Course contentsShow
← Data Science
Lesson 1234 of 2,14531. Data Visualization PrinciplesFree lesson

Color: Hue, Saturation, and Luminance

Using color effectively: hue for categories, luminance for ordered data, and avoiding saturation alone.

Color: Hue, Saturation, and Luminance

You'll learn how to use color's three dimensions strategically to encode different types of data accurately.

Understanding Color's Three Properties

Color isn't just "red" or "blue"—it has three distinct dimensions you can manipulate independently:

Hue is what we typically call "color": red, green, blue, purple, etc. It's the position around the color wheel.

Saturation is the intensity or purity of the color, ranging from vivid/vibrant to dull/grayish. Think of it as "colorfulness."

Luminance (or lightness/value) is how bright or dark the color appears, from near-black to near-white.

Using Each Dimension Appropriately

Use hue for categorical data. When you have distinct, unordered categories (like product types or regions), assign different hues. Our eyes easily distinguish red from blue from green without implying any ordering or magnitude.

Use luminance for ordered data. When visualizing quantities, rankings, or sequential data, vary the lightness of a single hue from light to dark. This naturally conveys "less to more" because we perceive lighter-to-darker as a progression. A light blue to dark blue gradient works beautifully for numerical scales.

Avoid relying on saturation alone. Changes in saturation (vivid to pale) are harder to perceive accurately than changes in luminance. Saturation differences can also disappear for colorblind viewers or in grayscale printing. Use saturation as a secondary enhancement, not your primary encoding channel.

Think of it like cooking: hue is your ingredient choice (tomato vs. spinach), luminance is cooking time (raw to fully cooked), and saturation is seasoning intensity (subtle to bold)—but you wouldn't rely only on seasoning intensity to tell ingredients apart!

Key Takeaway: Use hue to distinguish categories, luminance to show ordered/quantitative information, and avoid saturation as a standalone encoding—it's the least perceptually accurate of the three.