Tableau Public Vizes in the Wild

le février 17, 2010

Worldwide, users of Tableau Public are beginning to publish some amazing views. Bloggers, journalists, government statisticians and even community groups are taking advantage of Tableau Public. It is great to see the wide variety of users and the analyses they are doing. Enough of me- take a look at a sample of the vizes!

This viz, from Eric Lai at ComputerWorld shows the number of airborne battery incidents over the past twenty years. Of the named carriers, FedEx has by far the most incidents; perhaps this is a reflection of the fact that they transport so many? Visually, this view is interesting because it uses three color schemes. Under normal circumstances this would be a bad idea since it is easy to confuse colors between visualizations. However, the colors in these schemes were designed to differ from each other so they are more difficult to confuse. Therefore, the added analytical benefit of color can still be utilized.

The small but enlightening viz above was posted by the Pan American Health Organization. Generally, the further towards the top left of the viz, the more developed and healthy the nation is. People familiar with GapMinder will appreciate how you can see the relationships on the viz changing by moving the year slider. Once again, interactivity allows greater analytic capability than even a well designed static chart.

Anyone interested in the root cause of the dot-com bubble will find this viz, originally posted on the WiredBarn Blog, fascinating. Below "Medium", click on "Internet" to highlight the share of advertising devoted to the internet. Despite predictions of the internet taking over the world of advertising (and billionaires being made overnight) it looks like newspapers and television still reign supreme. The highlighting feature is especially useful on bar charts like this because it allows you to compare the sizes of the bars without straining your eyes.


Great stuff.

I notice on the final viz that the Year filter suffers from an all-too-easy problem. For humans, years are whole numbers without comma-separators. For Tableau, a year is normally recognised as a number, with decimals. Therefore, the default filter is a continuous, thousand-separator, slider scale. Not ideal for the human interactor!

My normal solution (from memory - I don't have access to Tableau from where I am right now) is to duplicate the Year dimension, and change the duplicate to a Discrete measure. Then change its default number format to have 0 decimals and no thousand separator. Keep the original Year dimension on the chart, but put the duplicate on the Filter shelf. Voila - whole number year filters!

Do correct me if I'm wrong - I can't check this right now. I'm also sure there's plenty of other ways around this.


Ah - I was wrong! You get the whole numbers with a *Continuous* dimension, with default number format set to zero decimal places and no thousand separators.