If you're going to be attending VIS 2016 and have a chance to watch Connor Gramazio's presentation, two things are pretty much guaranteed: a color palette that looks good, and one whose component colors are easy to tell apart from each other. Connor, a PhD student at Brown University's Department of Computer Science (Brown CS), recently released a new web-based tool, Colorgorical, which immediately made the front page of Hacker News and was featured on FlowingData.
Colorgorical solves one problem faced by almost everyone on the planet and another that presents itself to most people required to visualize information with charts, graphs, and other methods: how do you choose colors that look good together, but don't look too much like each other? It uses semi-random sampling to pick colors based on user-defined scores for functions that include:
- Perceptual Distance (colors that are easily distinguishable from each other),
- Name Difference (colors that have few common names),
- Name Uniqueness (colors whose names are not shared with other colors), and
- Pair Preference (colors that are predicted to be aesthetically pleasing when paired)
Users can also specify hue ranges or build off their own starting palettes. One of the advantages that Colorgorical has over prior approaches is that it allows customization (users can pick the number of colors, adjust the functions above, filter colors, adjust lightness, or even choose starting colors) and takes into account design environments such as Adobe Color and ACE that limit the user's color choices.
You can try Colorgorical for yourself here.
For more information, please click the link that follows to contact Brown CS Communication Outreach Specialist Jesse C. Polhemus.