Introduction to Visualization
Visualization is the communication of information using graphical
representations. A picture is worth a thousand words because a
picture can be processed in parallel by the brain whereas words
have to be processed sequentially. Moreover, pictures can be
understood in a language independent way.
There are various ways that data can be presented to the user.
- table of data
- map (region or subway)
- chart (weather, stocks)
- signs (highway)
- 3D models (buildings, human body, distribution of galaxies)
Importance of Visualization
The importance of visualization cannot be overstated. Simply put the
key areas that visualization is important are:
- synthesize and convey information
- aid in making decisions
- tell a story
Brief History of Visualization
Pictures have been used to communicate since prehistoric times. Some of
the oldest writing systems used pictures to encode symbols and words -
logograms. Limestone tablets found in Mesopotamia are pictographic and
the beginning of syllabic and cuneiform scripts.
From the Egyptians we get hieroglyphics - pictorial writing. These
writings were born out of necessity - travel, commerce, religion, and
communication.
John Snow's map of cholera deaths in London in 1854. Each stacked bar
represents one deceased individual. He found that cholera occurred
entirely among those who drank from the Broad Street water pump. He
had the handle of that pump removed ending the epidemic that had taken
more than 500 lives.
Minard produced a brilliant representation of linked geographic and
time series data. The map emphasizes the loss of troops during the
Napoleonic Russian expedition. The size of the French army went from
400,000 to 10,000.
A break through came in data visualization with the abstract
representation for axes. Examples include:
Vizualization Today
We use visualization to communicate both qualitative and quantitative
information. Modern visualizations use digital media. They provide
a rich description of the data. Consider the
Anscombe
Quartet. They have the same mean and standard deviation but are
very different distributions.
Visualization was once considered to be a sub-field of Computer Graphics.
It uses graphical primitives like points, lines, areas, and volumes. But
visualization goes beyond graphics. It is connected with data, most
often there is statistical processing and there is a narrative involved.
Milestones in Data
Visualization
Process of Visualization
This is a simplified overview of the process:
- Data - what type is it?
- What do you want to do with it?
- explore - look for interesting things
- test - confirm or disprove a hypothesis
- present - results of one's analysis
- How do you want to display the visualization - static or
interactive?
Visualization is part of a larger process - exploratory data analysis,
knowledge discovery, or visual analytics. The visualization pipeline
is as follows:
- Data Modeling: store data in a data frame or record for rapid
access and easy modification.
- Data Selection: chose a subset of data to be visualized.
- Data to Visual Mappings: map data values to graphical entities.
- Scene Parameter setting (view transformations): specify attributes
of the visualization that are relatively independent of the data
like color map selection.
- Rendering or generation of the visualization: specify projections,
shading, texture, axes, annotations.
Human Perception and Visualization
If the goal of the visualization is to accurately convey information with
pictures then human perceptual abilities have to be considered. Consider
the following
illusions.
Most people cannot gauge textures accurately but our abilities to perform
relative comparisons are much stronger than absolute ones. About half the
human brain deals with visual input and the processing is parallel and
continuous. The preattentive processing is fast and can identify differences
in color or texture. Other features the system deals with are line
orientation, length, width, size of an object, curvature, grouping, and
motion.
The Gestalt School of Psychology attempted to define a set of laws by which
we define patterns. These laws included rules about proximity, similarity,
continuity, closure, symmetry, foreground and background, and size.