In the plot on the right, the orange triangles “pop out”, making it easy to distinguish them from the circles. Consider this simple example: in which of these two plots is it easier to count the number of triangular points?

When you want to represent multiple categories in a plot, you typically should vary the color of the elements. Two colors with different hues will look more distinct when they have more saturation:Īnd lightness corresponds to how much light is emitted (or reflected, for printed colors), ranging from black to white: Vary hue to distinguish categories # Saturation (or chroma) is the colorfulness. It’s property of color that leads to first-order names like “red” and “blue”: Hue is the component that distinguishes “different colors” in a non-technical sense. But for analyzing the perceptual attributes of a color, it’s better to think in terms of hue, saturation, and luminance channels. We usually program colors in a computer by specifying their RGB values, which set the intensity of the red, green, and blue channels in a display.

General principles for using color in plots # Components of color #īecause of the way our eyes work, a particular color can be defined using three components. This chapter discusses both the general principles that should guide your choices and the tools in seaborn that help you quickly find the best solution for a given application. Seaborn makes it easy to use colors that are well-suited to the characteristics of your data and your visualization goals.

Color scatter plot python seaborn