partition chart Archives

In Ineffective Chart – Partition Chart, I discussed the partition chart that Chandoo spoke of in Partition Charts in Excel – alternative to pie charts [visualization hack]. I felt that the partition chart hid the data it was meant to present. Jorge Camoes remarked that it looked to him like the map of a fragmented hard drive’s file system.

In Partition Charts – Continuation – [More visualization fun] Chandoo has reworked his partition chart, defragging it to bring all same-colored squares together. The original chart

Partition Chart in Excel

now has just three contiguous colored regions

Partition Chart in Excel - Defragged

That looks less cluttered than the original chart with its distracting camouflage effect, but by eye, I still can’t really resolve the difference between 30% light orange and 22% brown, and those numbers aren’t really close enough to be uncertain about them. The pie chart I used in my post showed these values more readily.

Pie Chart in Excel

As always, the simplest and clearest way to show this data is in a bar or column chart:

Bar Chart in Excel

The defragged partition chart reminds me of the “square pie” we first read about in Eager Eyes (Women in IT – Squaring the Pie?), and which was taken up by Juice Analytics (Squaring the Pie), Dashboards by Example (The Square Pie Chart), Anil Dash (Pixels Are The New Pies), Info Clarity (A square pie), Bella Consults (Square pies taste bad), and others. Here are some examples:

Square Pie Chart
Women in IT

New York Times (left) cited by Anil Dash and Juice Analytics.
Wired Magazine (right) cited by Bella Consults and Dashboards by Example.

Produced and cited by Info Clarity

The square pie chart was all the rage for a while. Consensus seemed to be split between “they’re wicked cool” and they’re even less effective than round pie charts. My own opinion is that they are ineffective: the awkward shapes imposed by rows and columns of filled squares make comparison of areas even more difficult than comparing areas of rectangles with unequal lengths and widths or comparing areas of rotationally misaligned wedges in a pie chart.

Bottom line: you shouldn’t use square pies or “defragged” partition charts, use a bar chart instead.

In Partition Charts in Excel – alternative to pie charts [visualization hack], Chandoo offers a “partition chart” as an alternative to a pie chart. We all know pie charts are bad (don’t we???), so it’s good to look for alternatives. Here is part of Chandoo’s partition chart:

Partition Chart in Excel

This chart uses conditional formatting to achieve the mottled appearance. It’s almost reminiscent of camouflage, partly because of the mottled colors, and partly because the information is well hidden. The colors cover 22%, 30%, and 48% of the area of the chart. Without looking at the legend of the full partition chart below, can you at a glance match up the colors with their associated proportions? I couldn’t, either.

Here is Chandoo’s complete chart, including the legend:

Partition Chart in Excel

As I told Chandoo in a comment to his post:

While pie charts are somewhat ineffective at showing relative proportions, this partition chart is completely ineffective. If not for the legend, I would have guessed that the darkest color accounted for the largest, not the smallest, proportion. The pie at least puts all of its bits into one wedge per category, while the partition chart keeps all the bits separate, and they are scattered across the chart. It’s as if you spilled a package of M&Ms on the table and tried to determine the portions of each color.

Since there are only three categories, a pie chart isn’t a terrible choice for this data:

Pie Chart in Excel

As always, the simplest and clearest way to show this data is in a bar or column chart:

Bar Chart in Excel

Chandoo spoke of the partition chart as a way to look at geographic variations in data. He mentioned that it might be sales of your products in a number of cities. It makes more sense if you were looking at a map of counties, states, regions, or countries, and you were filling each geographical unit’s shape with the fill color. The user would not be expected to integrate the entire map to get an overall percentage. Instead the user could compare the colors of familiar shaped geographical units and make sense of the data. Wikipedia has some examples of this in Red states and blue states:

In this context, regions of different colors can effectively present information about, in this case, election results in different states. In contrast, the partition graph is a map of nothing.