Is it possible to control which colors are retrieved from matplotlib colormap?

user_3078502 0
Viewed 3102

I want to assign colors in a plot to nominal data represented by integers. I want to draw colors from a Qualitative colormap, specifically I want to draw five colors from Set3:

Matplotlib built-in Qualitative colormaps

The problem is that I want to use the first five colors, but the colormapper normalizes my data, which ranges from 1 to 5 for the five categorical values, and selects the 1st, 4th, 7th, 10th, and 12th colors from the 12-color set.

Basically, matplotlib.cm.get_cmap allows you to specify a number of colors, but normalizes across the range:

from matplotlib import cm
set3_5 = cm.get_cmap("Set3", lut = 5)

I want something like matplotlib.colors.ListedColormap which has a parameter N which truncates the color list after N items without normalizing. But I can't figure out how to pass a built-in colormap to ListedColormap.

matplotlib colors python
1 Answers

With cm.get_cmap("Set3").colors you get a list of the 12 colors in the colormap. This list can be sliced to get specific colors. It can be used as input for the ListedColormap.

Note that with a sequential colormap such as viridis, the list has 256 colors. You can get an evenly spaced subset with cm.get_cmap("viridis", 8).colors and then again take a slice, for example if you don't want to use the too bright colors.

Here is an example:

import matplotlib.pyplot as plt
import matplotlib
import numpy as np

cmap = matplotlib.colors.ListedColormap(matplotlib.cm.get_cmap("Set3").colors[:5])
plt.scatter(np.random.uniform(0, 10, 50), np.random.uniform(0, 10, 50), c=np.random.uniform(0, 10, 50), cmap=cmap)
plt.colorbar()
plt.show()

example

user_120464090
Related