The Graphic Continuum

Jon Schwabish and Severino Ribecca recently released a poster taxonomy of different types of charts, and how they all relate to each other. We think this is a great resource for designers everywhere, so we were especially interested in their take on the project. The Graphic Continuum began as I thought about the different ways we can plot data into different types of charts. My understanding of the different relationships between charts evolved over time by reading a variety of data visualization books, sketching different ideas and layouts, and presenting my ideas to different audiences. As my thoughts developed, I wanted to create a visualization that showed these different chart types and how they relate to one another, and I wanted to create something tangible that people could hold and point to as they worked with their data and their visualizations. I eventually teamed up with Severino Ribecca—who created and runs The Data Visualisation Catalogue—to help me with the design. Simply put, one of the biggest challenges of visualizing chart types is that there are just a lot of ways to visualize data. Take a column chart, for example, and bend it into a circle and you have a donut chart. Fill that in and you have a pie chart. Blow it up in all different directions—a nightingale. But how do you create a visualization of an inherently nonlinear, complex system of graphic types? Others have tried to create a classification system for graphic types (herehere, and maybe even here),but, and perhaps by necessity, each simplifies the number of graph and data types. The first drafts of The Graphic Continuum were laid out in a grid with a single dot in the top-left. But this layout lacked direction and a story and instead was a clutter of graphs; it wasn’t clear that people should start in the top-left graph and then make their way—some way, any way—through it. It was just a cluttered collection of graphs. In an attempt to better organize the space, we started dividing graphs into different categories: Comparing Categories, Distribution, Geospatial, Part-to-Whole, Relationships, and Time.  But this was no simple feat as there are functional overlaps between many chart types; for example, you can plot time series data as a line chart or as a column chart. We collected as many examples as we could and drew on a variety of resources to categorize the graphs based on what seemed to be the primary function of the chart. For example, time series data are primarily visualized using line charts and data that compare categories are primarily visualized using column charts. image 1 We then had an organizing principle in mind, but we were far from where we needed to be. A few other edits, layouts, and approaches didn’t get us any further: we added then subtracted a side bar that showed basic visual encodings; we tried coloring the groups based on these encodings instead of our five groups; we added then subtracted links within and across groups; we rotated the space vertically; and we played around with different colors, fonts, and title bars. image 4 Feeling stuck, we went analog. I hand-colored and cut out each graph in the original, grid version. Playing around in this way, I tried a circle layout, again relying on the basic five groups. We tried this in the electronic version, but it didn’t feel like a great use of the space. We felt we could do better than the circle layout. IMG_1781 image 6 After more toying around (and with helpful advice from some friends), we came to the current version. Using our thought process from the original format, we grouped the different graphs into our five categories and—this was crucial—we separated them across the layout. We could now add linking annotations to show connections between chart types. We went back and forth several times to figure out how many and which links to include, and what they should look like; we— then cut some graphs and text, added in some more, and iterated back and forth.

The Graphic Continuum



We believe The Graphic Continuum is a more comprehensive view of graphic types and how they can be classified into different categories. We hope you can use it as a tool to help decide how to choose the best graph for your data or to expose people to less common graphic types. Or, perhaps you can just use it as a piece of art in your home or office. Either way, we hope you find it useful and beautiful.   Jon Schwabish is an economist and data visualization creator. You can reach him at [email protected], at his website, or by following him on Twitter @jschwabish.

Guide to visually communicate complex data - Promotional Banner

Want to get more content like this for free?

Sign up to receive our content by email.

Email registered successfully
Oops! Invalid email, please check if the email is correct.