There was too much code used in the practice, so code-related content is only in the Jupyter notebook.
Sequential
- Suitable for sequential data
- Represented with continuous colors
Diverging
- Diverges from the center
- Suitable for opposing values (temperature), two different datasets (approval ratings)
- Usually darker colors toward the extremes
- The center color is neutral, not biased toward either side
Example
South Korea average temperature data
HSI
Matplotlib’s color API uses HSI, apparently. It’s “that color space” I learned in computer vision class in college..
- Hue: the color itself. Complementary colors exist — mixing complementary colors produces white. Hue differences are usually the easiest to notice.
- Saturation: the ratio of mixing between white and a pure color. Described as “light” or “vivid.”
- Lightness: brightness. Dark or bright.
Color palette
Modules provide default palettes, but there are various color palettes on GitHub. For professional use, Adobe Color is recommended. https://color.adobe.com/create/color-wheel
RColorBrewer palettes
https://www.datanovia.com/en/blog/top-r-color-palettes-to-know-for-great-data-visualization/
The instructor said these are well-differentiated and look good.