Source code has been slightly modified into fully object-oriented matplotlib interface. It now uses the "inferno" colormap, instead of "jet" which produces perceptual glitches.
##########
## Code for the figure
##########
import matplotlib.pyplot as plt
import numpy as np
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.cm import inferno as colormap
from matplotlib.colors import LogNorm
fig = plt.figure()
fig.clf()
ax = Axes3D(fig, azim=-128.0, elev=43.0)
s = 0.05
x = np.arange(-2.0, 2.0 + s, s)
y = np.arange(-1.0, 3.0 + s, s)
X, Y = np.meshgrid(x, y)
Z = (1.0 - X)**2 + 100.0 * (Y - X*X)**2
# Without using `` linewidth=0, edgecolor='none' '', the code may produce a
# graph with wide black edges, which will make the surface look much darker.
ax.plot_surface(X, Y, Z, rstride=1, cstride=1, norm=LogNorm(),
cmap=colormap, linewidth=0, edgecolor='none')
ax.set_xlim([-2, 2])
ax.set_ylim([-1, 3])
ax.set_zlim([0, 2500])
ax.set_xlabel("x")
ax.set_ylabel("y")
fig.savefig("Rosenbrock function.svg")
plt.show()
##########