René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"[Bug]: Megabyte-level memory leak when using imshow() in a loop","articleBody":"### Bug summary\n\nWhen creating, plotting to, and closing a figure in a scope (I tested with this in a function called in a loop), the memory usage steadily increases, sometimes by several megabytes for even simple calls to imshow.\n\n(p.s. in the code below, the generate_frame function gets called 3 times to \"warm up\" the Python interpreter, just in case. same results with or without those lines, though)\n\n### Code for reproduction\n\n```Python\nimport gc\nimport numpy as np\nimport os\nimport psutil\nimport matplotlib\nimport matplotlib.pyplot as plt\n\nmatplotlib.use('Agg')\n\ndef generate_frame(frame_idx: int):\n    plt.style.use(\"dark_background\")\n    fig, axes = plt.subplots(2, 3, figsize=(16, 8))\n    axes = axes.flatten()\n\n    for i in range(6):\n        data = np.random.rand(1_000, 2_000)\n        axes[i].imshow(data)\n\n    plt.savefig(\"test.png\", dpi=400, bbox_inches='tight', pad_inches=0.1)\n\n    plt.clf()\n    plt.close('all')\n    del fig\n    gc.collect()\n    return None\n\ndef get_memory():\n    return psutil.Process(os.getpid()).memory_info().rss / (1024**2)\n\nfor _ in range(3):\n    generate_frame(-1)\n\nprev = get_memory()\nfor i in range(0, 20):\n    generate_frame(i)\n    current = get_memory()\n    print(f\"Generated frame {i + 1}. Delta: {current - prev:.4f} MB of memory.\")\n    prev = current\n```\n\n### Actual outcome\n\n```\nGenerated frame 1. Delta: 4.1602 MB of memory.\nGenerated frame 2. Delta: 12.0195 MB of memory.\nGenerated frame 3. Delta: 0.2266 MB of memory.\nGenerated frame 4. Delta: 0.0859 MB of memory.\nGenerated frame 5. Delta: 4.1641 MB of memory.\nGenerated frame 6. Delta: 4.1758 MB of memory.\nGenerated frame 7. Delta: 0.0781 MB of memory.\nGenerated frame 8. Delta: 0.0430 MB of memory.\nGenerated frame 9. Delta: 0.0156 MB of memory.\nGenerated frame 10. Delta: 0.0117 MB of memory.\nGenerated frame 11. Delta: 0.1289 MB of memory.\nGenerated frame 12. Delta: 0.0117 MB of memory.\nGenerated frame 13. Delta: 0.3750 MB of memory.\nGenerated frame 14. Delta: 4.0859 MB of memory.\nGenerated frame 15. Delta: 4.1172 MB of memory.\nGenerated frame 16. Delta: 0.0234 MB of memory.\nGenerated frame 17. Delta: 0.0195 MB of memory.\nGenerated frame 18. Delta: 0.0195 MB of memory.\nGenerated frame 19. Delta: 0.0234 MB of memory.\nGenerated frame 20. Delta: 0.0234 MB of memory.\n```\n\n### Expected outcome\n\nUsing the same code but with no reference to matplotlib, gives:\n```\nGenerated frame 1. Delta: 4.0000 B of memory.\nGenerated frame 2. Delta: 12.0000 B of memory.\nGenerated frame 3. Delta: 176.0000 B of memory.\nGenerated frame 4. Delta: 504.0000 B of memory.\nGenerated frame 5. Delta: 164.0000 B of memory.\nGenerated frame 6. Delta: 0.0000 B of memory.\nGenerated frame 7. Delta: 160.0000 B of memory.\nGenerated frame 8. Delta: 0.0000 B of memory.\nGenerated frame 9. Delta: 28.0000 B of memory.\nGenerated frame 10. Delta: 16.0000 B of memory.\nGenerated frame 11. Delta: 0.0000 B of memory.\nGenerated frame 12. Delta: 24.0000 B of memory.\nGenerated frame 13. Delta: 24.0000 B of memory.\nGenerated frame 14. Delta: 0.0000 B of memory.\nGenerated frame 15. Delta: 0.0000 B of memory.\nGenerated frame 16. Delta: 0.0000 B of memory.\nGenerated frame 17. Delta: 0.0000 B of memory.\nGenerated frame 18. Delta: 0.0000 B of memory.\nGenerated frame 19. Delta: 0.0000 B of memory.\nGenerated frame 20. Delta: 0.0000 B of memory.\n```\n\nCode for that output:\n```py\nimport gc\nimport numpy as np\nimport os\nimport psutil\n\ndef generate_frame(frame_idx: int):\n    for i in range(6):\n        data = np.random.rand(100, 200)\n\n    gc.collect()\n    return None\n\ndef get_memory():\n    return psutil.Process(os.getpid()).memory_info().rss / (1024)\n\nfor _ in range(3):\n    generate_frame(-1)\n\nprev = get_memory()\nfor i in range(0, 20):\n    generate_frame(i)\n    current = get_memory()\n    print(f\"Generated frame {i + 1}. Delta: {current - prev:.4f} B of memory.\")\n    prev = current\n```\n\nIn other words, matplotlib should properly clean up references/resources to not generate a consistent upward trend in memory usage.\n\n### Additional information\n\nIn my real use-case, code very similar to this (but with colorbars. additional axes, etc.), using larger data, generates nearly 0.2 GB per frame as a consistent trend. I'm pretty sure the issue comes from matplotlib, as demonstrated with the example above.\n\n### Operating system\n\nMacOS\n\n### Matplotlib Version\n\n3.10.3\n\n### Matplotlib Backend\n\nAgg\n\n### Python version\n\n3.12.10\n\n### Jupyter version\n\nN/A\n\n### Installation\n\npip","author":{"url":"https://github.com/borisnezlobin","@type":"Person","name":"borisnezlobin"},"datePublished":"2025-08-06T23:23:07.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":2},"url":"https://github.com/30400/matplotlib/issues/30400"}