December 29th, 2022
New column charts, visual chart selector and more
Here is an overdue update of what is new. As of April 12, 2013, the public version of Datawrapper is now 1.2.9. Combined with the other enhancements since version 1.2 (released a few weeks earlier) the application has matured in many ways.
Firstly there are more chart types, more options per chart (already since 1.2.8). At the same time we invested (a lot) of work into the backend. Most of that will never been seen by anyone outside development. For example, Gregor Aisch created an incredible browser testing system, to check whether charts are shown properly on all kinds of OS/browser combinations, including the static images we use as fallbacks for older browsers.
In other enhancements we can now monitor whether the application runs smoothly or whether there are hic-ups. Given the growing number of created charts per day, that is important to understand whether users succeed or not and how we can assist. Finally, the publishing process, where the charts are shuffled to a server able to handle really high traffic, is now much smoother and quicker for the user.
Our overall goal in the project is unchanged: To build an application opening up data visualization for journalists and newsrooms. We see Datawrapper as a tool that does not stand in your way, if you need to publish on a deadline. Once you have the data, checked it and so on. If you know how to built charts, Datawrapper is super-simple. For all others working with this tool can help to dig into data more often and build experience.
Basically it is a maintenance release that fixed a few quirks of Datawrapper, namely:
Already available since Datawrapper 1.2.8 are these really cool chart variations. That includes some meaningful options for animated chart transitions, too.
Does not sound like a big step, but this new chart type provides a lot of new options to visualize.
Here are actual examples of what you can do now:
Simple column charts
Grouped column charts
Stacked column charts
100% stacked column charts
You will need a set of data with several rows/columns. Choose the chart type from the selector and then make sure to explore all the options in the context menue on the lower right of the page. Datawrapper by design encourages to check all possible variations of charts. You can go back and forth between the steps without loosing data. Choose the one that conveys the story best.
When to use stacked column charts? Here is a link to a background post written by Gregor Aisch, published at “School of Data”, why stacked columns are sometimes the best option to reveal the story from a set of data. applied to elections in Germany.
Wait, there is more:
This optimizes the labeling on the axis and gives you two options: Depending on which expert you follow, there are different suggestions for the perfect steps between labels on the axis:
Edward Tufte says, that it is helpful to let the y-axis end with the highest value, indicating the range of values for the user.
Donna M. Wong, author of “The Wall Street Journal Guide to Information Graphics” advices that the labeling should be in “natural steps”, e.g. 10, 20, 30, 40 and so on. Now you can choose what you like best. The feature is reachable in the context menue, again on the lower right, depending on the chart you have chosen.
The visual chart selector makes it very easy to see all available chart types (as of now) and select the one that might be most appropriate. Datawrapper draws a lot of it’s strength by being simple to use, yet helping you to create correct, clear charts. If you are very experienced with charting you will plan ahead before even throwing data into the tool – so the chart selector is just a good, handy overview. Many of our users though simple take a table of data and see what might make sense as a visualization. With Datawrapper both approaches work (still, the better you prepare your charts the easier it gets). What is great about the chart selector is that you can play around with many variations.
Note: We assume that some users might be confused by the option to “transpose” the data as this sometimes results in funny looking charts. Actually, our pro users love this feature. Because Datawrapper makes it very easy and quick to “flip” the data around, many spreadsheet programs can do that to, but only by taking a number of steps, not just one click. This feature can be accessed in two spots of the Datawrapper process: Firstly in step 2, where you can define whether you want your data to be visualized based on the rows or the columns. In the visualize step you have a second option to change this back. If you are an expert, it will be very clear what Datawrapper does here in an easy way. If you are a beginner, just don’t get confused and play around with this feature. You will see that some charts simply don’t make sense, are not clear, etc. So, change it back. No big deal.
You will get the best results out of Datawrapper the better your data has been prepared in a spreadsheet before uploading. Still, you might want to import the exact values and then they the numbers might be a bit to big to be understood. 65.750.820 is an arbitrary example. It’s correct, but who in the world will understand or remember it? With the “multiply/divide” feature you can easily divide such a number by “millions” and have a clearer chart, that says “65 Million”. Or “60,7 Million” if that is an added information in that particular story. It actually depends when it is better to shorten such numbers values or when to display the actual number down to nitty-gritty details.
In short, with this feature, Datawrapper enables you to do both.
Thanks to many of you out there for feedback and support. A big, special thanks to the teams at The Guardian Data Blog, Le Monde, Neue Zürcher Zeitung and Der Standard.
As a result of journalists working with Datawrapper we have seen big leaps in terms of visits to charts created with Datawrapper in the last six month, since we released Version 1.0. In March we had several days with >250.000 visits per day. Our monthly delivered visits (across all domains using Datawrapper) has by now crossed two million per month and it looks like the next million is not far away. For us, this uptake creates a lot of motivation to keep on working.
For the next months we already have a very detailed plan for development sprints. What we try to do is delivering updates in smaller steps. Stick to basic, correct charts. Have quick response times should there be questions. Enhance the quality further. Work with awesome people around the world.
Follow us on Twitter for more frequent updates and links: @Datawrapper