visualization of time-oriented datavda.univie.ac.at/Teaching/Vis/13s/LectureNotes/16_time_wolfgang.pdf · WOLFGANG AiGNER! visualization of time-oriented data 29 Low-level Task List

1 WOLFGANG AiGNER! visualization of time-oriented data

Wolfgang Aigner [email protected]

http://ifs.tuwien.ac.at/~aigner/

Version 4.0 June 3, 2013

visualization of time-oriented data

2

About me

Basic and applied research + consulting projects

Focus: interactive visualization and data analysis of time-oriented data

60+ peer-reviewed publications

10+ years teaching experience

active member of scientific community InfoVis:Wiki – www.infovis-wiki.net TimeViz Browser – survey.timeviz.net EvalBench – www.evalbench.org


Contents

Introduction What makes time a special data dimension?

What is time-oriented data?

Visualization Techniques Some selected techniques

Then: YOU decide, what we will look at

Roundup & Conclusions


Data types 1-dimensional

2-dimensional

3-dimensional

Temporal

Multi-dimensional

Tree

Network

= 4D space “the world we are living in”

[Shneiderman, 1996]


Spatial + temporal dimensions Every data element we measure is related and

often only meaningful in context of space + time

Example: price of a computer where? when?


Differences between space and time Space can be traversered “arbitrarily”

we can move back to where we came from

Time is unidirectional we can’t go back or forward in time

Humans have senses for perceiving space visually, touch

Humans don’t have senses for perceiving time


…travel in time virtually.

Interactive visualization Gives us the ability to…


Time-oriented data?

Event calendar

Snow height & sunshine hours

Organization chart

iPod price

next >

9 WOLFGANG AiGNER! visualization of time-oriented data ^ up

Event calendar

10 WOLFGANG AiGNER! visualization of time-oriented data ^ up

Snow height & sunshine hours


Organization chart

^ up

time

1998 2000 2002


iPod price

^ up


What is time?

“If no one asks me, I know. But if I wanted to explain it to one who asks me, I plainly do not know.” -- Augustinus (AD 354-430, The Confessions)

”Die Empfindung der Zeit hängt davon ab, auf welcher Seite

der geschlossenen Klotür man sich befindet.”

-- Albert Einstein



no formal definition

what is considered as time-oriented data depends on the intended task

a possible definition:

Data, where changes over time or

temporal aspects play a central role or are of interest.


Visualization of time-oriented data

What? time & data 1 Why? user tasks 2 How? visualization & interaction 3


Modeling time


Scale

ordinal only order is known

discrete every element of time has a unique predecessor and successor comparable to Integer

continuous between any two elements in time there might be another one

in between dense time comparable to Float

A B C

D


Arrangement

linear

each element of time has a unique predecessor and a unique successor

cyclic

summer is before winter, but winter is also before summer


Viewpoints

ordered

multiple perspectives

branching

Past Definite time - data

element assignment

Present Currently valid

state

Future Planning

Temporal uncertainty

Alternative scenarios


Modeling time


Granularity


Calendar

System of granularities and mappings between them

representational scheme

for human readability and usability

Calendars Gregorian Academic (semester, trimester, ...) Financial (Quarters, Fiskal, ...) ...


Example: Granularity paradoxon


Time primitives

anchored

instant - single point in time

interval - duration between 2 instants

unanchored

span - duration of time

^ up


Determinacy

determinate complete knowledge of temporal attributes

indeterminate incomplete knowledge of temporal attributes

no exact knowledge

i.e. “time when the earth was formed”

future planning

i.e. “it will take 2-3 weeks”

imprecise event times

i.e. “one or two days ago”

multiple granularities


Characterizing data


Relating data & time





Low-level Task List 1/2

Existence of a data element Does a data element exist at a specific time? Example: Was a measurement made in July, 1960? !

Temporal location When does a data element exist in time? Example: Is there a lecture taking place on November 24, 2005?

Time interval How long is the time span from beginning to end of the data element? Example: How long was the processing time for data set A?

Temporal texture How often does a data element occur? Example: How often was Jane sick last year?!

[McEachren, 1995]!


Low-level Task List 2/2

Rate of change How fast is a data element changing or how much difference is there from

data element to data element over time?

Example: How much did the price of gasoline change since last September?

Sequence In what order do data elements appear?

Example: Did the explosion happen before or after the car accident?

Synchronization Do data elements exist together?

Example: Is Jill’s birthday on Easter Monday this year?

[McEachren, 1995]!


Task Taxonomy 1/2 [Andrienko & Andrienko, 2006]!


Task Taxonomy 2/2 [Andrienko & Andrienko, 2006]!





Visualization roots Statistics

Visualization of time-series.

The time-series plot is the most frequently used form of graphic design. [Tufte, 1983]

Mostly one parameter over time.

t!

y!


Early time-series plot

Part of a text for monastery schools 10th or 11th century (!) Inclinations of the planetary orbits over time 800 years before other time-series plots appeared


Train schedule

Paris to Lyon (1880s)


Visualizing time-oriented data


Visual mapping of time Dynamic: Time → Time (Animation)

probably the most natural form of mapping no “conversion” of concepts needed in between well suited for

keeping track of changes

following trends and movements

not well suited for

analytic and explorative tasks

no direct comparison of parameters between different points in

time is possible

Static: Time → Space

mapping of time to visual features direct comparison of parameters between different points in time is

possible


Visual variables

position most common mapping

the most accurately perceived visual feature

length second most accurate attribute

typically, the length of an object denotes the duration, as for example in timelines!


Visual variables angle, slope

analog-clock-based visualizations

!

connection connecting arrows or lines

“before element” --> “after element”

!

text, label simple text labelling

often combined with “connection”!


Visual variables

line (thickness) Increasing or decreasing with time

color (brightness, saturation, hue) brightness most appropriate

“fading away” against the background

transparency


Visual variables

area

enclosure

size

texture

shape

less suited


Interaction facilitates active discourse with the data and visualization

see think

modify

[Card et al., 1983]


Interacting with time

specific interaction techniques

+

task & interaction taxonomies

[VisuExplore project]

[VisuExplore project: measure tool]




+


[CHI09 workshop, VisuExplore project]

[Animated Scatterplot project]




+


[CHI09 workshop, VisuExplore project]

[CareCruiser project]




48

SELECTED VISUALIZATION TECHNIQUES

49

Instant

50

Sparklines

Simple, word-like graphics intended to be integrated into text

adds richer information about the development of a variable over time that words themselves could hardly convey

can be integrated seamlessly into paragraphs of text, can be laid out as tables, or can be used for dashboards

emphasis of first, last, min, and max value using colored dots

[Tufte, 2006]

51

TimeSearcher

visualization tool for time-series data

timebox query model rectangular regions that specify constraints over time series data sets

x-axis extent: time period of interest

y-axis extent: constraint on the range of values

combinations of multiple timeboxes

data + query envelope

[Hochheiser, 2002; Hochheiser and Shneiderman, 2002]

DEMO http://www.cs.umd.edu/hcil/timesearcher/

52

Horizon Graph

visualization technique for comparing a large number of time-dependent variables

based on the two-tone pseudo coloring

[Reijner, 2005]

53

Cycle Plot

technique to make seasonal and trend components visually discernable

showing individual trends as line plots embedded within a plot that shows the seasonal pattern

mean value for each weekday as grey line

[Cleveland, 1994]

54

Spiral Graph

encodes time-series data along a spiral

emphasizing the cyclic character of the time domain

data values are visualized using symbols for nominal data (left) as well as color and line thickness for quantitative data (right)

[Weber et al., 2001]

55

ThemeRiver

Visualize thematic variations over time.

Across a large collection of documents.

River Metaphor: the “river” flows through time.

Changing width to depict changes.

Themes or topics are colored “currents”.

[Havre et al., 2000]

56

Space-Time Cube

two spatial dimensions (latitude and longitude) along the x-axis and the y-axis

time along the z-axis

[Kraak, 2003]

57

Interval

58

Gantt Chart

Left: indented list of tasks

Right: timelines show position and duration of tasks

59

PlanningLines

representation of temporal uncertainties

facilitates the identification of (un)defined attributes

supports in maintaining logical constraints (e.g., bars may not extend caps)

gives a visual impression of the individual and overall uncertainties

[Aigner et al., 2005]

60

www.timeviz.net

Wolfgang Aigner • Silvia Miksch Heidrun Schumann • Christian Tominski

Visualization of Time-Oriented Data with a foreword by Ben Shneiderman

Springer

1st Edition., 2011, XVIII, 286 p. 221 illus., 198 in color. Hardcover, ISBN 978-0-85729-078-6.

Table of Contents Introduction • Historical Background • Time & Time-Oriented Data • Visualization Aspects • Interaction Support • Analytical Support • Survey of Visualization Techniques • Conclusion

NEW BOOK!


62

3 Key Criteria

data frame of reference – abstract vs. spatial variables – univariate vs. multivariate

time arrangement – linear vs. cyclic time primitives – instant vs. interval

vis mapping – static vs. dynamic dimensionality – 2D vs. 3D

63

64

65

survey.timeviz.net


Visualization design

[Aigner, Miksch Schumann, Tominski, 2011]

What is presented? time and data

Why is it presented? user tasks

How is it presented? visual representation


Roundup What makes time a special data dimension?


Concepts of time-oriented data

Visual variables for representing time

User Tasks for time-oriented data

Visualization techniques


Conclusions time-oriented data covers a very broad field

what is considered as time-oriented data is task dependent

a lot of different techniques available

visualizations are task driven

periodic behavior is very common but relatively underexplored

Generally:

Visualization sparks interest for further investigation


"Die Zeit verlängert sich für alle, die sie zu nutzen verstehen.” -- Leonardo Da Vinci

70

Documents

visualization of time-oriented datavda.univie.ac.at/Teaching/Vis/13s/LectureNotes/16_time_wolfgang.pdf · WOLFGANG AiGNER! visualization of time-oriented data 29 Low-level Task List