How to animate a seaborn's heatmap or correlation matrix?

Question

I am relatively new to python (coming from Matlab). As one project, I am trying to create an animated plot of a correlation matrix over time. To make the plots nice, I am trying seaborn. I struggled to get the animation done at all (having issues with Matplotlib backend on a mac), but a very basic animation now works using this code from the web:

import numpy as np
from matplotlib import pyplot as plt
from matplotlib import animation

nx = 50
ny = 50

fig = plt.figure()
data = np.random.rand(nx, ny)
im = plt.imshow(data)

def init():
    im.set_data(np.zeros((nx, ny)))

def animate(i):
    #xi = i // ny
    #yi = i % ny
    data = np.random.rand(nx, ny)
    im.set_data(data)
    return im

anim = animation.FuncAnimation(fig, animate, init_func=init, frames=50, repeat = False)

Now, I was trying to adapt this to seaborn, but did not succeed. It seems that seaborn works on subplots and to animate these was far harder. The best thing I got once was a kind of recursive plot, where seaborn.heatmaps were plotted on top of each other. Also, the im.set_data method was not available.

Any suggestions are highly appreciated.

Answer 1

I replaced plt.imshow (casting data via set_data didn't work) with seaborn.heatmap.

import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from matplotlib import animation

fig = plt.figure()
data = np.random.rand(10, 10)
sns.heatmap(data, vmax=.8, square=True)

def init():
      sns.heatmap(np.zeros((10, 10)), vmax=.8, square=True, cbar=False)

def animate(i):
    data = np.random.rand(10, 10)
    sns.heatmap(data, vmax=.8, square=True, cbar=False)

anim = animation.FuncAnimation(fig, animate, init_func=init, frames=20, repeat = False)

This creates the recursive plot I struggled with.