Python: Update plot instead of creating a new one

Question

I am trying to code the Ising Model using Python.

I think, I've coded it correctly, but I have a problem with the animation or the plotting. I seem to plot a new image of every configuration, instead of updating the existing one, resulting in a lot of saved images that I don't need. I just want a single plot that is being updated, if that's possible.

I know, I am plotting inside the loop, but I don't recall that being an issue, when I want to plot every iteration. Can it be a problem with Seaborn's heatmap?

I've attached my code:

import numpy as np
import numpy.random as npr
import matplotlib.pyplot as plt
import seaborn as sns

#Constants
J = 1
h = 1
kbT = 1
beta = 1

#Grid
L = 20 #Dimensions
N = L**2 #Total number of grid points


#Initial configuration
spins = 2*np.random.randint(2, size = (L,L))-1


E = []
i = 0
plt.figure()

while i < 100000:
    for i in range(1,N):
        i += 1
        s = tuple(npr.randint(0, L, 2)) # Random initial coordinate
    
        # x and y coordinate
        (sx, sy) = s
        # Periodic boundary condition
        sl = (sx-1, sy) 
        sr = ((sx+1)%L, sy)
        sb = (sx, sy-1)
        st = (sx, (sy+1)%L)
        # Energy
        E =   spins[s] * ( spins[sl] + spins[sr] + spins[sb] + spins[st] )
        if E <= 0 : # If negative, flip
            spins[s] *= -1
        else:
            x = np.exp(-E/kbT) # If positve, check condition
            q = npr.rand()
            if x > q: 
                spins[s] *= -1
    # Plot (heatmap)
    sns.heatmap(spins, cmap = 'magma')
    plt.pause(10e-10)
    plt.draw()
    plt.show()

I think you are looking for the function `ion`, you can use it as `plt.ion()`, take a look [here](https://stackoverflow.com/questions/12822762/pylab-ion-in-python-2-matplotlib-1-1-1-and-updating-of-the-plot-while-the-pro/12826273) — Carlo Zanocco, Sep 05 '20 at 12:06

score 1 · Accepted Answer · answered Sep 05 '20 at 12:20

I think the function ion and clf could do the trick.

import numpy as np
import numpy.random as npr
import matplotlib.pyplot as plt
import seaborn as sns

#Constants
J = 1
h = 1
kbT = 1
beta = 1

#Grid
L = 20 #Dimensions
N = L**2 #Total number of grid points

#Initial configuration
spins = 2*np.random.randint(2, size = (L,L))-1

E = []
i = 0

plt.ion()
plt.figure()
plt.show()

while i < 100000:
    for i in range(1,N):
        i += 1
        s = tuple(npr.randint(0, L, 2)) # Random initial coordinate
    
        # x and y coordinate
        (sx, sy) = s
        # Periodic boundary condition
        sl = (sx-1, sy) 
        sr = ((sx+1)%L, sy)
        sb = (sx, sy-1)
        st = (sx, (sy+1)%L)
        # Energy
        E =   spins[s] * ( spins[sl] + spins[sr] + spins[sb] + spins[st] )
        if E <= 0 : # If negative, flip
            spins[s] *= -1
        else:
            x = np.exp(-E/kbT) # If positve, check condition
            q = npr.rand()
            if x > q: 
                spins[s] *= -1
    # Plot (heatmap)
    plt.clf()
    sns.heatmap(spins, cmap = 'magma')
    plt.pause(10e-10)

With the function ion you are making interactive the plot, so you need to:

Make it interactive
Show the plot
Clear the plot in your cycle

Here the reference for the ion function.

Reference for clf is here

That did the trick! Thank you. Now I only get one "
", instead of a lot. — JohnGoodWill, Sep 05 '20 at 12:29

Python: Update plot instead of creating a new one

1 Answers1