Class 3: Data Formatting, Word Clouds, Network Visualizations
1. Class 3:
1) Data formatting and manipulation continued….
Last class: dealing with quantitative data, case study: statistical distributions
Today: dealing with qualitative nominal data, case study: text data
Making Word Clouds:
Tagcloud
Word cram
Processing example with and without Word Cram
BREAK
2) Network visualizations: Example with Gephi
BREAK
3) Design considerations:
• Cognition and Perception
• Gestalt Principles
• Bertin’s semiotics and use of metaphors
2. Another valuable resource to add to the list of useful websites:
http://www.visualcomplexity.com/vc/
For cleaning up data, see:
google refine (https://code.google.com/p/google-refine/ )
3. DATA
Quantitative
(Numerical)
Qualitative
(Descriptive)
Nominal
Data has no
natural order.
Includes objects,
names, and
concepts.
Examples:
gender, race,
religion, sport
Ordinal
Data can be
arranged in
order or rank
Examples: sizes
(small, medium,
large), attitudes
(strongly
disagree,
disagree,
neutral, agree,
strongly agree),
house number.
Continuous
Data is
measured on a
continuous
scale.
Examples:
Temperature,
length, height
Discrete
Data is
countable, and
exists only in
whole numbers
Examples:
Number of
people taking
this class,
Number of
candy bars
collected on
Halloween.
5. Word Cloud Workshop
Tools:
1) http://tagcrowd.com/ (and many, many others. Wordle, etc)
2) Coding with Processing
a) Using the word cram library (http://wordcram.org/ )
b) Without the word cram library
6. Visualizing word frequency and order
Image Source:
http://www.openbible.info/blog/2009/03/phrase-net-bible-visualizations/
11. Download here - http://gephi.github.io/
Network Visualization Tool: Gephi
Get sample data sets here - https://github.com/gephi/gephi/wiki/Datasets
Network Visualization Workshop
12. Design Considerations
1) Perception and Cognition
• Perception is fragmented
• Eyes are constantly scanning and constructing reality
Image from: Ware, Colin. Visual thinking: For design. Morgan Kaufmann, 2010.
Original Study: Daniel J. Simons and Daniel T. Levin. 1998. “Failure to detect changes to people during a
real world interaction.” Psychonomic Bulletin and Review. 5: 644–669.
https://www.youtube.com/embed/FWSxSQsspiQ
13. Visual thinking is about finding patterns
Bottom-up processes gather information and build patterns.
Top-down processes determine where you look and what you pull out from the patterns
Image from: Ware, Colin. Visual thinking: For design. Morgan Kaufmann, 2010.
14. Pre-Attentive Processing
• Bottom-up
• Fast, automatic
• Instinctive
• Efficient
• Multitasks
• Top-down
• Slow, deliberate
• focused
Attentive Processing
Goal of information design:
• help humans process information as efficiently as possible.
• make as much use of pre-attentive processing as possible.
https://www.youtube.com/watch?v=vJG698U2Mvo
17. 2) Gestalt Principles
• Figure/Ground
• Proximity
• Similarity
• Symmetry
• Continuity
• Closure
Visual information is understood holistically before it is examined separately.
Our brains use Gestalt principles create unity in a composition.
Designers reinforce unity by applying Gestalt principles.
An image composed of units that are unrelated in size style orientation and color
appears chaotic and unresolved.
20. Similarity
Our perceptual tendency to conceptually group objects that are similar in
shape, size, color, or texture.
Image Source:
http://graphicdesign.spokanefalls.edu/tutorials/process/gestaltprinciples/gestaltprinc.htm
Image Source:
http://facweb.cs.depaul.edu/sgrais/gestalt_principles.htm
22. Continuity
Our perceptual tendency to see lines as continuous even when they are
intersected; and to see two groupings of similar things as one interrupted group.
23. Closure
Our perceptual tendency to fill in missing parts of an object so it appears whole.
Image Source:
http://facweb.cs.depaul.edu/sgrais/gestalt_principles.htm
Panda image Source:
http://graphicdesign.spokanefalls.edu/tutorials/process/gestaltprinciples/gestaltprinc.htm
24. 3) Jacques Bertin: Semiology of Graphics, 1967
Graphic by: Sheelagh Carpendale
25. Visual variables for quantitative data (used to represent quantities)
Position
Size
Value
Time
Distance
26. A word on size: Area vs. Length
Image sources: http://si.wsj.net/public/resources/images/BF-AI151_ALIBAB_G_20140905165108.jpg
http://www.zerohedge.com/sites/default/files/images/user5/imageroot/2013/08/20130826_TSLA.jpg
27. When making bubble visualizations:
use area to represent quantities, not radius or diameter
The bigger circle has twice the
diameter as the smaller circle,
and four times the area
The bigger circle has twice the
area as the smaller circle.
28. Visual variables for qualitative data (used to represent a category)
Texture
Colour
Orientation
Shape
30. Data mapping
• One dimensional data: easy. Example: Simple bar graph
• Multi-dimensional data: more tricky
– Map each data dimension to a different visual variable
– Chose visual variables that suit data type
– Avoid interference with pre-attentive perception
– Map primary data to primary visual feature; secondary to secondary, etc