Lectures 5&6: analyze idiom structure marks (geoms) Points - PowerPoint PPT Presentation

Visual encoding Definitions: Marks and channels Lectures 5&6: • analyze idiom structure • marks (geoms) Points Lines Areas – geometric primitives Perception & Color, Marks and Channels • channels (aesthetics) Rules of Thumb Position Color – control appearance of marks (Geoms and Aesthetics) Horizontal Vertical Both – can redundantly code with multiple channels Tamara Munzner Department of Computer Science Shape Tilt Perceptual Principles University of British Columbia DSCI 532, Data Visualization 2 Week 3, Jan 16 / Jan 18 2018 Size Length Area Volume www.cs.ubc.ca/~tmm/courses/mds-viz2-17 @tamaramunzner 2 3 4 Visual encoding Channels: Expressiveness types and effectiveness rankings Channels/Aesthetics: Matching Types Channels/Aesthetics: Rankings • analyze idiom structure Magnitude Channels: Ordered Attributes Identity Channels: Categorical Attributes Magnitude Channels: Ordered Attributes Identity Channels: Categorical Attributes Magnitude Channels: Ordered Attributes Identity Channels: Categorical Attributes Position on common scale Spatial region Position on common scale Spatial region Position on common scale Spatial region –as combination of marks/geoms and channels/aesthetics Position on unaligned scale Color hue Position on unaligned scale Color hue Position on unaligned scale Color hue Length (1D size) Motion Length (1D size) Motion Length (1D size) Motion Tilt/angle Shape Tilt/angle Shape Tilt/angle Shape Area (2D size) Area (2D size) Area (2D size) • expressiveness principle • expressiveness principle 1: 
 2: 
 3: 
 4: 
 Depth (3D position) Depth (3D position) Depth (3D position) –match channel/aesthetics & data –match channel/aesthetics & data vertical position vertical position vertical position vertical position characteristics characteristics Color luminance Color luminance Color luminance horizontal position horizontal position horizontal position • effectiveness principle color hue color hue Color saturation Color saturation Color saturation size (area) –encode most important attributes with Curvature Curvature Curvature highest ranked channels mark: line mark: point mark: point mark: point Volume (3D size) Volume (3D size) Volume (3D size) 5 6 7 8 Channels/Aesthetics: Spatial position Accuracy: Fundamental Theory Accuracy: Vis experiments Discriminability: How many usable steps? Cleveland & McGill’s Results • must be sufficient for number of Magnitude Channels: Ordered Attributes Identity Channels: Categorical Attributes Position on common scale Spatial region attribute levels to show –linewidth: few bins but salient Position on unaligned scale Color hue Positions Length (1D size) Motion 1.0 1.5 2.0 2.5 3.0 Log Error Tilt/angle Shape Crowdsourced Results Area (2D size) • expressiveness principle Angles Depth (3D position) –match channel and data characteristics Circular Color luminance • effectiveness principle [Crowdsourcing Graphical areas Perception: Using Mechanical Turk –encode most important attributes with Color saturation to Assess Visualization Design. Rectangular areas Heer and Bostock. Proc ACM Conf. highest ranked channels (aligned or in a Human Factors in Computing Curvature treemap) –spatial position ranks high for both Systems (CHI) 2010, p. 203– [mappa.mundi.net/maps/maps 014/telegeography.html] 212.] 1.0 1.5 2.0 2.5 3.0 Volume (3D size) 9 10 after Michael McGuffin course slides, http://profs.etsmtl.ca/mmcguffin/ Log Error 11 12 Separability vs. Integrality Popout Popout Grouping Marks as Links • find the red dot Containment Connection –how long does it take? • containment Position Size Width Red • parallel processing on many individual • connection Hue (Color) Hue (Color) Height Green channels –speed independent of distractor count Identity Channels: Categorical Attributes –speed depends on channel and amount of • proximity difference from distractors Spatial region –same spatial region • serial search for (almost all) combinations Color hue • similarity Fully separable Some interference Some/signi fj cant Major interference –speed depends on number of distractors interference –same values as other Motion categorical channels 2 groups each 2 groups each 3 groups total: 4 groups total: • many channels: tilt, size, shape, proximity, shadow direction, ... integral area integral hue • but not all! parallel line pairs do not pop out from tilted pairs Shape 13 14 15 16

Relative vs. absolute judgements Relative luminance judgements Relative color judgements Further reading • perceptual system mostly operates with relative judgements, not absolute • perception of luminance is contextual based on contrast with • color constancy across broad range of illumination conditions • Visualization Analysis and Design. Munzner. AK Peters Visualization Series, CRC Press, 2014. surroundings –that’s why accuracy increases with common frame/scale and alignment – Chap 5: Marks and Channels –Weber’s Law: ratio of increment to background is constant • On the Theory of Scales of Measurement. Stevens. Science 103:2684 (1946), 677–680. • filled rectangles differ in length by 1:9, difficult judgement • Psychophysics: Introduction to its Perceptual, Neural, and Social Prospects. • white rectangles differ in length by 1:2, easy judgement Stevens. Wiley, 1975. • Graphical Perception: Theory, Experimentation, and Application to the Development of Graphical Methods. Cleveland and McGill. Journ. American Statistical Association 79:387 (1984), 531–554. • Perception in Vision . Healey. http://www.csc.ncsu.edu/faculty/healey/PP B B • Visual Thinking for Design. Ware. Morgan Kaufmann, 2008. A A B A length position along position along • Information Visualization: Perception for Design, 3rd edition. Ware. Morgan Framed unaligned aligned scale Kaufmann /Academic Press, 2004. common scale 17 18 19 20 after [Graphical Perception: Theory, Experimentation, and Application to the Development of Graphical Methods. Cleveland and McGill. Journ. American Statistical Association 79:387 (1984), 531–554.] http://persci.mit.edu/gallery/checkershadow http://www.purveslab.net/seeforyourself/ Idiom design choices: Encode Categorical vs ordered color Decomposing color Encode • first rule of color: do not talk about color! –color is confusing if treated as monolithic Arrange Map from categorical and ordered Express Separate attributes • decompose into three channels Color What? Hue Saturation Luminance –ordered can show magnitude Luminance v Color Theory Order Align • luminance: how bright Why? Saturation Size, Angle, Curvature, ... • saturation: how colorful How? –categorical can show identity Hue Use • hue: what color Shape • channels have different properties Motion Direction, Rate, Frequency, ... [Seriously Colorful: Advanced Color Principles & Practices. –what they convey directly to perceptual system Stone.Tableau Customer Conference 2014.] –how much they can convey: how many discriminable bins can we use? 21 22 23 24 Spectral sensitivity Luminance Opponent color and color deficiency Color spaces Corners of the RGB • need luminance for edge detection • perceptual processing before optic nerve • CIE L*a*b*: good for computation color cube – L* intuitive: perceptually linear luminance –fine-grained detail only visible through –one achromatic luminance channel (L*) L from HLS All the same – a*b* axes: perceptually linear but nonintuitive luminance contrast –edge detection through luminance contrast • RGB: good for display hardware Luminance values –legible text requires luminance contrast! –2 chroma channels – poor for encoding –red-green (a*) & yellow-blue axis (b*) L* values • HSL/HSV: somewhat better for encoding • intrinsic perceptual ordering • “color blind”: one axis has degraded acuity – hue/saturation wheel intuitive Luminance information Chroma information Luminance information Chroma information –8% of men are red/green color deficient – beware: only pseudo-perceptual! – lightness (L) or value (V) ≠ luminance or L* –blue/yellow is rare • Luminance, hue, saturation – good for encoding Wavelength (nm) – but not standard graphics/tools colorspace IR UV [Seriously Colorful: Advanced Color Principles & Practices. Stone.Tableau Customer Conference 2014.] [Seriously Colorful: Advanced Color Principles & Practices. Visible Spectrum Stone.Tableau Customer Conference 2014.] 25 26 27 [https://en.wikipedia.org/wiki/HSL_and_HSV] 28 Designing for color deficiency: Check with simulator Designing for color deficiency: Avoid encoding by hue alone Color deficiency: Reduces color to 2 dimensions Designing for color deficiency: Blue-Orange is safe • redundantly encode – vary luminance – change shape Normal Protanope Normal Deuteranope Protanope Tritanope Deuteranope simulation vision http://rehue.net Change the shape Vary luminance [Seriously Colorful: Advanced Color Principles & Practices. Deuteranope Tritanope Stone.Tableau Customer Conference 2014.] [Seriously Colorful: Advanced Color Principles & Practices. Stone.Tableau Customer Conference 2014.] [Seriously Colorful: Advanced Color Principles & Practices. Stone.Tableau Customer Conference 2014.] 29 [Seriously Colorful: Advanced Color Principles & Practices. Stone.Tableau Customer Conference 2014.] 30 31 32