The longer I work in the field of information visualization, the more I come to appreciate the simplicity, versatility and effectiveness of basic bar charts. They are not attention grabbing, but by encoding the information on the most powerful visual channels (size and position, with color as an option) they are remarkably accurate. They can fit anywhere from primary elements in basic infographics to small-multiples bar chart tables, and work well in different orientations and aspect ratios.
While experimenting with the paper.js graphics library recently, I used bar charts to create this date/time bar chart clock. All of the bars will fill up completely at the end of the century. This seems like such a simple idea that I am sure that someone else must have done this before, but quite a few minutes of googling and flipping through books did not turn up anything. If you are aware of another variation on this date/time clock please contact me and let me know.
I recently wrote a whitepaper for IBM on ways in which visualization can be used to effectively understand big data. Obviously, that’s a “big” topic. For this whitepaper I focused more on the complexity of big data than on the sheer scale, and on the ways in which visualization can be used to capture complex facets of the data in a “Customer 360″ scenario where an organization may be looking to get a more complete picture of various aspects of their customers including time, social networks, key social influencers and customer relations. It starts out with a simple stacked bar chart, and builds from there to increasingly complex visualization examples. The whitepaper is available @ ibm.biz/bigdatavis.
IBM just posted a quick blog post of mine on their Business Analytics blog site. The post addresses the general theme of visual navigation, comparing old school approached with external controls vs. integrated approaches that combine the visual views and navigational controls within a single metaphor. To illustrate these concepts I compare the use of traditional dashboard tabbed hierarchies with a single hierarchical visualization that shows multiple levels of detail simultaneously. Read the short article for more details.
It’s been awhile since I published on this blog, which is because I joined IBM’s new Center for Advanced Visualization last December and have been busy getting up to speed on everything that is happening there. It is an exciting opportunity for me, I will be working to help shape IBM’s recent and significant investments in effective visualization and visual analytics across IBM. There will be much more for me to post about that in the future, but for starters here is a link to my first post on their Business Analytics blog. It is a somewhat humorous comparison between visualization chart types and vehicle types. I often encounter novice chart users who don’t know what type of chart to use for their data. This post is meant to reinforce some of the conventional wisdom in the visualization community as to how to select chart types, hopefully in a memorable way.
For those of you lacking the time to click through for the entire article: Pie Chart == Horse & Buggy in my system. You’ll have to click through to find out what vehicle matches the 3D Pie/Bar chart however.
Now that I’ve had a chance to recover a bit from the deluge of London 2012 olympic-related infographics and visualizations, I thought it would be interesting to see how the top countries compare in terms of total number of medals won over the history of their participation in the summer games. I used Tableau to do an initial analysis, and then selected the top 10 countries for the D3 animated line graph that is shown here. Each line shows the accumulated total of medals for that country for all years up to and including that year. The circles are sized proportionally to the average number of medals won each year over the years that the country was participating. Mouse over the chart elements for details.
I had expected the USA to show up well on this graph, but was surprised at just how dominant they have been since 1896. This certainly illustrates one aspect of “the American Century”. China looks to be an up and coming nation in the medal count, however it remains to be seen if their rise will continue over the long haul or if it will flatten out (or disappear) as has happened with other countries. The US has won an average of 96 medals per year, which is topped only by the old Soviet Union at 112 medals per year.
This graphic was inspired by the awesome view of Mariano Rivera’s all time saves record by the New York Times. Note that the line for Germany does not include any medals from the period when it was split into West and East Germany. Arguably those years should be added in, however I am going with the official medal counts as reported by the IOC. I also used a linear scaling on the circle sizes rather than a sqrt scale, in order to emphasize the differences between averages.