How to animate the colorbar in matplotlib

While I’m not sure how to do this specifically using an ArtistAnimation, using a FuncAnimation is fairly straightforward. If I make the following modifications to your “naive” version 1 it works.

Modified Version 1

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.animation as animation
from mpl_toolkits.axes_grid1 import make_axes_locatable

fig = plt.figure()
ax = fig.add_subplot(111)

# I like to position my colorbars this way, but you don't have to
div = make_axes_locatable(ax)
cax = div.append_axes('right', '5%', '5%')

def f(x, y):
    return np.exp(x) + np.sin(y)

x = np.linspace(0, 1, 120)
y = np.linspace(0, 2 * np.pi, 100).reshape(-1, 1)

frames = []
for i in range(10):
    x       += 1
    curVals  = f(x, y)
    frames.append(curVals)

cv0 = frames[0]
cf = ax.contourf(cv0, 200)
cb = fig.colorbar(cf, cax=cax)
tx = ax.set_title('Frame 0')

def animate(i):
    arr = frames[i]
    vmax     = np.max(arr)
    vmin     = np.min(arr)
    levels   = np.linspace(vmin, vmax, 200, endpoint = True)
    cf = ax.contourf(arr, vmax=vmax, vmin=vmin, levels=levels)
    cax.cla()
    fig.colorbar(cf, cax=cax)
    tx.set_text('Frame {0}'.format(i))

ani = animation.FuncAnimation(fig, animate, frames=10)

plt.show()

The main difference is that I do the levels calculations and contouring in a function instead of creating a list of artists. The colorbar works because you can clear the axes from the previous frame and redo it every frame.

Doing this redo is necessary when using contour or contourf, because you can’t just dynamically change the data. However, as you have plotted so many contour levels and the result looks smooth, I think you may be better off using imshow instead – it means you can actually just use the same artist and change the data, and the colorbar updates itself automatically. It’s also much faster!

Better Version

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.animation as animation
from mpl_toolkits.axes_grid1 import make_axes_locatable

fig = plt.figure()
ax = fig.add_subplot(111)

# I like to position my colorbars this way, but you don't have to
div = make_axes_locatable(ax)
cax = div.append_axes('right', '5%', '5%')

def f(x, y):
    return np.exp(x) + np.sin(y)

x = np.linspace(0, 1, 120)
y = np.linspace(0, 2 * np.pi, 100).reshape(-1, 1)

# This is now a list of arrays rather than a list of artists
frames = []
for i in range(10):
    x       += 1
    curVals  = f(x, y)
    frames.append(curVals)

cv0 = frames[0]
im = ax.imshow(cv0, origin='lower') # Here make an AxesImage rather than contour
cb = fig.colorbar(im, cax=cax)
tx = ax.set_title('Frame 0')

def animate(i):
    arr = frames[i]
    vmax     = np.max(arr)
    vmin     = np.min(arr)
    im.set_data(arr)
    im.set_clim(vmin, vmax)
    tx.set_text('Frame {0}'.format(i))
    # In this version you don't have to do anything to the colorbar,
    # it updates itself when the mappable it watches (im) changes

ani = animation.FuncAnimation(fig, animate, frames=10)

plt.show()