Color Brewer

01 Jun 2015

by: Emily Halket

The Basics

I want to tell you about color brewer and why you should start using it today!

Essentially, color brewer allows you to create a color palette that conveys an additional layer of meaning when displaying your data.

There are options for sequential, diverging, or qualitative palettes. Within each type of palette, there are a variety of color options.

The sequential palettes vary in lightness. Within a blue sequential palette, high values would shown in a saturated blue while low values would be shown in a light blue.

The diverging palettes emphasize the middle and end values. Intermediate values are assigned colors based on where they fall in relation to the middle point and relative end value. For instance, the mid point might be assigned to white, while the maximum is assigned orange and the minimum is assigned purple. Values above the midpoint will vary between white and orange, with increasing values becoming a more saturated orange.

For both sequential and diverging palettes, the difference between colors represents the magnitude of difference between values.

The qualitative palettes provide colors that are qualitatively different from each other, with no magnitude of difference implied.

My Pitch

I cannot even express how much I love color brewer. I place a high value on manipulating the aesthetics of a visualization in order to convey the most information possible, in the simplest form possible. As human viewers, we have experience based heuristics that allow us to quickly make meaning from visual displays. So it makes perfect sense to take advantage of these heuristics to add meaning to a visualization without having to add verbal explanation. As exemplified by this post, verbal explanation is often cumbersome and inefficient.

In my zeal for color brewer I would maybe even go as far to say that it can be irresponsible to not consider the implications of your color choices when you display data. If you just choose colors because you like them, you might be misleading your viewers into assuming nonexistent relationships between your variables. This might be a stretch, but I’m really just trying to convince you to think before you display!

Truly though, you don’t have to think too much because color theorists have done it for you. All you have to think about is the type of data you are displaying, and one would hope you’ve already thought about that!

Go Check it Out!

So now that you’re fully sold on mindfully choosing a color palette, head over to color brewer to check out your options. If you go through the site, you can even choose colors that are color blind or photocopier safe! If you’re an R user, also check out the package documentation to read how to quickly implement in R.