matplotlib colorbar limits for contourf

Question

How do you set hard limits on colorbar used with contourf? The code below works as expected with colorbar limits set to [-3, 3] when using plot_surface, but with contourf the limits are not at the tips of the colorbar. I need this to create a gif from multiple images with a constant colorbar.

import matplotlib.pyplot as plt 
import numpy as np

fig = plt.figure()

ax = fig.gca(projection='3d')
CHI = np.linspace(-45, 45, 35);
M = np.linspace(0, 1, 35) 
CHI, M = np.meshgrid(CHI, M)
R = 10*2*M*np.sin(  2 * np.deg2rad(CHI) )

cont = ax.contourf(CHI, M, R)
#cont = ax.plot_surface(CHI, M, R)
cont.set_clim(vmin=-3, vmax=3)

ax.set_xlim(-45,45)
cbar = plt.colorbar(cont, ticks=[-3,-2,-1,0,1,2,3])
plt.show()

Answer 1

You could pass levels parameter to matplotlib.pyplot.contourf in order to specify the number and positions of the contour regions. Then you can set extend = 'both' in order to draw the countour regions outside levels range you used:

import matplotlib.pyplot as plt
import numpy as np

fig = plt.figure()

ax = fig.gca(projection='3d')
CHI = np.linspace(-45, 45, 35);
M = np.linspace(0, 1, 35)
CHI, M = np.meshgrid(CHI, M)
R = 10*2*M*np.sin(  2 * np.deg2rad(CHI) )

levels = [-3, -2, -1, 0, 1, 2, 3]

cont = ax.contourf(CHI, M, R, levels = levels, extend = 'both')

ax.set_xlim(-45,45)
cbar = plt.colorbar(cont)
plt.show()