matplotlib: alternating non-overlapping dashes in contour line-CodePudding

I would like to mark one particular contour of a function with alternating and non-overlapping red and green dashes. My first idea was to use the parametrized dash minilanguage, but that gives an error:

import numpy as np
import matplotlib.pyplot as plt

# prepare the data
delta = 0.025
x = np.arange(-1.0, 1.0, delta)
y = np.arange(-1.0, 1.0, delta)
X, Y = np.meshgrid(x, y)
Z1 = np.exp(-(X ** 2) - Y ** 2)
Z2 = np.exp(-((X - 1) ** 2) - (Y - 1) ** 2)
Z = (Z1 - Z2) * 2

# the actual plotting
plt.contour(X, Y, Z, levels=[1.1], linestyles=(0, (5, 5)), colors="red", alpha=0.5)
plt.contour(X, Y, Z, levels=[1.1], linestyles=(5, (5, 5)), colors="green", alpha=0.5)

plt.show()

ValueError: Do not know how to convert [0] to dashes

Any hints?

CodePudding user response：

The [] brackets were missing:

import numpy as np
import matplotlib.pyplot as plt

# prepare the data
delta = 0.025
x = np.arange(-1.0, 1.0, delta)
y = np.arange(-1.0, 1.0, delta)
X, Y = np.meshgrid(x, y)
Z1 = np.exp(-(X ** 2) - Y ** 2)
Z2 = np.exp(-((X - 1) ** 2) - (Y - 1) ** 2)
Z = (Z1 - Z2) * 2

# the actual plotting
plt.contour(X, Y, Z, levels=[1.1], linestyles=[(0, (5, 5))], colors="red")
plt.contour(X, Y, Z, levels=[1.1], linestyles=[(5, (5, 5))], colors="green")

plt.show()

Thanks to @JohanC for the reply!

CodePudding user response：

To expand on Trenton's comment, linestyles in the context of the contour() function is not the same as in regular plots, as noted in the documentation: matplotlib contour():

linestyles : {None, 'solid', 'dashed', 'dashdot', 'dotted'}, optional

Only applies to contour

Because contour is intended to generate a whole group of lines which themselves convey additional information, there is reduced ability to define the appearance of those lines, and so you can only choose from those 5 options (or a list of those options, applying each element to successive contours).