There are hundreds of colorschemes in this package, and they're useful for many different purposes. However, if you're intending to use a colorscheme for communicating features of a scientific dataset, you should choose it with care.
You should choose a perceptually uniform colorscheme: a set of colors arranged so that equal steps in data are perceived by the viewer as equal steps in the color space.
Researchers[Kovesi][ZhouHansen] have found that the human brain perceives changes in the lightness parameter as changes in the data much better than, for example, changes in hue. So sequential colorschemes with monotonically increasing lightness values will be better interpreted by the viewer.
The Lab color space represents a color with three components: Lightness, RedGreen, and YellowBlue. The Lightness parameter can be used to indicate how uniform the colors will be perceived by viewers.
In the following diagrams, the Lightness Lab component of each color step is plotted in
x moves through the colorscheme. You can see how the lightness increases evenly in the recommended schemes.
Good choices include
ColorCET schemes (
findcolorscheme("colorcet") will return the very long names to save you typing them):
Fabio Crameri's Scientific colorschemes:
For diverging colorschemes, the lightness values of the extremes should be broadly equivalent. As well as the
diverging- ColorCET colorschemes, there are suitable schemes in Scientific, ColorBrewer, and others.
Colorschemes with rapid changes in lightness are less suitable, because the viewer's interpretation of a region of data might be influenced by the coloring, rather than by the data values.