PyPlot ColorMesh creates plot with wrong color mapping

Question

I try to create a Color Mesh plot using a map of xy-coordinates and colors in this way:

from matplotlib.colors import ListedColormap
import numpy as np
%pylab inline

colors = ListedColormap(['red', 'blue', 'yellow'])
xx,yy = np.meshgrid(np.arange(1, 6, 1), np.arange(1, 6, 1))
zz = np.array([[1,1,1,1,1],
               [1,1,1,1,1],
               [1,1,1,1,1],
               [2,2,0,0,0],
               [2,2,0,0,0]])
pyplot.pcolormesh(xx, yy, zz, cmap = colors)

It works ok when there is a list of three colors and I try to map xy-points into all of that colors (like in the code above):

But when there is a list of three colors, and I try to map points only into two of them, mapping goes wrong:

zz = np.array([[1,1,1,1,1],
               [1,1,1,1,1],
               [1,1,1,1,1],
               [0,0,0,0,0],
               [0,0,0,0,0]])
pyplot.pcolormesh(xx, yy, zz, cmap = colors)

It should map into colors 0 (red) and 1 (blue), but I get plot with 0 (red) and 2 (yellow) colors. Where is the mistake?

Answer 1

Colormaps are normalized between 0 and 1. When they are used in a plot, the normalization is transfered to the minimum and maximum value of the data. If 0 and 1 are minimum and maximum, 0 will be the first color of the map (red) and 1 will the last (yellow).

What you need is a normalization which takes the desired colormap behaviour into account. The easiest option is to use vmin and vmax

zz = np.array([[1,1,1,1,1],
               [1,1,1,1,1],
               [1,1,1,1,1],
               [0,0,0,0,0],
               [0,0,0,0,0]])
plt.pcolormesh(xx, yy, zz, cmap = colors, vmin=0,vmax=colors.N)