Graphics Reference
In-Depth Information
Computational Statistics
and Data Visualization
1.1
his book is the third volume of the Handbook of Computational Statistics and cov-
ers the field of data visualization. In line with the companion volumes, it contains
a collection of chapters by experts in the field to present readers with an up-to-date
andcomprehensiveoverviewofthestateoftheart.Datavisualizationisanactivearea
of application and research, and this is a good time to gather together a summary of
current knowledge.
Graphic displays are oten very effective at communicating information. hey are
also very oten not effective at communicating information. Two important reasons
for this state of affairs are that graphics can be produced with a few clicks of the
mousewithout any thought and thedesign of graphics is nottaken seriously in many
scientific textbooks. Some people seem to think that preparing good graphics is just
a matter of common sense (in which case their common sense cannot be in good
shape), while others believe that preparing graphics is a low-level task, not appropri-
ate for scientific attention. his volume of the Handbook of Computational Statistics
takes graphics for data visualization seriously.
Data Visualization and Theory
1.1.1
Graphics provide an excellent approach for exploring data and are essential for pre-
senting results. Although graphics have been used extensively in statistics for a long
time, there is not a substantive body of theory about the topic. Quite a lot of atten-
tionhasbeenpaidtographics forpresentation, particularlysincethesuperb topics of
Edward Tute.However, this knowledge is expressed in principles to be followed and
not in formal theories. Bertin's work from the
s is oten cited but has not been
developed further. his is a curious state of affairs. Graphics are used a great deal in
many different fields, and one might expect more progress to have been made along
theoretical lines.
Sometimes in science the theoretical literature for a subject is considerable while
thereislittle appliedliterature tobefound.heliterature ondata visualization isvery
much the opposite. Examples abound in almost every issue of every scientific jour-
nal concerned with quantitative analysis. here are occasionally articles published in
a more theoretical vein about specific graphical forms, but little else. Although there
is a respected statistics journal called the Journal of Computational and Graphical
Statistics, most of the paperssubmitted there are in computational statistics. Perhaps
this is because it is easier to publish a study of a technical computational problem
than it is to publish work on improving a graphic display.
Presentation and Exploratory Graphics
1.1.2
he differences between graphics for presentation and graphics for exploration lie
in both form and practice. Presentation graphics are generally static, and a single
