Transcript PowerPoint Presentation - Space, Time and Antony

Space, Time, and Antony
Space, Time and Antony
Visualizing Then and Now,
Here and There
1
© 2009 IBM Corporation
Space, Time, and Antony
Who is Antony Unwin?
• There is no ‘H’ in ‘Antony’
• Otherwise you might find …
2
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
3
© 2009 IBM Corporation
Space, Time, and Antony
Hmmm…
• Without a statistical background, we might
just pick the most likely results:
–
–
–
–
“ANTHONY”
Bimodal distribution of ages
Born in 1908, 1978
Happy 32nd or 102nd birthday!
• Does “Anthony Unwin” have a web page?
4
© 2009 IBM Corporation
Space, Time, and Antony
Anthonyunwin.com
QuickTime™ and a
decompressor
are needed to see this picture.
5
© 2009 IBM Corporation
Space, Time, and Antony
Fortunately, we are statisticians
•
WWAD [What would Antony Do]?
1.
2.
3.
4.
5.
6.
7.
6
Get good data
Look at the data
Validate the data
Build models
Look at the model results
Validate the model results
Don’t trust a model more than the data
© 2009 IBM Corporation
Space, Time, and Antony
Some Data on Antony
• Sample Papers from 1991 onwards
• Look at the data
7
© 2009 IBM Corporation
Space, Time, and Antony
All Years
8
© 2009 IBM Corporation
Space, Time, and Antony
1991
9
© 2009 IBM Corporation
Space, Time, and Antony
1999
10
© 2009 IBM Corporation
Space, Time, and Antony
2001
11
© 2009 IBM Corporation
Space, Time, and Antony
2007
12
© 2009 IBM Corporation
Space, Time, and Antony
All Years
13
© 2009 IBM Corporation
Space, Time, and Antony
Space and Time
• Base data shown is TEXT x TIME
• This relates back to early work I did with
Antony on DIAMOND FAST, exploring
time series data.
• The “eyeballing time series” paper exhibits
several features later made popular by
Shneiderman …
14
© 2009 IBM Corporation
Space, Time, and Antony
Visual Information-Seeking
Mantra
Overview first
Zoom and filter
Then details-on-demand
Shneiderman, 1996
15
© 2009 IBM Corporation
Space, Time, and Antony
Unwin, 1988
• Eyeballing Time Series
“Explore, Interrogate and Manipulate
Data Series Graphically”
16
© 2009 IBM Corporation
Space, Time, and Antony
Overview First
17
© 2009 IBM Corporation
Space, Time, and Antony
Overview First
QuickTime™ and a
decompressor
are needed to see this picture.
18
QuickTime™ and a
decompressor
are needed to see this picture.
© 2009 IBM Corporation
Space, Time, and Antony
Zoom and Filter
• “Changing time scaling … can be done dramatically by a factor
of 10 with a single click or more precisely by dragging a scaling
bar. Here are two views of monthly Canadian unemployment
data. The graph on the left of Figure 2 shows all 20 years of the
series (240 points) but has been crushed into a very short timescale. The overall trend and range of seasonality are clearly
shown.
• The graph on the right of Figure 2 shows only some of the
series, just under two years, at a scale 100 times bigger. The
points have been marked to emphasise the effect”
QuickTime™ and a
decompressor
are needed to see this picture.
19
© 2009 IBM Corporation
Space, Time, and Antony
Details on Demand
QuickTime™ and a
decompressor
are needed to see this picture.
20
© 2009 IBM Corporation
Space, Time, and Antony
+ Modeling
21
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
22
© 2009 IBM Corporation
Space, Time, and Antony
Space and Time
• Statistical Graphics are a SPACIAL layout
of glyphs based on a mapping from data.
• Studying geographic visualization informs
on statistical visualization.
• A second project followed Diamond Fast:
REGARD
23
© 2009 IBM Corporation
Space, Time, and Antony
What did REGARD introduce?
• Linked views:
– “Details-on-demand” on steroids
– Mix Statistical Graphics and Domain-Specific
Graphics
– New ways of displaying data subsets
• Spatial Data
– Layers: Multiple elements for different data
24
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
25
© 2009 IBM Corporation
Space, Time, and Antony
Model Spatial
Correlation
• Domain views:
variogram cloud,
map view
• Link to map
view with edge
element
QuickTime™ and a
decompressor
are needed to see this picture.
• Show selected
subset in linked
views
26
© 2009 IBM Corporation
Space, Time, and Antony
Linking
QuickTime™ and a
decompressor
are needed to see this picture.
Districts of the city of Dublin showing areas
with high levels of average income.
27
© 2009 IBM Corporation
Space, Time, and Antony
Showing
Selections
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
28
© 2009 IBM Corporation
Space, Time, and Antony
REGARD
QuickTime™ and a
decompressor
are needed to see this picture.
29
© 2009 IBM Corporation
Space, Time, and Antony
Plus ça change, plus c'est la même chose
•Wills, G. (2010) Visualizing Time Springer, 2010
•Wills, G., and Wilkinson, L. (2010) Autovis: Automatic Visualization
Information Visualization, December 18, 2008
•Wilkinson, L and Wills, G. (2008) Scagnostics Distributions Journal of
Computational and Graphical Statistics, Volume 17; pp.473-491
•Graham J. Wills (2008) Linked Data Views Handbook of Data Visualization;
Chen, Haerdle and Unwin
•Wills, Graham J. (2002) Visualization Handbook of Data Mining and Knowledge
Discovery
•Wills, Graham (2009) Visualizing Network Data Encyclopedia of Database Systems,
Wills, Graham (2009) Visualizing Hierarchies Encyclopedia of Database Systems,
•Graham Wills (2006) Networks Graphics of Large DataSets; Unwin, Theus and
Hofmann (eds)
•Wills, Graham J. and Keim, D. (2002) Data Visualization For Domain Exploration:
Interactive Statistical Graphics Handbook of Data Mining and Knowledge Discovery.
30
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
Space and Time;
Overviews
Visualizing Time; Chapter 7
31
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
Aligned Views;
Overviews
Visualizing Time; Chapter 6
32
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
Details on
demand
Visualizing Time; Chapter 10
33
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
Don’t
trust the
model
more
than the
data
Visualizing Time; Chapter 9
34
© 2009 IBM Corporation
Space, Time, and Antony
Final Lessons
Antony is also close to
data; stay there also
Look at the data many
ways with appropriate
tools and plots
35
© 2009 IBM Corporation
Space, Time, and Antony
Final Lessons
Variables + Models =
INFORMATION
Statistics, analysis and
procedures must be close
to representations and
must be investigated for
features, errors and ideas
36
© 2009 IBM Corporation
Space, Time, and Antony
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
QuickTime™ and a
decompressor
are needed to see this picture.
37
© 2009 IBM Corporation