question-mark
Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Matplotlib Figures not cleaned up during loop

See original GitHub issue

It seems like while generating plots in a loop, the memory grows over time when running the script in spyder.

I have a funny use-case where I’m making a lot of large plots, and then when I’m done I like to interactively look at the last plot that was generated.

To reproduce using this little script:

import sys
import gc
import numpy as np
import matplotlib
matplotlib.use('Qt4Agg')
import matplotlib.pyplot as plt

arr = np.random.rand(10000, 10000)

def memplot_plot(arr):
    plt.matshow(arr)    
    plt.close()

print '*' * 80
for i in range(1, 10):
    print 'i : ' + str(i)    
    print  len( gc.get_objects())
    print sys.getsizeof(gc.get_objects())
    memplot_plot(arr)    
    gc.collect()
print '*' * 80

From the command-line

*****************************************
i : 1
61424
514568
i : 2
61710
514568
i : 3
61713
514568
i : 4
61716
514568
i : 5
61719
514568
i : 6
61722
514568
i : 7
61725
514568
i : 8
61728
514568
i : 9
61731
514568
*****************************************

From spyder, running in a dedicated terminal with interaction afterwards

********************************************************************************
i : 1
64416
578936
i : 2
67685
578936
i : 3
70737
578936
i : 4
73789
651352
i : 5
76841
651352
i : 6
79893
651352
i : 7
82945
732816
i : 8
85997
732816
i : 9
89049
732816
********************************************************************************

And for fun, my windows task manager. The four small blips are four runs from the commandline (python, python -i, ipython and ipython -i). The last one is from Spyder. And this is not clean because I’m also working in the background. mpl_memory_leak

matplotlib.get_backend()
Out[2]: u'Qt4Agg'

matplotlib.__version__
Out[3]: '1.4.3'

Let me know if you need any more information. Same behavior was observed on a colleague’s machine.

Issue Analytics

  • State:closed
  • Created 8 years ago
  • Comments:8 (6 by maintainers)

github_iconTop GitHub Comments

1reaction
goanpecacommented, Nov 28, 2019

So, should we close?

0reactions
ccordoba12commented, Nov 28, 2019

This is very old, let’s close it.

Read more comments on GitHub >

github_iconTop Results From Across the Web

Matplotlib runs out of memory when plotting in a loop
From @geographica's answer it seems that clf() clears the figure, but doesn't release all references to it, unlike close(fig) . It won't hold ......
Read more >
How To Clear A Plot In Python (without closing the window)
Click to clear a plot using two methods - clearing Pyplot Figure or clearning PyPlot axes, without closing the plot window.
Read more >
Interactive figures and asynchronous programming - Matplotlib
Once the event loop is stopped (leaving any still open figure windows non-responsive) you will be able to use the prompt again. Re-starting...
Read more >
How to update a plot on same figure during the loop?
Here, figure.canvas.flush_events() is used to clear the old figure before plotting the updated figure. Real time plotting with Matplotlib.
Read more >
How to clear the memory completely of all Matplotlib plots?
Using the following methods, we can clear the memory occupied by Matplotlib plots. plt.figure() - Create a new figure or activate an ...
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Dev.to Post

No results found

github_iconTop Related Hashnode Post

No results found