René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"[MNT]: Avoid using pyplot in tests","articleBody":"### Summary\n\nInspired by #31076: Can we remove the pyplot overhead from most tests?\n\nMany tests create figures. We have 1154 usages of `plt.subplots()` and 441 usages of `plt.figure()` in the tests. So we create around 1500 figures.\n\nAs a rough estimation, based on\n\n```\nIn [1]: import matplotlib.pyplot as plt\n\nIn [2]: from matplotlib.figure import Figure\n\nIn [3]: def bare_figure():\n   ...:     fig = Figure()\n   ...: \n\nIn [4]: def pyplot_figure():\n   ...:     plt.figure()\n   ...:     plt.close()\n   ...: \n\nIn [5]: %timeit bare_figure()\n250 μs ± 2.81 μs per loop (mean ± std. dev. of 7 runs, 1,000 loops each)\n\nIn [6]: %timeit pyplot_figure()\n10.2 ms ± 493 μs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n```\nI infer that the pyplot infrastructure for figure management costs around 10ms. This would mean, we can save around 15sec if we do not use pyplot.\n\nIf our test suite runs ~8min that'd be 3%. This is something but not not too much.\n\nSwitching would mean to replace\n```\nfig = plt.figure()  --\u003e fig = Figure()\nfig, ax = plt.subplots()  --\u003e ax = Figure().add_subplot()  # assuming fig is not used, otherwise, split the commands.\n```\n\nI'm not clear it's worth the churn, but wanted to share the idea (even if it's just to document it and decide not to act on it). What do you think?\n\n### Proposed fix\n\nNone","author":{"url":"https://github.com/timhoffm","@type":"Person","name":"timhoffm"},"datePublished":"2026-02-04T21:55:25.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":7},"url":"https://github.com/31079/matplotlib/issues/31079"}