Redrawing a plot in IPython / matplotlib

Question

I use IPython/Matplotlib, and I want to create functions that can plot various graphs in the same plotting window. However, I have trouble with redrawing. This is my program test_plot_simple.py:

from __future__ import division
import numpy as np
import matplotlib.pyplot as plt

x = np.arange(10)
y2 = (x**2)/(10**2)
ye = (2**x)/(2**10)
fig, ax = plt.subplots()

def p_squared():
    ax.plot(x,y2, 'r')
    plt.show()

def p_exp():
    ax.plot(x,ye, 'r')
    plt.show()

I start IPython as $ python --matplotlib

On the IPython command line I do

In [1]: run test_plot_simple.py
In [2]: p_squared()
In [3]: p_exp()

After the second line, the squared graph is shown. But nothing happens after the second. Why is the plt.show() not working here?

cosmosis · Accepted Answer · 2014-11-06 16:32:11Z

1

It appears as though you call subplots without actually taking advantage of them, namely that you are trying to over plot on the same canvas. See here for a more thorough explanation. That being said, all you need is the following in order to have the functionality I think you want:

from __future__ import division
import numpy as np
import matplotlib.pyplot as plt

x = np.arange(10)
y2 = (x**2)/(10**2)
ye = (2**x)/(2**10)

def p_squared():
    plt.plot(x,y2, 'r')
    plt.show()

def p_exp():
    plt.plot(x,ye, 'r')
    plt.show()

Now both the p_squared() and p_exp() calls produce plots. Hope this helps.

answered Nov 6, 2014 at 16:32

cosmosis

6,3273 gold badges35 silver badges28 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

ukrutt Over a year ago

This helps in the simple example. However, for other reasons I need to go with an object-oriented approach where I can keep track of the axis from the very start. Thus I do indeed need to do the plotting via ax.plot().

ukrutt · Accepted Answer · 2014-11-06 16:50:48Z

After some digging I think I found the right way to go about this. It seems that show() is not really intended for this purpose, but rather draw() is. And if I want to keep it object-oriented, I should draw via my figure or my axis. It seems to me that something like this is the best approach:

from __future__ import division
import numpy as np
import matplotlib.pyplot as plt

x = np.arange(10)
y2 = (x**2)/(10**2)
ye = (2**x)/(2**10)
fig, ax = plt.subplots()
fig.show()

def p_squared():
    ax.plot(x,y2, 'r')
    fig.canvas.draw()

def p_exp():
    ax.plot(x,ye, 'r')
    fig.canvas.draw()

I.e., use fig.canvas.draw() in lieu of plt.show() (or fig.show(), for that matter.)

I still need one show() - I chose to do that right away after the figure has been created.

Collectives™ on Stack Overflow

Redrawing a plot in IPython / matplotlib

2 Answers 2

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related