Customize Matplotlib Colormap

0
votes

I want to construct a colormap that mainly consists of three colors, similar to the colormap 'brg': enter image description here

However, I'd like to modify it in two aspects:

  1. I'd like to replace blue, red and green by blue, black and red, respectively.
  2. I'd like to extend the area of the middle color. As a norm I am currently using matplotlibs 'Normalize', I guess I could choose a non linear normalization, but I'd rather just 'add red (black)' to the middle of the colormap and stick with the linear normalization.
1
pythonpython-3.xmatplotlibcolorscolormap
Thanks JohanC, that did the trick. Go ahead and put it as an answer, so that I can mark it if you want – Pickniclas

1 Answers

1
votes

The simplest way to create such a colormap is via LinearSegmentedColormap.from_list('', ['b', 'k', 'k', 'r']). This divides the colors evenly, so pure black occupying 1/3rd of the range.

The function also allows to position the colors, via tuples of the form (position, color) with position ranging from 0 at the left (bottom) to 1 at the right (top). Below is an example where black occupies 10% of the range.

import matplotlib.pyplot as plt
from matplotlib.colors import LinearSegmentedColormap
from matplotlib.cm import ScalarMappable
import numpy as np

cmap1 = LinearSegmentedColormap.from_list('', ['b', 'k', 'k', 'r'])
plt.colorbar(ScalarMappable(cmap=cmap1), orientation='horizontal',
             ticks=np.linspace(0,1,11), label='pure black occupies 1/3rd')
width = 0.1
cmap2 = LinearSegmentedColormap.from_list('', [(0, 'b'), ((1 - width) / 2, 'k'), ( (1 + width) / 2, 'k'), (1, 'r')])
plt.colorbar(ScalarMappable(cmap=cmap2), orientation='horizontal',
             ticks=np.linspace(0,1,11), label=f'pure black occupies {width}')
plt.show()

illustrative plot