Author: Julie Steele, and Noah Iliinsky
Publisher: O'Reilly, 2010
Aimed at: Any programmer who needs to represent data
Pros: Includes some genuinely good ideas and is an entertaining read
Cons: Little code, can seem a little repetitive,
Reviewed by: David Conrad
The book is a collection of essays so it is a mixed bag both in terms of style and especially in subject matter - read on to discover more
The "Beautiful" series from O'Reilly has been a mixed bag but this one is a lot of fun. The first thing to say is that this is not about statistics and statistical data presentation - it really is about visualization. That is taking some data and trying to find a way to represent it graphically so that the end user can understand and perhaps draw some conclusions form their understanding.
The book is a collection of essays so it is a mixed bag both in terms of style and especially in subject matter. The first essay - "On Beauty" - isn't a general as its title might suggest. It considers what makes a good visualization by comparing examples to the periodic table and the classic London underground map. The second essay is in the same vein in that it considers how to best to represent data as an area based chart.
For me the third essay is where the book really started to be interesting. Its about Wordle - a word charting program. It hardly contains any code but it does consider the algorithms needed to create word charts that go beyond the simple key word map and make use of typography to present information in new and clever ways.
From here we have essays on color and dimensionality, creating a better New York Subway map, plotting flight pattern data, social graphs, representing search patterns, visualising the editing patterns in Wikipedia, tree ring plots and more. The only problem is that by the time you reach chapter 16 you have seen a lot of variations on network diagrams and things can seem a little repetitive. However there are some genuinely good ideas on display and most of the material is very well worth working through and a lot of it is entertaining.
Chapter 17 takes us into a different area - an immersive visualisation system that is used to allow for example scientists to get to grips with their abstract world of atoms and electron density plots. Chapter 18 takes another turn and look at Postmortem visualization - i.e. tomographic scans of dead people. The final chapter returns to lighter material and considers animation and its role in visualization.
I can't say that I enjoyed every essay equally and I have to admit that my concentration wandered at various points in the book, but overall it was a good experience. The only problem is that this is not a cheap book - in part because it is produced with high quality color printing and perhaps in part because the profits are going to charity - Architecture for Humanity.
What is missing from this book is code and detailed descriptions of methodology. This makes it more accessible and more fun to read but it reduces it to the status of a coffee table book. It also doesn't really do justice to the the cutting edge visualization going on in science and engineering.
As long as all you want is an entertaining read on the subject of visualization this book works.