Absolute Judgements  Perceptual system fundamentally based on relative judgements and not absolute ones (Weber's Law)  e.g. position along a scale can be perceived more accurately than pure length judgement without a scale
  27. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 27

    March 3, 2020 Relative Luminance Perception  Perception of luminance is contextual based on the contrast with surrounding colours
  28. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 28

    March 3, 2020 Colour (Hue) Perception  Our visual systems evolved to provide colour constancy  same surface identifiable across illumination conditions  visual system might work against simple colour encodings [Visualization Analysis & Design, Tamara Munzner, 2014]
  29. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 29

    March 3, 2020 Mapping Colour  Luminance and saturation are magnitude channels while hue is an identity channel  luminance can be used for two to four levels (bins)  saturation can be used for up to three levels (bins) - strongly interacts with size channel  saturation and hue are non- separable channels
  30. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 30

    March 3, 2020 Mapping Colour …  Can use up to moderate six to twelve distinguishable hue steps (bins) for small separated regions  includes background colour and default object colours  use easy nameable colours: e.g. red, blue, green, yellow, orange, brown, pink, magenta, purple and cyan  Use highly saturated colours for small regions  Use low-saturation colours (pastels) for large regions
  31. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 31

    March 3, 2020 Comparing HSL Lightness  Computed HSL lightness L is the same for all six colours  true luminance as measured by an instrument  perceived luminance L* represents what we see - more sensitive to certain wavelengths (green and yellow) as shown earlier with the spectral sensitivity
  32. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 32

    March 3, 2020 No Implicit Order for Hue  Sometimes learned hue orders (not at perception level)  green-yellow-red from traffics lights  rainbow colour ordering
  33. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 33

    March 3, 2020 Colourmaps  A colourmap defines a mapping between colours and data values  Colourmaps can be categorical or ordered (sequential or diverging)  use magnitude channels of luminance and saturation for ordered data
  34. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 34

    March 3, 2020 Colourmap Categorisation [Visualization Analysis & Design, Tamara Munzner, 2014]
  35. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 35

    March 3, 2020 Categorical Colourmaps  Categorical colourmaps (qualitative colourmaps) are normally segmented (not continous)  effective for categorical data (next best channel after position)  Good resource for creating colourmaps is ColorBrewer
  36. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 36

    March 3, 2020 Ineffective Categorical Colourmap Use
  37. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 37

    March 3, 2020 Example of Using Additional Channels  Dataset with 27 categorical levels from 7 categories
  38. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 38

    March 3, 2020 Example of Using Additional Channels …
  39. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 39

    March 3, 2020 Example of Using Additional Channels …
  40. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 40

    March 3, 2020 Example of Using Additional Channels …
  41. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 41

    March 3, 2020 Ordered Colourmaps  Sequential colourmap ranges from a minimum value to a maximum value  use luminance (with or without hue) or saturation channel  Diverging colourmap  use two hues at the endpoints and a neutral colour (e.g. white or grey) as a midpoint
  42. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 42

    March 3, 2020 Rainbow versus Two-Hue Colour Map  How many hues to use in continous colourmaps?  high-level structure versus local neighbourhoods (fine grained details)  rainbow colourmap makes it easier to discuss specific (nameable) subranges
  43. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 43

    March 3, 2020 Rainbow Continous Colourmaps  Problems of rainbow continous colourmaps  hue is use to indicate order (despite being an identity channel)  scale is not perceptually linear  fine details cannot be perceived via the hue channel - luminance channel much better (luminance contrast required for edge detection in our eyes)
  44. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 44

    March 3, 2020 Rainbow Continous Colourmaps …  The three problems of rainbow continous colourmaps can be addressed by using monotonically increasing luminance colourmaps  multiple hues are ordered according to their luminance from lowest to highest  Rainbow colourmap  standard rainbow colour- map (a) vs. perceptually linear rainbows (b) with decreased dynamic range  segmented rainbow for categorical data (c)
  45. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 45

    March 3, 2020 Bivariate Colourmaps  Safest use of colour channel is to visually encode a single attribute (univariate)  In the colourmap categorisation we have seen colourmaps encoding two separate attribute (bivariate)  if one of the two attributes is binary then it is straightforward to create a comprehensible bivariate colourmap - choose base set of hues and vary the saturation  if both attributes are categorical with multiple levels the results will be poor  combinations of sequential and diverging attributes should be used carefully - appear frequently in vis solutions but some people have difficulties to interpret their meaning
  46. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 46

    March 3, 2020 Colourblind-Safe Colourmaps  A safe strategy is to avoid using the hue channel only  e.g. vary luminance or saturation in addition to hue in categorical colourmaps  Avoid colourmaps emphasising red-green (divergent red-green ramps)  Use colour blindness simulators and tools such as Viz Palette
  47. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 47

    March 3, 2020 Size Channels  Suitable for ordered data and interacts with most other channels  length (1D) - judgment of length is very accurate  area (2D) - judgement of area is less accurate  volume (3D) - volume channel is quite inaccurate
  48. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 48

    March 3, 2020 Angle (Tilt) Channel  Encode magnitude information based on the orientation of a mark  angle: orientation of a line with respect to another line  tilt: orientation against the global frame of the display  Accuracy of our perception of an angle is not uniform  very accurate near exact horizontal, vertical or diagonal positions
  49. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 49

    March 3, 2020 Other Channels  Shape channel  commonly applied to point marks  can also be applied to line marks (e.g. dotted or dashed lines)  can distinguish between dozens up to hundreds bins - strong interaction between shape and size channel  Motion channels  direction of motion  velocity of motion  flicker and blinking frequency  very separable from all other static channels  strongly draws attention - hard to ignore and should be used carefully
  50. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 50

    March 3, 2020 Other Channels …  Texture and stippling channel  texture can be simplified by considering it as a combination of the following three perceptual dimensions - orientation, scale and contrast  texture can be used to show categorical attributes as well as ordered attributes  Stippling fills regions of drawings with short strokes - e.g. dashed or dotted lines - used for area marks in older printing (to simulate grey)
  51. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 51

    March 3, 2020 Exercise 4  Preprocessing and Statistics in R
  52. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 52

    March 3, 2020 Further Reading  This lecture is mainly based on the book Visualization Analysis & Design  chapter 5 - Marks and Channels  chapter 10 - Map Color and Other Channels
  53. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 53

    March 3, 2020 References  Visualization Analysis & Design, Tamara Munzner, Taylor & Francis Inc, (Har/Psc edition), May, November 2014, ISBN-13: 978-1466508910  Semiology of Graphics: Diagrams, Networks, Maps, Jacques Bertin, ESRI PR (1st edition), January 2010, ISBN-13: 978-1466508910  Information Visualization: Perception for Design, Colin Ware, Morgan Kaufmann (3rd edition) May 2012, ISBN-13: 978-0123814647
  54. Beat Signer - Department of Computer Science - bsigner@vub.ac.be 54

    March 3, 2020 References …  ColorBrewer  http://colorbrewer2.org  Viz Palette  https://projects.susielu.com/viz-palette
  55. 2 December 2005 Next Lecture Data Processing and Visualisation Toolkits