New Ways to Visualize Your Data

New Ways to Visualize Your Data

  • Comments 3
  • Likes

Posted by Rob Knies

Histogram data visualization in Excel

If you are feeling hungry, you go to the kitchen. If you’d like to take a swim, you head to a swimming pool. If you want to catch a movie, you’re bound for a theater.

And, Danyel Fisher says, if you’re interested in data, you open Excel.

“Excel is where data lives,” says Fisher, a researcher with the Visualization and Interaction for Business and Entertainment (VIBE) team at Microsoft Research Redmond. “When people have data to organize, in any form, it usually passes through Excel at some point—sometimes, just as a quick way to look at it, and sometimes, with tools like Flash Fill and charting and sorting—that’s where it stays.

“Data visualizations are incredibly powerful and fun ways for users to understand their data.”

Excel users—and they are legion—are about to find out. On May 23, Fisher and his colleagues, including Roland Fernandez, principal research software-design engineer, published three new data-visualization tools to Apps for Office, a new channel where people can download free apps for the new Office. Those embeddable apps will let Excel users experiment with new data visualizations using their own data. Everything you need to know about the new tools can be found here.

Offerings from Apps for Office are available at the Office Store—just click “add” to download an app, then start Excel and locate the new app under the Insert tab. Such apps, which can be used in Word, Excel,
Outlook, PowerPoint, and Project, can help with tasks such as translations, connecting with social media, managing finances, or visualizing data. The apps, which can be offered for free or sold, are web-based, written with HTML and JavaScript, and can use web protocols such as OAuth and REST. And the apps are easy to build and publish.

VIBE has been exploring new visualization types as part of its work to develop new ways to display and interact with large amounts of data. The new downloads represent a first wave of this research.

Histogram data visualizationHistogram, which last month became VIBE’s first visualization release, shows the distribution of data via a bar graph of data-point frequency. Such visualizations commonly are used by scientists, and this add-in makes it much easier to create them in Excel.

“Histograms are basic visualizations that many people like to use for basic analytics,” Fisher says. “We wanted to make them easy and lightweight for users.”


A related app, Histogram 2D, examines the distribution of two values with a visualization also known as a density plot.Histogram 2D data visualization

“It gives the frequencies of two different variables,” Fisher says. “That is a really useful analytic technique.”

Treemap data visualizationA third download, Treemap, has an interesting history.

“In 2008,” Fisher recalls, “we released a Treemapper app for Excel, which allowed users to install an Excel add-in to create treemaps. While it was fairly common to have special apps that would allow a user to create treemaps based on disk usage, it was pretty hard to create a treemap based on your own data. We got a lot of positive feedback on that—from teams inside Microsoft and from outside.”

The Treemapper code grew stale over time, though, and had to be retired. The new Treemap app brings such functionality back to Excel, assisting with data organized hierarchically.

Streamgraph data visualizationThe final visualization tool now available for download might be the most eye-catching. Streamgraph examines data volume over time, such as tracking how many people are employed in specific economic sectors over a predetermined span of time.

“Steamgraph is one of the most beautiful visualization techniques to merge in the last decade,” Fisher says. “When you have sliding temporal data like this, it’s a great way to see data that changes over time.”

There is, of course, a research agenda that benefits from the work—beyond the satisfaction of delivering useful, appealing visualizations to Excel users.

“First,” Fisher says, “we’re learning about how people interact with these tools. We’re collecting minimal, anonymous telemetry information to learn what we did right—and what we did wrong. Telemetry can help us find where people are crashing or giving up and what things work for them.

“Second, we’re getting a real-world test of our original idea. We proposed an idea for an ecosystem of visualizations in Excel. It’s valuable for a researcher to find out whether a proposal like that makes sense—and to see how it changes over time.”

In the end, though, while the researchers get what they need to advance their research studies, Excel users get new ways to examine their data.

“There’s been a lot of user-interface design work,” Fisher concludes. “It’s a little less common for users to work with these more complex data visualizations, and we’re trying to learn what seems to work best for people.”

Your comment has been posted.   Close
Thank you, your comment requires moderation so it may take a while to appear.   Close
Leave a Comment
  • In my many years of data analysis and data management I have never seen a more popular and user friendly tool than Excel.

    I'm ready to ramp up and start deploying these tools!

  • Apps for Office is a really nice way for MS Research to add significant value to Office.   Thanks!  

  • Hi Rob, great overview! also check SmartCharts app for excel in office store :) features a bundle of data views & a few other dataviz nuggets ;) very proud of it http://office.microsoft.com/en-us/store/smartcharts-for-excel-WA104120594.aspx Rui Quintino/DevScope team www.devscope.net