Visualization Analysis & Design

Transcription

1 Visualization Analysis & Design Tamara Munzner Department of Computer Science University of British Columbia UBC STAT 545A Guest Lecture October , Vancouver BC

2 Why have a human in the loop? Computer-based visualization systems provide visual representations of datasets designed to help people carry out tasks more effectively. Visualization is suitable when there is a need to augment human capabilities rather than replace people with computational decision-making methods. don t need vis when fully automatic solution exists and is trusted many analysis problems ill-specified don t know exactly what questions to ask in advance possibilities long-term use for end users (e.g. exploratory analysis of scientific data) presentation of known results stepping stone to better understanding of requirements before developing models help developers of automatic solution refine/debug, determine parameters help end users of automatic solutions verify, build trust 2

3 Why use an external representation? Computer-based visualization systems provide visual representations of datasets designed to help people carry out tasks more effectively. external representation: replace cognition with perception [Cerebral: Visualizing Multiple Experimental Conditions on a Graph with Biological Context. Barsky, Munzner, Gardy, and Kincaid. IEEE TVCG (Proc. InfoVis) 14(6): , 2008.] 3

4 Why represent all the data? Computer-based visualization systems provide visual representations of datasets designed to help people carry out tasks more effectively. summaries lose information, details matter confirm expected and find unexpected patterns assess validity of statistical model Anscombe s Quartet Identical statistics x mean 9 x variance 10 y mean 7.5 y variance 3.75 x/y correlation

5 Why analyze? imposes structure on huge design space SpaceTree TreeJuxtaposer What? scaffold to help you think systematically about choices analyzing existing as stepping stone to designing new most possibilities ineffective for particular task/data combination Why? How? [SpaceTree: Supporting Exploration in Large Node Link Tree, Design Evolution and Empirical Evaluation. Grosjean, Plaisant, and Bederson. Proc. InfoVis 2002, p ] [TreeJuxtaposer: Scalable Tree Comparison Using Focus +Context With Guaranteed Visibility. ACM Trans. on Graphics (Proc. SIGGRAPH) 22: , 2003.] Tree Actions Present Locate Identify SpaceTree Encode Navigate Select Filter Aggregate Targets Path between two nodes TreeJuxtaposer Encode Navigate Select Arrange 5

6 Analysis framework: Four levels, three questions domain situation who are the target users? abstraction translate from specifics of domain to vocabulary of vis what is shown? data abstraction often don t just draw what you re given: transform to new form why is the user looking at it? task abstraction idiom how is it shown? visual encoding idiom: how to draw interaction idiom: how to manipulate algorithm domain abstraction algorithm efficient computation 6 idiom [A Nested Model of Visualization Design and Validation. Munzner. IEEE TVCG 15(6): , 2009 (Proc. InfoVis 2009). ] domain abstraction idiom algorithm [A Multi-Level Typology of Abstract Visualization Tasks Brehmer and Munzner. IEEE TVCG 19(12): , 2013 (Proc. InfoVis 2013). ]

7 Why is validation difficult? different ways to get it wrong at each level Domain situation You misunderstood their needs Data/task abstraction You re showing them the wrong thing Visual encoding/interaction idiom The way you show it doesn t work Algorithm Your code is too slow 7

8 Why is validation difficult? solution: use methods from different fields at each level anthropology/ ethnography Domain situation Observe target users using existing tools Data/task abstraction problem-driven work design Visual encoding/interaction idiom Justify design with respect to alternatives computer science cognitive psychology anthropology/ ethnography Algorithm Measure system time/memory Analyze computational complexity Analyze results qualitatively Measure human time with lab experiment (lab study) Observe target users after deployment ( ) Measure adoption technique-driven work [A Nested Model of Visualization Design and Validation. Munzner. IEEE TVCG 15(6): , 2009 (Proc. InfoVis 2009). ] 8

9 Datasets What? Attributes What? Data Types Items Attributes Links Positions Grids Attribute Types Categorical Why? Data and Dataset Types Tables Networks & Trees Fields Geometry Clusters, Sets, Lists Ordered Ordinal Items Items (nodes) Grids Items Items How? Attributes Links Attributes Positions Attributes Positions Quantitative Dataset Types Tables Attributes (columns) Networks Fields (Continuous) Grid of positions Ordering Direction Sequential Items (rows) Cell containing value Link Node (item) Cell Attributes (columns) Diverging Multidimensional Table Trees Value in cell Cyclic Value in cell Geometry (Spatial) Dataset Availability Static Dynamic Position 9

10 Three major datatypes Dataset Types Tables Networks Spatial Attributes (columns) Fields (Continuous) Geometry (Spatial) Items (rows) Cell containing value Link Node (item) Cell Grid of positions Position Multidimensional Table Trees Attributes (columns) Value in cell Value in cell visualization vs computer graphics geometry is design decision 10

11 Types: Datasets and data Dataset Types Tables Networks Spatial Attributes (columns) Fields (Continuous) Geometry (Spatial) Items (rows) Cell containing value Link Node (item) Cell Grid of positions Position Attributes (columns) Attribute Types Categorical Ordered Ordinal Value in cell Quantitative 11

12 Actions Why? Targets Analyze All Data Consume Trends Outliers Features Discover Present Enjoy Produce Annotate Record Derive tag Attributes One Many Distribution Dependency Correlation Similarity Search Extremes {action, target} pairs discover distribution compare trends Location known Location unknown Target known Lookup Locate Target unknown Browse Explore Network Data Topology locate outliers browse topology Query Identify Compare Summarize Paths What? Spatial Data Shape Why? How? 12

13 Actions: Analyze, Query analyze consume discover vs present aka explore vs explain enjoy aka casual, social produce query annotate, record, derive how much data matters? one, some, all Analyze Consume Discover Present Enjoy Produce Annotate Record Derive tag Query Identify Compare Summarize independent choices analyze, query, (search) 13

14 Derive don t just draw what you re given! decide what the right thing to show is create it with a series of transformations from the original dataset draw that one of the four major strategies for handling complexity exports imports trade balance trade balance = exports imports Original Data Derived Data 14

15 Analysis example: Derive one attribute Strahler number centrality metric for trees/networks derived quantitative attribute draw top 5K of 500K for good skeleton [Using Strahler numbers for real time visual exploration of huge graphs. Auber. Proc. Intl. Conf. Computer Vision and Graphics, pp , 2002.] Task 1 Task 2 In Tree Out Quantitative attribute on nodes In Tree In Quantitative attribute on nodes Out Filtered Tree Removed unimportant parts What? Why? What? Why? How? In Tree Out Quantitative attribute on nodes Derive In Tree In Quantitative attribute on nodes Out Filtered Tree Summarize Topology Reduce Filter 15

16 Targets All Data Network Data Trends Outliers Features Topology Attributes Paths One Many Distribution Dependency Correlation Similarity Spatial Data Shape Extremes 16

17 Encode How? Encode Manipulate Facet Reduce Manipulate Facet Reduce Arrange Express Separate Map from categorical and ordered attributes Change Juxtapose Filter Order Align Color Hue Saturation Luminance Select Partition Aggregate Size, Angle, Curvature,... Use Navigate Superimpose Embed Shape Motion Direction, Rate, Frequency,... 17

18 How to encode: Arrange space, map channels Encode Arrange Express Order Separate Align Map from categorical and ordered attributes Color Hue Saturation Luminance Size, Angle, Curvature,... Use Shape Motion Direction, Rate, Frequency,... 18

19 Encoding visually analyze idiom structure 19

20 Definitions: Marks and channels marks geometric primitives Points Lines Areas Position Color channels Horizontal Vertical Both control appearance of marks Shape Tilt Size Length Area Volume 20

21 Encoding visually with marks and channels analyze idiom structure as combination of marks and channels 1: vertical position 2: vertical position horizontal position 3: vertical position horizontal position color hue 4: vertical position horizontal position color hue size (area) mark: line mark: point mark: point mark: point 21

22 Channels: Expressiveness types and effectiveness rankings Magnitude Channels: Ordered Attributes Position on common scale Position on unaligned scale Length (1D size) Tilt/angle Identity Channels: Categorical Attributes Spatial region Color hue Motion Shape Area (2D size) Depth (3D position) Color luminance Color saturation Curvature Volume (3D size) 22

23 Channels: Matching Types Magnitude Channels: Ordered Attributes Position on common scale Identity Channels: Categorical Attributes Spatial region Position on unaligned scale Color hue Length (1D size) Motion Tilt/angle Shape Area (2D size) Depth (3D position) Color luminance expressiveness principle match channel and data characteristics Color saturation Curvature Volume (3D size) 23

24 Channels: Rankings Magnitude Channels: Ordered Attributes Position on common scale Identity Channels: Categorical Attributes Spatial region Position on unaligned scale Color hue Length (1D size) Motion Tilt/angle Shape Area (2D size) Depth (3D position) Color luminance Color saturation Curvature expressiveness principle match channel and data characteristics effectiveness principle encode most important attributes with highest ranked channels Volume (3D size) 24

25 Accuracy: Fundamental Theory 25

26 Accuracy: Vis experiments Cleveland & McGill s Results Positions Log Error Crowdsourced Results Angles Circular areas Rectangular areas (aligned or in a treemap) after Michael McGuffin course slides, Log Error [Crowdsourcing Graphical Perception: Using Mechanical Turk to Assess Visualization Design. Heer and Bostock. Proc ACM Conf. Human Factors in Computing Systems (CHI) 2010, p ] 26

27 Separability vs. Integrality Position Hue (Color) Size Hue (Color) Width Height Red Green Fully separable Some interference Some/significant interference 2 groups each 2 groups each 3 groups total: integral area Major interference 4 groups total: integral hue 27

28 Grouping Marks as Links Containment Connection containment connection Identity Channels: Categorical Attributes proximity same spatial region similarity same values as other categorical channels Spatial region Color hue Motion Shape 28

29 How to encode: Arrange position and region Encode Arrange Express Separate Map from categorical and ordered attributes What? Why? Order Align Color Hue Saturation Luminance How? Size, Angle, Curvature,... Use Shape Motion Direction, Rate, Frequency,... 29

30 Arrange tables Express Values Axis Orientation Rectilinear Parallel Radial Separate, Order, Align Regions Separate Order Align Layout Density Dense Space-Filling 1 Key 2 Keys 3 Keys Many Keys List Matrix Volume Recursive Subdivision 30

31 Idioms: dot chart, line chart one key, one value data 2 quant attribs mark: points dot plot: + line connection marks between them channels aligned lengths to express quant value separated and ordered by key attrib into horizontal regions task find trend connection marks emphasize ordering of items along key axis by explicitly showing relationship 1 Key 2 Keys List Matrix Many Keys Recursive Subdivision between one item and the next Year Year

32 Choosing bar vs line charts depends on type of key attrib bar charts if categorical line charts if ordered Female Male Female Male do not use line charts for categorical key attribs violates expressiveness principle implication of trend so strong that it overrides semantics! The more male a person is, the taller he/she is year-olds 12-year-olds year-olds 12-year-olds after [Bars and Lines: A Study of Graphic Communication. Zacks and Tversky. Memory and Cognition 27:6 (1999), ] 32

33 Idiom: glyphmaps rectilinear good for linear vs nonlinear trends radial good for cyclic patterns Axis Orientation Rectilinear Parallel Radial : I con plots for 1 2 iconic time series shapes ( linear [Glyph-maps increasing, for Visually decreasing, Exploring 3Temporal shifted, Patterns in Climate single Data peak, and Models. single Wickham, dip, Hofmann, Wickham, and Cook. Environmetrics 23:5 (2012), ] ned linear and nonlinear, seasonal trends with different scales, and a combined linear and seasonal trend) in ean coordinates, time series icons ( left) and polar coordinates, star plots ( right). 33

34 Idiom: heatmap two keys, one value data 2 categ attribs (gene, experimental condition) 1 quant attrib (expression levels) marks: area separate and align in 2D matrix indexed by 2 categorical attributes channels color by quant attrib (ordered diverging colormap) task find clusters, outliers scalability 1 Key 2 Keys List Matrix Many Keys Recursive Subdivision 1M items, 100s of categ levels, ~10 quant attrib levels 34

35 Arrange spatial data Use Given Geometry Geographic Other Derived Spatial Fields Scalar Fields (one value per cell) Isocontours Direct Volume Rendering Vector and Tensor Fields (many values per cell) Flow Glyphs (local) Geometric (sparse seeds) Textures (dense seeds) Features (globally derived) 35

36 Idiom: choropleth map use given spatial data when central task is understanding spatial relationships data geographic geometry table with 1 quant attribute per region encoding use given geometry for area mark boundaries sequential segmented colormap 36

37 Population maps trickiness beware! [ ] 37

38 Idiom: topographic map data geographic geometry scalar spatial field 1 quant attribute per grid cell derived data isoline geometry isocontours computed for specific levels of scalar values Land Information New Zealand Data Service 38

39 Idioms: isosurfaces, direct volume rendering data scalar spatial field task 1 quant attribute per grid cell shape understanding, spatial relationships isosurface derived data: isocontours computed for specific levels of scalar values direct volume rendering [Interactive Volume Rendering Techniques. Kniss. Master s thesis, University of Utah Computer Science, 2002.] D F C transfer function maps scalar values to color, opacity [Multidimensional Transfer Functions for Volume Rendering. Kniss, Kindlmann, and Hansen. In The Visualization Handbook, edited by Charles Hansen and Christopher Johnson, pp Elsevier, 2005.] B E 39

40 Idiom: similarity-clustered streamlines data 3D vector field derived data (from field) streamlines: trajectory particle will follow derived data (per streamline) curvature, torsion, tortuosity signature: complex weighted combination compute cluster hierarchy across all signatures encode: color and opacity by cluster tasks find features, query shape scalability [Similarity Measures for Enhancing Interactive Streamline Seeding. McLoughlin,. Jones, Laramee, Malki, Masters, and. Hansen. IEEE Trans. Visualization and Computer Graphics 19:8 (2013), ] millions of samples, hundreds of streamlines 40

41 Arrange networks and trees Node Link Diagrams Connection Marks NETWORKS TREES Adjacency Matrix Derived Table NETWORKS TREES Enclosure Containment Marks NETWORKS TREES 41

42 Idiom: force-directed placement visual encoding link connection marks, node point marks considerations spatial position: no meaning directly encoded left free to minimize crossings proximity semantics? tasks sometimes meaningful sometimes arbitrary, artifact of layout algorithm tension with length long edges more visually salient than short explore topology; locate paths, clusters scalability node/edge density E < 4N 42

43 Idiom: adjacency matrix view data: network transform into same data/encoding as heatmap derived data: table from network 1 quant attrib weighted edge between nodes 2 categ attribs: node list x 2 [NodeTrix: a Hybrid Visualization of Social Networks. Henry, Fekete, and McGuffin. IEEE TVCG (Proc. InfoVis) 13(6): , 2007.] visual encoding cell shows presence/absence of edge scalability 1K nodes, 1M edges [Points of view: Networks. Gehlenborg and Wong. Nature Methods 9:115.] 43

44 Connection vs. adjacency comparison adjacency matrix strengths predictability, scalability, supports reordering some topology tasks trainable node-link diagram strengths topology understanding, path tracing intuitive, no training needed empirical study node-link best for small networks matrix best for large networks if tasks don t involve topological structure! [On the readability of graphs using node-link and matrix-based representations: a controlled experiment and statistical analysis. Ghoniem, Fekete, and Castagliola. Information Visualization 4:2 (2005), ] 44

45 Idiom: radial node-link tree >$%8#/.#"0<"'/,( F%8$%($8B#$$<"'/,( F*)*&$B#$$<"'/,( G/8$<)%=B#$$<"'/,( D"8)"&B#$$<"'/,( D"%8/0<"'/,( 3("*=$8-#$"<"'/,( B#$$6"9<"'/,( data tree encoding link connection marks point node marks radial axis orientation tasks angular proximity: siblings distance from center: depth in tree understanding topology, following paths scalability 1K - 10K nodes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

46 Idiom: treemap data tree 1 quant attrib at leaf nodes encoding area containment marks for hierarchical structure rectilinear orientation size encodes quant attrib tasks query attribute at leaf nodes scalability 1M leaf nodes 46

47 Connection vs. containment comparison marks as links (vs. nodes) common case in network drawing 1D case: connection ex: all node-link diagrams emphasizes topology, path tracing networks and trees 2D case: containment ex: all treemap variants emphasizes attribute values at leaves (size coding) only trees Containment Node Link Diagram Connection Treemap [Elastic Hierarchies: Combining Treemaps and Node-Link Diagrams. Dong, McGuffin, and Chignell. Proc. InfoVis 2005, p ] 47

48 How to encode: Mapping color Encode Arrange Express Separate Map from categorical and ordered attributes What? Why? Order Align Color Hue Saturation Luminance How? Size, Angle, Curvature,... Use Shape Motion Direction, Rate, Frequency,... 48

49 Color: Luminance, saturation, hue 3 channels identity for categorical hue magnitude for ordered luminance saturation Luminance values Saturation Hue RGB: poor for encoding HSL: better, but beware lightness luminance Corners of the RGB color cube L from HLS All the same Luminance values 49

50 Categorical color: Discriminability constraints noncontiguous small regions of color: only 6-12 bins [Cinteny: flexible analysis and visualization of synteny and genome rearrangements in multiple organisms. Sinha and Meller. BMC Bioinformatics, 8:82, 2007.] 50

51 Ordered color: Rainbow is poor default problems perceptually unordered perceptually nonlinear benefits fine-grained structure visible and nameable [A Rule-based Tool for Assisting Colormap Selection. Bergman,. Rogowitz, and. Treinish. Proc. IEEE Visualization (Vis), pp , 1995.] [Why Should Engineers Be Worried About Color? Treinish and Rogowitz [Transfer Functions in Direct Volume Rendering: Design, Interface, Interaction. Kindlmann. SIGGRAPH 2002 Course Notes] 51

52 Ordered color: Rainbow is poor default problems perceptually unordered perceptually nonlinear benefits fine-grained structure visible and nameable alternatives [A Rule-based Tool for Assisting Colormap Selection. Bergman,. Rogowitz, and. Treinish. Proc. IEEE Visualization (Vis), pp , 1995.] large-scale structure: fewer hues [Why Should Engineers Be Worried About Color? Treinish and Rogowitz [Transfer Functions in Direct Volume Rendering: Design, Interface, Interaction. Kindlmann. SIGGRAPH 2002 Course Notes] 52

53 Ordered color: Rainbow is poor default problems perceptually unordered perceptually nonlinear benefits fine-grained structure visible and nameable alternatives [A Rule-based Tool for Assisting Colormap Selection. Bergman,. Rogowitz, and. Treinish. Proc. IEEE Visualization (Vis), pp , 1995.] large-scale structure: fewer hues fine structure: multiple hues with monotonically increasing luminance [eg viridis R/python] [Why Should Engineers Be Worried About Color? Treinish and Rogowitz [Transfer Functions in Direct Volume Rendering: Design, Interface, Interaction. Kindlmann. SIGGRAPH 2002 Course Notes] 53

54 Viridis colorful, perceptually uniform, colorblind-safe, monotonically increasing luminance viridis/vignettes/intro-to-viridis.html 54

55 Ordered color: Rainbow is poor default problems perceptually unordered perceptually nonlinear benefits fine-grained structure visible and nameable alternatives [A Rule-based Tool for Assisting Colormap Selection. Bergman,. Rogowitz, and. Treinish. Proc. IEEE Visualization (Vis), pp , 1995.] large-scale structure: fewer hues fine structure: multiple hues with monotonically increasing luminance [eg viridis R/python] segmented rainbows for binned or categorical [Why Should Engineers Be Worried About Color? Treinish and Rogowitz [Transfer Functions in Direct Volume Rendering: Design, Interface, Interaction. Kindlmann. SIGGRAPH 2002 Course Notes] 55

56 Encode How? Encode Manipulate Facet Reduce Manipulate Facet Reduce Arrange Express Separate Map from categorical and ordered attributes Change Juxtapose Filter Order Align Color Hue Saturation Luminance Select Partition Aggregate Size, Angle, Curvature,... Use Navigate Superimpose Embed Shape Motion Direction, Rate, Frequency,... 56

57 How to handle complexity: 3 more strategies + 1 previous Manipulate Facet Reduce Derive Change Juxtapose Filter Select Partition Aggregate change view over time facet across multiple views Navigate Superimpose Embed reduce items/attributes within single view derive new data to show within view 57

58 Datasets What? Attributes domain Data Types Items Attributes Links Positions Grids Actions Why? Attribute Types Categorical Targets abstraction Data and Dataset Types Tables Items Attributes Dataset Types Tables Items (rows) Attributes (columns) Networks & Trees Items (nodes) Links Attributes Cell containing value Analyze Search Multidimensional Table Value in cell Consume Discover Produce Fields Geometry Clusters, Sets, Lists Grids Positions Attributes Annotate Networks Record Encode Derive Location known Location unknown tag Trees Present Link Node (item) Items Arrange Order Target known Positions Express Lookup Use Locate Fields (Continuous) Cell Enjoy Items Grid of positions Separate Attributes (columns) Value Align in cell Target unknown Browse Explore All Data Ordered Ordinal Trends Quantitative Attributes Outliers Features Ordering Direction One Many Manipulate Facet Reduce Sequential Distribution Dependency Correlation Similarity Map Change Juxtapose Filter from categorical and ordered attributes Diverging Extremes Color Hue Saturation Luminance Cyclic Select Partition Aggregate Network Data Topology How? Encode Manipulate Facet Reduce Size, Angle, Curvature,... Shape Navigate Superimpose idiom algorithm Embed Geometry (Spatial) Position Query Identify Compare Summarize Motion Paths Direction, Rate, Frequency,... Spatial Data What? 58

59 More this talk book page (including tutorial lecture slides) 20% promo code for book+ebook combo: HVN17 illustrations: Eamonn Maguire papers, videos, software, talks, full courses grad vis course Jan 17: CPSC 547, Tue/Thu 3:30 - students from outside CS are welcome Visualization Analysis and Design. Munzner. A K Peters Visualization Series, CRC Press, Visualization Series,