Matplotlib/MulticoloredLine - SciPy wiki dump (original) (raw)

Defining colors manually

colored_line.py is a simple illustration of how to make the color of each segment of a line depend on some property of the data being plotted.

An up to date version of the script can be found here.

colored_line.png

Here is the script:

1 2 ''' 3 Color parts of a line based on its properties, e.g., slope. 4 ''' 5 import numpy as np 6 import matplotlib.pyplot as plt 7 from matplotlib.collections import LineCollection 8 from matplotlib.colors import ListedColormap, BoundaryNorm 9 10 x = np.linspace(0, 3 * np.pi, 500) 11 y = np.sin(x) 12 z = np.cos(0.5 * (x[:-1] + x[1:]))
13 14 15 16 cmap = ListedColormap(['r', 'g', 'b']) 17 norm = BoundaryNorm([-1, -0.5, 0.5, 1], cmap.N) 18 19 20 21 22 23 points = np.array([x, y]).T.reshape(-1, 1, 2) 24 segments = np.concatenate([points[:-1], points[1:]], axis=1) 25 26 27 28 lc = LineCollection(segments, cmap=cmap, norm=norm) 29 lc.set_array(z) 30 lc.set_linewidth(3) 31 plt.gca().add_collection(lc) 32 33 plt.xlim(x.min(), x.max()) 34 plt.ylim(-1.1, 1.1) 35 plt.show()

Note that the number of segments is one less than the number of points.

An alternative strategy would be to generate one segment for each contiguous region of a given color.

Using a smooth, builtin colormap

If you have a parametric curve to display, and want to represent the parameter using color.

colored_line2.png

1 import numpy as np 2 import matplotlib.pyplot as plt 3 from matplotlib.collections import LineCollection 4 5 t = np.linspace(0, 10, 200) 6 x = np.cos(np.pi * t) 7 y = np.sin(t) 8 9 10 11 12 13 points = np.array([x, y]).T.reshape(-1, 1, 2) 14 segments = np.concatenate([points[:-1], points[1:]], axis=1) 15 16 17 18 lc = LineCollection(segments, cmap=plt.get_cmap('copper'), 19 norm=plt.Normalize(0, 10)) 20 lc.set_array(t) 21 lc.set_linewidth(3) 22 23 plt.gca().add_collection(lc) 24 plt.xlim(-1, 1) 25 plt.ylim(-1, 1) 26 plt.show()


CategoryCookbookMatplotlib

SciPy: Cookbook/Matplotlib/MulticoloredLine (last edited 2015-10-24 17:48:23 by anonymous)