How to create a donut chart
This Academy article is outdated. It explains how to create the donut charts that we archived on December 14, 2018. We replaced them with shiny new donut charts! We will update this Academy article accordingly in January 2019.
The donut chart is a variation of a pie chart. The main difference is that it displays the total number of all parts in the middle, as an additional information. That's why Donut charts can be used well to show the relative shares of absolute data (that does not add up to 100%).
Pie charts and donut charts are easy to create and easy to understand - people understand these "at a glance". One can easily understand how many parts there are, and how they compare to each other. Very similar to a pie chart a donut chart must not feature more than three or a maximum of five parts. If you have more than three parts always consider to use a bar or column chart.
Below is an example where the simple comparison of the donut chart works - two values tell the story of how close the outcome of the EU Referendum in the UK was.
Preparing and importing the data
If you want to create this chart type, your data needs to be in a certain format. You'll need:
- One header row containing descriptive labels.
- One column containing at least two categories. This will determine the label in the donut slices. In our case, that's "Remain" and "Leave".
- One column containing numeric values. The values in the second column will define the size of the donut slices. In the final chart, Datawrapper turns them into shares of 100% (48% and 52% in our case). The two slices will later be distinguished by colors.
The most important thing you have to keep in mind is that a donut chart always represents a whole, i.e. 100%. Therefore, you can only use data that is based on exclusive values. Making a donut chart of a survey that allows multiple answers will lead to a misleading chart. Use a bar chart instead, but never a donut chart.
That's the data we used to create the chart at the top of the page:
EU Referendum Votes Remain 16141241 Leave 17410742
Source: The Electoral Commission
Once you prepared your data, create a new chart in Datawrapper. You can do so by going to our homepage and clicking on "Create a chart". In Step 1: Upload, copy & paste your dataset, upload it as a .csv or an Excel sheet. After pasting the data above into Datawrapper, it will look like this:
Click "Proceed" at the bottom right to go to the next step:
Check & Describe
In the second step, you can check if your dataset was imported correctly and make changes to it - if necessary. If you did not upload a header row, you have to untick "First row as label" to avoid losing your first row of data. Always remember to do this if you don't have descriptive row and column headers.
In step 2, your data looks like this. You can see that Datawrapper correctly recognized your numbers as numbers (and not as text or dates) because they are colored in blue and are right-alined. To learn more about the Datawrapper's automatic recognition of data formats, visit this article.
To change the number from the very unhandy "16141241" to a better readable "16.14m", click on the column header "B". To the left, you'll see some options. Divide the number by a million, round the numbers to 2 places and append an "m" to the result:
Click on "Proceed" at the bottom left to go to Step 3: Visualize:
In this step, you see a first chart. It's probably a line chart. We want to change that. To do so, click on the "Donut chart" symbol in the grid of available chart types:
You will now see a donut chart, without a title, descriptions or customized colors. Maybe you want to further refine, annotate & define this chart. We cover this in a separate short tutorial found here.