Comp/Phys/Mtsc 715 Information Visualization and Tufte 4/5/20112 - - PDF document

▶
comp phys mtsc 715
SMART_READER_LITE
LIVE PREVIEW

Comp/Phys/Mtsc 715 Information Visualization and Tufte 4/5/20112 - - PDF document

Jun 25, 2023 134 likes •419 views

4/5/2012 Comp/Phys/Mtsc 715 Information Visualization and Tufte 4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor 1 Preview Videos Vis 2004: robbins.mpg Comparing two 2D time-varying neural responses Vis 2004: theisel.avi

slide-1
SLIDE 1

4/5/2012 1

Information Visualization and Tufte

Comp/Phys/Mtsc 715

4/5/20112 Infovis, Genomics Tufte 1 Comp/Phys/Mtsc 715 Taylor

Preview Videos

  • Vis 2004: robbins.mpg

– Comparing two 2D time-varying neural responses

  • Vis 2004: theisel.avi

– Flow topology for time-varying 2D flow fields

4/5/20112 Infovis, Genomics Tufte 2 Comp/Phys/Mtsc 715 Taylor

Information Display and Spatial Embeddings

  • The data we have looked at so far has existed in some natural

spatial embedding

  • Subset of the possible kinds of information that exist
  • Key challenges in information visualization

– How to embed in 2D or 3D? – How do you reduce high-dimensional data? – How do you convey relationships among entities?

  • Choice of embedding often depends on goal
  • Often, spatial embedding changes

– Level-of-detail – Focus + context techniques magnify portions of data

4/5/20112 Infovis, Genomics Tufte 3 Comp/Phys/Mtsc 715 Taylor

slide-2
SLIDE 2

4/5/2012 2

Focus + Context

  • May only want to show part of data in detail
  • Knowing how the detailed data relates to the

rest of the data set is still important

  • Focus + context techniques attempt to do this

– More screen space dedicated to showing one or more parts of the data – Rest of the data shown in less resolution

4/5/20112 Infovis, Genomics Tufte 4 Comp/Phys/Mtsc 715 Taylor

Interactive Data Analysis

  • Tie data analysis into visualization interface
  • Show results of queries/algorithms
  • Allows pattern detection in subsets of the data
  • Enable finding answers to different questions

at different points in data exploration

4/5/20112 Infovis, Genomics Tufte 5 Comp/Phys/Mtsc 715 Taylor

Techniques for small data sets…

4/5/20112 Infovis, Genomics Tufte 6 Comp/Phys/Mtsc 715 Taylor

slide-3
SLIDE 3

4/5/2012 3

Flow Maps

  • Phan et al., IEEE Vis 2005
  • Quantifies movement of items between

nodes in a graph

4/5/20112 Infovis, Genomics Tufte 7 Comp/Phys/Mtsc 715 Taylor

Cartograms 1

  • Heilmann et al., InfoVis 2004
  • Distort maps so area reflects population

4/5/20112 Infovis, Genomics Tufte 8 Comp/Phys/Mtsc 715 Taylor

Cartograms 2

  • Dorling, Barford,

Newman, IEEE TVCG, 12(5), 2006

  • Distorts geographic

map based on scalar data values

  • Net emigration
  • ver last 50 years

4/5/20112 Infovis, Genomics Tufte 9 Comp/Phys/Mtsc 715 Taylor

slide-4
SLIDE 4

4/5/2012 4

Tracking Avian Flu Outbreaks

  • Proulx et al., IEEE VAST, 2006

4/5/20112 Infovis, Genomics Tufte 10 Comp/Phys/Mtsc 715 Taylor

Tracking Avian Flu Outbreaks

  • Proulx et al., IEEE VAST, 2006

4/5/20112 Infovis, Genomics Tufte 11 Comp/Phys/Mtsc 715 Taylor

Multi-Dimensional Scaling

4/5/20112 Infovis, Genomics Tufte 12 Comp/Phys/Mtsc 715 Taylor

slide-5
SLIDE 5

4/5/2012 5

Multi-Dimensional Scaling

  • Family of algorithms to reduce high-dimensional data

down to 2D or 3D

– So we can visualize it – Reduce complexity of higher-dimensional relationships

  • Aims to preserve distances in higher dimensional

space after the dimension reduction

– Close things in high-dimension space are close in 2D or 3D – Far things remain far apart

4/5/20112 Infovis, Genomics Tufte 13 Comp/Phys/Mtsc 715 Taylor

H-BLOBs

  • Sprenger, Brunella, Gross; Vis 2000
  • Splatting in 3D

4/5/20112 Infovis, Genomics Tufte 14 Comp/Phys/Mtsc 715 Taylor

ThemeView

  • Card, SIGGRAPH ‘96

4/5/20112 Infovis, Genomics Tufte 15 Comp/Phys/Mtsc 715 Taylor

slide-6
SLIDE 6

4/5/2012 6

Tree Visualization

4/5/20112 Infovis, Genomics Tufte 16 Comp/Phys/Mtsc 715 Taylor

Tree Visualization

  • Trees model hierarchical information very well
  • Structure found in many areas

– Corporation management hierarchy – Governments – Taxonomy – Computer science

  • Many visualization techniques

4/5/20112 Infovis, Genomics Tufte 17 Comp/Phys/Mtsc 715 Taylor

Node-link Diagrams

Tree Graph

Nodes Link Leaf nodes

4/5/20112 Infovis, Genomics Tufte 18 Comp/Phys/Mtsc 715 Taylor

slide-7
SLIDE 7

4/5/2012 7

Visual Grammar of Node-Link Diagrams

4/5/20112 Infovis, Genomics Tufte 19 Comp/Phys/Mtsc 715 Taylor

ConeTrees

  • Robertson et al., Comm. ACM 36(4), pp. 57-71

4/5/20112 Infovis, Genomics Tufte 20 Comp/Phys/Mtsc 715 Taylor

Tree “Extruded” in 3D

  • Wattenberg and Fisher,

IEEE InfoVis 2003

  • Tree structure showing

salient features of text

  • Tree extruded into 3D

with varying levels of detail in text

4/5/20112 Infovis, Genomics Tufte 21 Comp/Phys/Mtsc 715 Taylor

slide-8
SLIDE 8

4/5/2012 8

Nature-Inspired Graph Layout

  • Carpendale and Agarawala, Info Vis 2004

4/5/20112 Infovis, Genomics Tufte 22 Comp/Phys/Mtsc 715 Taylor

Showing Hierarchical Clusters

  • Pretorius and van

Wijk, TVCG 12(5), 2006

4/5/20112 Infovis, Genomics Tufte 23 Comp/Phys/Mtsc 715 Taylor

Hyperbolic Trees

  • Lamping and Rao,

UIST 1994

  • Focus+context

scheme for viewing large hierarchies

  • Smooth animated

changes between different foci

4/5/20112 Infovis, Genomics Tufte 24 Comp/Phys/Mtsc 715 Taylor

slide-9
SLIDE 9

4/5/2012 9

TreeMaps

Herman et. al. IEEE Trans on Vis & CG, 6(1), 2000

4/5/20112 Infovis, Genomics Tufte 25 Comp/Phys/Mtsc 715 Taylor

Tree Maps

  • www.smartmoney.com

4/5/20112 Infovis, Genomics Tufte 26 Comp/Phys/Mtsc 715 Taylor

Cushion Maps

  • Van Wijk and van de Wetering, InfoVis 1999
  • Multiple levels of

lighting group subtrees

4/5/20112 Infovis, Genomics Tufte 27 Comp/Phys/Mtsc 715 Taylor

slide-10
SLIDE 10

4/5/2012 10

4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor 28

Voronoi Tree Maps

  • Balzer and Deussen, IEEE Vis 2005

4/5/20112 Infovis, Genomics Tufte 29 Comp/Phys/Mtsc 715 Taylor

GraphSplatting

  • van Liere and de Leeuw, IEEE TVCG 9(2), 2003
  • Add up 2D Gaussians centered at nodes

– Density field – All 2D scalar techniques available

4/5/20112 Infovis, Genomics Tufte 30 Comp/Phys/Mtsc 715 Taylor

Rainbow

slide-11
SLIDE 11

4/5/2012 11 Visual Bibliography of Tree Vis (1) by Hans-Jorg Schulz

  • http://www.informatik.uni-rostock.de/~hs162/treeposter/poster.html

4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor 31

Visual Bibliography of Tree Vis (2) by Hans-Jorg Schulz

  • http://www.informatik.uni-rostock.de/~hs162/treeposter/poster.html

4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor 32

Graph Visualization

4/5/20112 Infovis, Genomics Tufte 33 Comp/Phys/Mtsc 715 Taylor

slide-12
SLIDE 12

4/5/2012 12

Graph Visualization

  • Huge topic in it’s own right
  • Goals:

– Show all nodes clearly – Show all edges and avoid edge overlap

4/5/20112 Infovis, Genomics Tufte 34 Comp/Phys/Mtsc 715 Taylor

Airline Flight Network

  • Czechoslovakia

Air Transport Company flight network

  • Edward R. Tufte,

Envisioning Information

4/5/20112 Infovis, Genomics Tufte 35 Comp/Phys/Mtsc 715 Taylor

Facebook Friend Wheel

  • Fletcher 2008

4/5/20112 Infovis, Genomics Tufte 36 Comp/Phys/Mtsc 715 Taylor

slide-13
SLIDE 13

4/5/2012 13

Facebook Friend Wheel

  • Fletcher 2008

4/5/20112 Infovis, Genomics Tufte 37 Comp/Phys/Mtsc 715 Taylor

Facebook Friend Wheel

  • Fletcher 2008

Russ Taylor

4/5/20112 Infovis, Genomics Tufte 38 Comp/Phys/Mtsc 715 Taylor

Radial Focus+Context Graphs

  • Jankun-Kelly and Ma, IEEE InfoVis 2003
  • For showing visual nodes (images, web

pages)

  • Level highlighting

– More space allocated to level of interest

4/5/20112 Infovis, Genomics Tufte 39 Comp/Phys/Mtsc 715 Taylor

slide-14
SLIDE 14

4/5/2012 14

Social Network Identity Resolution

  • Bilgic et al., IEEE VAST, 2006

4/5/20112 Infovis, Genomics Tufte 40 Comp/Phys/Mtsc 715 Taylor

Social Networks in Time

  • Card et al., VAST 2006
  • Expand tree in regions
  • f interest
  • Highlight nodes

matching search criteria

  • Slider widget to

change time

  • Animation to show

changes in time

4/5/20112 Infovis, Genomics Tufte 41 Comp/Phys/Mtsc 715 Taylor

Social Networking: Vizster

  • Heer and Boyd, InfoVis 2005

4/5/20112 Infovis, Genomics Tufte 42 Comp/Phys/Mtsc 715 Taylor

slide-15
SLIDE 15

4/5/2012 15 Social Network Vis with Matrices and Graphs

  • Henry and

Fekete, IEEE TVCG 12(5), 2006

  • Reordered

matrices better for some tasks

  • Node-link

diagrams better for others

4/5/20112 Infovis, Genomics Tufte 43 Comp/Phys/Mtsc 715 Taylor

Semantic Relationships

  • Wong, et al., IEEE VAST 2006

4/5/20112 Infovis, Genomics Tufte 44 Comp/Phys/Mtsc 715 Taylor

Graph Edge Reduction

  • Perer and

Shneiderman, IEEE TVCG 12(5), 2006

  • Grouping by

country

  • Interconnection

among groups

4/5/20112 Infovis, Genomics Tufte 45 Comp/Phys/Mtsc 715 Taylor

slide-16
SLIDE 16

4/5/2012 16

Small World Graphs

  • van Ham and van Wijk, InfoVis 2004

4/5/20112 Infovis, Genomics Tufte 46 Comp/Phys/Mtsc 715 Taylor

Viewing Large Graphs

  • Abello, van Ham, Krishnan, IEEE TVCG 12(5), 2006

4/5/20112 Infovis, Genomics Tufte 47 Comp/Phys/Mtsc 715 Taylor

Semi-transparent Lines

  • Wong et al., InfoVis 2005

4/5/20112 Infovis, Genomics Tufte 48 Comp/Phys/Mtsc 715 Taylor

slide-17
SLIDE 17

4/5/2012 17 EdgeLens: Managing Edge Congestion in Graphs

  • Wong, Carpendale, Greenberg,

InfoVis 2003

4/5/20112 Infovis, Genomics Tufte 49 Comp/Phys/Mtsc 715 Taylor

Topological Fish-Eye Lens for Large Graph Visualization

  • Ganser, Koren, North, InfoVis 2004
  • Approximates graph structure with less

detail

  • Points of focus are shown in higher detail

4/5/20112 Infovis, Genomics Tufte 50 Comp/Phys/Mtsc 715 Taylor

Reducing Clutter in Graphs

  • Kumar and

Garland, IEEE TVCG, 12(5) 2006

  • Remove edges

within clusters

  • Replace edges in

cliques with star edge glyphs

4/5/20112 Infovis, Genomics Tufte 51 Comp/Phys/Mtsc 715 Taylor

slide-18
SLIDE 18

4/5/2012 18

Parallel Coordinates

4/5/20112 Infovis, Genomics Tufte 52 Comp/Phys/Mtsc 715 Taylor

Parallel Coordinates

  • http://davis.wpi.edu/~xmdv/vis_parcoord.html

4/5/20112 Infovis, Genomics Tufte 53 Comp/Phys/Mtsc 715 Taylor

Parallel Coordinates with Color Map

  • Xie et al., IEEE VAST 2006

4/5/20112 Infovis, Genomics Tufte 54 Comp/Phys/Mtsc 715 Taylor

slide-19
SLIDE 19

4/5/2012 19 Parallel Coordinates with Transfer Functions

  • Johansson et al., InfoVis 2005

4/5/20112 Infovis, Genomics Tufte 55 Comp/Phys/Mtsc 715 Taylor

Parallel Coordinates over Time

  • Brodbeck and

Girardin, IEEE InfoVis 2003

  • Fading out

(lowering saturation) shows how data change through time

4/5/20112 Infovis, Genomics Tufte 56 Comp/Phys/Mtsc 715 Taylor

Showing Clusters in Parallel Coordinates

  • Artero et al., IEEE Info Vis,

2004

4/5/20112 Infovis, Genomics Tufte 57 Comp/Phys/Mtsc 715 Taylor

slide-20
SLIDE 20

4/5/2012 20 Focus+Context in Parallel Coordinates

  • Brodbeck and Girardin, IEEE InfoVis 2003
  • Increase horizontal resolution in region of

focus

4/5/20112 Infovis, Genomics Tufte 58 Comp/Phys/Mtsc 715 Taylor

Curvy PC

  • Graham and

Kennedy IV’03

  • Easier to follow

(good continuity)

  • Cubic splines
  • Separates

intersections

4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor

Curvy PC (2)

  • Graham and

Kennedy IV’03

  • Easier to follow

(good continuity)

  • Cubic splines
  • Separates

intersections

4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor

slide-21
SLIDE 21

4/5/2012 21

Curvy PC (3)

  • Moustafa and

Wegman, Graphics of large data sets

  • Zero-derivative, not

just smooth

  • Smooth, but

maintain clustering

not actually their image, but same idea

4/5/20112 Infovis, Genomics Tufte 61 Comp/Phys/Mtsc 715 Taylor

Whisker and Star Plots

  • Each angle encodes a different property of the

data

  • Star plots connect the lines

Whisker Plot Star Plot

4/5/20112 Infovis, Genomics Tufte 62 Comp/Phys/Mtsc 715 Taylor

Star Glyphs + Parallel Coordinates

  • Fanea, Carpendale, Isenberg, InfoVis 2005
  • Star glyphs show how population varies in one

dimension

Where are the clusters?

4/5/20112 Infovis, Genomics Tufte 63 Comp/Phys/Mtsc 715 Taylor

slide-22
SLIDE 22

4/5/2012 22

Miscellaneous

4/5/20112 Infovis, Genomics Tufte 64 Comp/Phys/Mtsc 715 Taylor

Situational Awareness Visualization

  • Livnat et al., InfoVis

2005

  • What, when, where?
  • What - around circle
  • When - which ring
  • Where - indicated by

spatial embedding inside rings

4/5/20112 Infovis, Genomics Tufte 65 Comp/Phys/Mtsc 715 Taylor

Showing Time-Varying Data With Flocking Boids

  • Moere, InfoVis 2004
  • Particle motion

controlled by data values

  • Attraction/repulsion

based on similarity

Unpredictable spatial embedding

4/5/20112 Infovis, Genomics Tufte 66 Comp/Phys/Mtsc 715 Taylor

slide-23
SLIDE 23

4/5/2012 23

Theme River

  • Havre, Hetzler, Nowell; InfoVis 2000

4/5/20112 Infovis, Genomics Tufte 67 Comp/Phys/Mtsc 715 Taylor

NameVoyager

  • Wattenberg,

InfoVis 2005

  • Shows baby name

popularity over time in a stacked graph

  • Thickness of strip

indicates frequency of name

4/5/20112 Infovis, Genomics Tufte 68 Comp/Phys/Mtsc 715 Taylor

Tufte

4/5/20112 Infovis, Genomics Tufte 69 Comp/Phys/Mtsc 715 Taylor

slide-24
SLIDE 24

4/5/2012 24

Edward R. Tufte

  • Three books on information display

– The Visual Display of Quantitative Information – Envisioning Information – Visual Explanations

  • Beautiful and terrible examples
  • Rules of thumb
  • Design guidelines

4/5/20112 Infovis, Genomics Tufte 70 Comp/Phys/Mtsc 715 Taylor

Tufte: Visual Display

  • Train Schedule: Good visual organization

4/5/20112 Infovis, Genomics Tufte 71 Comp/Phys/Mtsc 715 Taylor

Tufte: Visual Display

  • Multi-layered Graph: Show all the data

4/5/20112 Infovis, Genomics Tufte 72 Comp/Phys/Mtsc 715 Taylor

slide-25
SLIDE 25

4/5/2012 25

Tufte: Visual Explanations

  • Storm Visualization Redesigned: Chartjunk

4/5/20112 Infovis, Genomics Tufte 73 Comp/Phys/Mtsc 715 Taylor

Aside: How strong a grid?

  • Bartram, Cheung, Stone; TVCG 2011

– Tried a variety of densities – Asked users “set too high” – Asked users “set too low” – 20% opaque within good range

4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor 74

Tufte: Visual Explanations

  • Medical Illustration Redesigned: Chartjunk

– Finer call-out lines, label directly on diagram

4/5/20112 Infovis, Genomics Tufte 75 Comp/Phys/Mtsc 715 Taylor

slide-26
SLIDE 26

4/5/2012 26

Tufte: Visual Explanations

  • O-Ring Damage Redesigned

– Order by the important variable

4/5/20112 Infovis, Genomics Tufte 76 Comp/Phys/Mtsc 715 Taylor

Tufte: Visual Explanations

  • O-Ring Damage Redesigned

– Order by the important variable

4/5/20112 Infovis, Genomics Tufte 77 Comp/Phys/Mtsc 715 Taylor 4/5/20112 Infovis, Genomics Tufte Comp/Phys/Mtsc 715 Taylor

slide-27
SLIDE 27

4/5/2012 27

Credits

  • Node-link diagram discussion: Information

Visualization by Colin Ware

4/5/20112 Infovis, Genomics Tufte 79 Comp/Phys/Mtsc 715 Taylor