IAT 355 Visual Analytics

Space: View Transformations Lyn Bartram So much data, so little space: 1

• Rich data (many dimensions) • Huge amounts of data

• Overplotting [Few] • patterns and relations across sets • Visual fragmentation • Decoding too many different visual forms

IAT 355 | View transformations 2 Recall: Dimensional division

• “splitting” dimensions across multiple linked views

• Small multiples • Trellis displays • Scatterplot matrices

IAT 355 | View transformations 3 Small multiples

• use the same basic graphic or chart to display difference slices of a data set • rich, multi-dimensional data without trying to cram all that information into a single, overly-complex chart. • Singular design reduces decoding effort.

• E. Tufte “The Visual Display of Quantitative Information,” p. 42 and “Envisioning Information,” p. 29

IAT 355 | View transformations 4 Small multiples

IAT 355 | View transformations 5 Small multiples

• Small multiples to convey n-dims.

IAT 355 | View transformations 6 Horizon graphs

IAT 355 | View transformations 7 Trellis plots

IAT 355 | View transformations 8 Scatter plot matrices

IAT 355 | View transformations 9 Multiple Views

• “Guidelines for Using Multiple Views in Information Visualization” • Baldonado, Woodruff and Kichinsky AVI 00

IAT 355 | View transformations 10 Multiple Views: 8 Guidelines

• Rule of Diversity: • Use multiple views when there is a diversity of attributes • Rule of Complementarity: • Multiple views should bring out correlations and/or disparities • Rule of Decomposition: “Divide and conquer”. • Help users visualize relevant chunks of complex data • Rule of Parsimony: • Use multiple views minimally

IAT 355 | View transformations 11 8 Guidelines Cont’d

• Rule of Space/Time Resource • Optimization: Balance spatial and temporal benefits of presenting and using the views • Rule of Self Evidence: • Use cues to make relationships apparent. • Rule of Consistency: • Keep views and state of multiple views consistent • Rule of attention management: • Use perceptual techniques to focus user attention

IAT 355 | View transformations 12 View transformations

• Location probes • Details on demand • Brushing • Magic lenses

• Viewpoint controls • Zoom, pan, clip • Overview+detail

• Distortions (Focus+Context) • Bifocal : Perspective Wall, Document Lens • Polyfocal : Table Lens • Based on levels of interest: Fish Eye Lens

IAT 355 | View transformations 13 So much data, so little space: 2

Location probes • Details on demand • Brushing • Magic lenses

IAT 355 | View transformations 14 Details on Demand

Popup windows Adjacent frames So much data, so little space: 2

Location probes • Details on demand • Brushing • Magic lenses

IAT 355 | View transformations 16 So much data, so little space: 2

Location probes • Details on demand • Brushing • Magic lenses

IAT 355 | View transformations 17 Magic lenses

Magic lenses are probes that give an alternate view of a region in the Visual Structure in which the objects reveal additional properties

IAT 355 | View transformations 18 Magic lens / tool glass

Figure 4.29 Illustrating the concept of a magic lens. (a) shows a conventional map of an area, (b) shows the location of services (gas, water and electricity pipes) in the same area, and (c) a (movable) magic lens shows services in an area of interest, in context Figure 4.30 A molecular surface of the protein transferase coloured by electrostatic potential bound to DNA shown as a schematic. (ID = 10mh). The magic lens window allows a view of the atomic structure bonding to be shown, with the bound ligand structure highlighted as cylinders, thereby providing a view inside the protein Source: By kind permission of Tom Oldfield and Michael Hartshorn So much data, so little space: 3

• Scale - Many data sets are too large to visualize on one screen • May simply be too many cases • May be too many variables • May only be able to highlight particular cases or particular variables, but viewer’s focus may change from time to time

IAT 355 | View transformations 21 Important Concepts

• Portals • Lenses • Sticky objects • Semantic zooming

IAT 355 | View transformations 22 Portals

• Views onto another place in the world • Implemented typically as separate rectangular region • Zooming, panning, I/O all work independently in there • Can be used to create overviews or focus regions

IAT 355 | View transformations 23 Lenses

• Rectangular regions/objects that can be moved around on display • Objects that alter the appearance and behavior of objects seen through them

IAT 355 | View transformations 24 Sticky Objects

• Objects in the world that do not respond to the basic zoom/pan interface physics • Objects are “stuck” to the display • They never change position • They never change size

IAT 355 | View transformations 25 Common Solution- Scroll/Pan

• Provide a larger, virtual screen by allowing user to move to different areas • Requires one or more of • Dedicated mouse button/wheel • Peripheral scroll bars – Takes screen space – Requires mouse move • Only get to see one piece

IAT 355 | View transformations 26 Panning and Zooming

• Panning • Smooth movement of camera across scene (or scene moves and camera stays still) • Zooming • Increasing or decreasing the magnification of the objects in a scene • Useful for changing focal point

IAT 355 | View transformations 27 Pan and Zoom

How to show a lot of information in a small space? • Multiple Levels of Resolution • The view changes depending on the “distance” from the viewer to the objects

• Distortion-based techniques • Keep a steady overview, make some objects larger while simultaneously shrinking others

IAT 355 | View transformations 28 Zooming

• Standard Zooming • Get close in to see information in more detail • Example: Google earth zooming in

• Intelligent Zooming • Show semantically relevant information out of proportion • Smart speed up and slow down • Example: speed-dependent zooming, Igarishi & Hinkley

• Semantic Zooming • Zooming can be conceptual as opposed to simply reducing pixels • Example tool: Pad++ and Piccolo projects • http://hcil.cs.umd.edu/video/1998/1998_pad.mpg

IAT 355 | View transformations 29 Speed-dependent Zooming by Igarashi & Hinkley 2000

http://www-ui.is.s.u-tokyo.ac.jp/~takeo/video/autozoom.mov http://www-ui.is.s.u-tokyo.ac.jp/~takeo/java/autozoom/ autozoom.htm

IAT 355 | View transformations 30 Pad++

• An infinite 2D plane • Can get infinitely close to the surface too • Navigate by panning and zooming • Pan: • move around on the plane • Zoom: • move closer to and farther from the plane • http://hcil.cs.umd.edu/video/1998/1998_pad.mpg

IAT 355 | View transformations 31 The Role of Portals

• All this panning and zooming can get confusing (maybe even dizzying) • Portals allow for zooming a small piece of the dataset while keeping everything else in the same position • Pad++ is one big stretchy sheet • A portal is more like a special window into a piece of the sheet • That window behaves independently of the rest

IAT 355 | View transformations 32 Space-Scale Diagrams (Furnas & Bederson 95)

• Original figure, shown at various scales • Horizontal axis is standard, vertical is scale

IAT 355 | View transformations 33 Space-Scale Diagrams (Furnas & Bederson 95)

• User has a fixed-sized viewing window • Moving it through 3D space yields all possible sequences of pan & zoom

IAT 355 | View transformations 34 Space-Scale Diagrams (Furnas & Bederson 95)

• If you move the origin of the 2D plane, the properties of the original 2D picture do not change • Therefore, the absolute angles between the rays should not be assigned any meaning

IAT 355 | View transformations 35 Space-Scale Diagrams (Furnas & Bederson 95)

• We can think of this in terms of 1D • When zoomed out, you can see wider set of points

IAT 355 | View transformations 36 Space-Scale Diagrams (Furnas & Bederson 95)

What about panning and zooming at the same time? • Panning is linear • Zooming is logarithmic • The two effects interact • If you compute the two separately and run them in parallel you get problems • When zooming in, things go exponentially fast • Panning can’t keep up – The target “runs away” out of view

IAT 355 | View transformations 37 How to Pan While Zooming?

IAT 355 | View transformations 38 How to Pan While Zooming?

IAT 355 | View transformations 39 Navigation in Pad++

• How to keep from getting lost? • Animate the traversal from one object to another using “hyperlinks” • If the target is more than one screen away, zoom out, pan over, and zoom back in • Goal: help viewer maintain context

IAT 355 | View transformations 40 Semantic Zooming

• Zooming that is not simply a change in size or scale like simple magnification • Objects change fundamental appearance/presence at different zoom levels • Zooming is like step function with boundaries where a semantic transition takes place

IAT 355 | View transformations 41 Standard vs. Semantic Zooming

• Geometric (standard) zooming: • The view depends on the physical properties of what is being viewed

• Semantic Zooming: • When zooming away, instead of seeing a scaled-down version of an object, see a different representation • The representation shown depends on the meaning to be imparted.

IAT 355 | View transformations 42 Concept of Semantic Zoom

• Infinitely scalable painting program • close in, see flecks of paint • farther away, see paint strokes • farther still, see the holistic impression of the painting • farther still, see the artist sitting at the easel

IAT 355 | View transformations 43 Examples of Semantic Zoom

• Information Maps • zoom into restaurant • see the interior • see what is served there • maybe zoom based on price instead! • see expensive restaurants first • keep zooming till you get to your price range

• Browsing an information service • Charge user successively higher rates for successively more detailed information

IAT 355 | View transformations 44 CZSaw

IAT 355 | View transformations 45 View Infinity

Model-based discrete semantic zoom

View Infinity: A Zoomable Interface for Feature-Oriented Software Development

IAT 355 | View transformations 46 Panning and Zooming

• Is it actually useful? • Is it better to show multiple simultaneous views? • Is it better to use distortion techniques?

• Would keeping a separate global overview help with navigation? • The research literature suggests yes, that overview +detail is usually better than pan & zoom.

IAT 355 | View transformations 47 Large Scale

• One of the fundamental challenges in information visualization • How to allow user to navigate through, and analyze a data set that is too large to fit in the display • Potential solutions lie in • Representation • Interaction • Eventually you will run out of pixels!

IAT 355 | View transformations 48 The “Overview” Concept

• Providing an overview of the data set can be extremely valuable • Helps present overall patterns • Assists user with navigation and search • Orients activities • Generally start with overview

IAT 355 | View transformations 49 Details

• Viewers also will want to examine details, individual cases and variables • How to allow user to find and focus on details of interest? • Generally provide details on demand

IAT 355 | View transformations 50 Overview + Detail

• Overview + Detail displays can be combined via either time or space • Time - Alternate between overview and details sequentially in same place • Space - Use different portions of screen to show overview and details • Develop visualization and interface techniques to allow flexible alternation

IAT 355 | View transformations 51 Worthy Objective

• Allow viewer to examine cases and/or variables in detail while still maintaining context of those details in the larger whole • Concession • You simply can’t show everything at once • Be flexible, facilitate a variety of user tasks • Visualization + Navigation

IAT 355 | View transformations 52 Example

IAT 355 | View transformations 53 Managing detail

• Single window with horizontal and vertical panning • Works only when image/space is not too much larger than the window

IAT 355 | View transformations 54 Single Window

• Single view with Selectable Zoom area • Selected zone is new view • Magnification and adjustment can follow • Context switch disorienting

IAT 355 | View transformations 55 Single Window

• Main + mini-map • Sometimes the Overview gets the most space • Depends on the user’s familiarity with the object of interest • Panning in one affects the other • Could be extended to 3 or more levels • Issue: How big are different views and where do they go?

IAT 355 | View transformations 56 Overview + Detail • K. Hornbaek et al., Navigation patterns and Usability of Zoomable User Interfaces with and without an Overview, ACM TOCHI, 9(4), December 2002.

IAT 355 | View transformations 57 Overview + Detail

• K. Hornbaek et al., Navigation patterns and Usability of Zoomable User Interfaces with and without an Overview, ACM TOCHI, 9(4), December 2002.

• A study on integrating Overview + Detail on a Map search task • Incorporating panning & zooming as well. • They note that panning & zooming does not do well in most studies. • Results seem to be • Subjectively, users prefer to have a linked overview • But they aren’t necessarily faster or more effective using it • Well-constructed representation of the underlying data may be more important. • More research needed as each study seems to turn up different results, sensitive to underlying test set.

IAT 355 | View transformations 58 Lens Technique

• Enlarged image floats over the overview • Neighbor objects obscured by the detail view

IAT 355 | View transformations 59 Variable lens focus view

Focus is at high magnification, periphery at low magnification • All in one view • Distortion can be disorienting

IAT 355 | View transformations 60 Overviews

• How to deal with approximate view? • Reduce the data elements • Eliminate • Sample • Aggregate • Reduce the visual representation • Need to render to sub-pixel resolution • Accumulate visual contributions per pixel

IAT 355 | View transformations 61 Focus + Context

• How is this different from Overview + Detail? • Focus + Context is an InfoVis term: • Present the Detail and the overview in the same window

IAT 355 | View transformations 62 Focus + Context Methods

• Filtering • Selective aggregation • Micro-macro readings • Highlighting • Distortion

IAT 355 | View transformations 63 Prototype example (a) An information space containing documents, emails, etc.

• Bifocal Display – Spence & Apperley, 1980

• Fisheye View - George Furnas, direction 1981 of view

(b) The same space wrapped around two uprights.

(c) Appearance of the information space when viewed from an appropriate direction

IAT 355 | View transformations 64 Definition

• Fisheye View • Magnify an area of interest without obscuring its neighboring unmagnified imagery • Why fisheye? • The fisheye camera lens

IAT 355 | View transformations 65 Fisheye Terminology

• Focal point • Distance from focus • Level of detail • Degree of interest function (DOI)

IAT 355 | View transformations 66 Level of Detail

• A number the determines the quantity of visual info you are going to draw for one data element • In maps: The quantity of imagery that fits in X pixels

IAT 355 | View transformations 67 Degree of Interest

• Function that determines how items in display are drawn DOI = Level of Detail – Distance From Focus DOI = Level of Detail / Distance From Focus 1.0, 1.0

Query Position 0.4, 0.7

Focal Point 0.8, 0.1

IAT 355 | View transformations 68 DoI Function

• Can take on various forms • Continuous - Smooth interpolation away from focus • Filtering - Past a certain point, objects disappear • Step - Levels or regions dictating rendering 0

IAT 355 | View transformations 69 Examples: 1D

• Fisheye Menus – Bederson • Dynamically change size of menu item & provide focus area around the pointer • Items near cursor displayed at full size • Items further away on either side are smaller • Uses a distortion function so items will always fill menu • Efficient mechanism for long menus • Need to “Lock Focus” to hit nearby targets (on right)

IAT 355 | View transformations 70 Elastic Presentation Space

IAT 355 | View transformations 71 Semantic fisheye

SpectraVis: Information Visualization for Supernova Spectra

IAT 355 | View transformations 72 Tasks

• Image generation: overview is important, but most of time is spent at detail level • Example: CAD • Open-ended exploration: overview not always complete; navigation must be fluent and easily mastered • Example: Interactive Map • Diagnostic: high detail, fluent panning and complete image coverage • Example: Circuit Design, Map directions

IAT 355 | View transformations 73 Tasks

• Navigation: global view with increased magnification detail areas; panning and zooming less important • Example: Geographic Information System

• Monitoring: Global view with multiple detailed views for local troubleshooting; window management is critical • Example: Network management

IAT 355 | View transformations 74 Efficiency Measures

• Spatial indexing • Hierarchy of objects based on bounding boxes • Clustering • Restructure hierarchy to maintain a balanced tree, speed for indexing • Level of detail • Render items depending on how large they are on screen, don’t draw small ones

IAT 355 | View transformations 75 Papers

• “Space-Scale Diagrams: Understanding Multiscale Interfaces” George Furnas – Fisheye Benjamin Bederson - Pad ++ CHI 1995 • Lyn Bartram, Albert Ho, John Dill, and Frank Henigman. 1995. The continuous zoom: a constrained fisheye technique for viewing and navigating large information spaces. In Proceedings of the 8th annual ACM symposium on User interface and software technology (UIST '95). ACM, New York, NY, USA, 207-215 • S. Carpendale. Examining comprehension issues in elastic presentation space, Information Design Journal, Volume 10, Number 1, 2001 , pp. 58-68(11).

IAT 355 | View transformations 76