Reviewed October 2014
Sarah Seymore, Digital Metadata Technician
Digital Scholarship Center, University of Oregon Libraries
University of Oregon
With information visualizations continuing to gain interest as memorable, data-instilled graphics—and even as works of art—librarians have the opportunity to support services that help foster the creation and display of data visualizations. The online, open source information visualization tool RAW, by DensityDesign Lab, offers a simplified process for creating graphs, charts, and other popular visualizations without having to write or edit code. RAW is particularly useful for novices because of its ease-of-use and customization. For art students or art historians, RAW can serve as an experimental portal to more advanced designs or as a tool to create personalized infographics.
RAW’s user-friendly interface for the popular data visualization library Data-Driven Documents has four straightforward steps for adding data, choosing a design template, and editing the information parameters and visual layout. Data is uploaded to RAW in a spreadsheet or plain text file with delimiter-separated values. It can also be typed in the input box at the top of the application’s homepage as long as it contains a set of delimiters such as commas, colons, semicolons, or tabs.
There are fifteen visualization styles for users to select from or, alternatively, they can add their own. Templates include short descriptions suggesting the types of data that are most suitable for each type of layout. The limitations for different types of data are evident with most charts requiring numerical values. For text-only, hierarchical information, it is best to use dendrogram, or tree diagram, charts. RAW suggests other websites to generate pie charts, line graphs, and histograms since its templates are more complex to recreate.
“Map your dimensions” is where the information is organized and when users might realize that their data is not compliant with a particular design. This step allows users to revise seamlessly with the visualization automatically updating as the data mappings are being manipulated. RAW’s categories list what data types can be used in each field, and it is easy to experiment with the categories to determine which fields are appropriate matches given the data type.
A visualization’s design elements can be manipulated online and after downloading. The customization is different for each template but ranges from width, height, color, padding of the text, and other spacing and sizing options. The completed visualization can be embedded into a website with the generated HTML, downloaded as a JSON model, or saved as a .png or .svg file, which are customizable with graphics editing software like GIMP, Inkscape, or Adobe Creative Suite.
RAW allows all visualizations to be published as long as they are properly credited. In order to protect users’ privacy, RAW does not store or save data so navigating away from the website while generating a visualization will result in the loss of any work up to that point. However, the streamlined interface and customizable features make design sessions brief with little required troubleshooting. If needed, documentation is on GitHub with sub-pages for advanced programmers who want to contribute their templates or users who are having trouble formatting their data. The FAQ section is helpful for beginners and for troubleshooting common errors.
Tools like RAW allow for experimentation with data visualizations without devoting a lot of time to design or learning coding languages. RAW is useful for a wide audience from researchers publishing their own data, students comparing information for a project, librarians showcasing their statistics, artists creating designs, or anyone else who wants to make their data visually compelling.