Error raised on mouse cursor hover with matplotlib>=3.10

[nvaytet]

Describe the issue

Using matplotlib>=3.10.*, making any plot with ipympl raises an error when hovering the mouse over the axes.

Example:

import matplotlib.pyplot as plt
%matplotlib widget

fig, ax = plt.subplots()

Hovering the mouse over the plot raises

KeyError                                  Traceback (most recent call last)
File [~/miniforge3/lib/python3.12/site-packages/ipympl/backend_nbagg.py:278](http://localhost:8888/home/nvaytet/miniforge3/lib/python3.12/site-packages/ipympl/backend_nbagg.py#line=277), in Canvas._handle_message(self, object, content, buffers)
    275     self.manager.handle_json(content)
    277 else:
--> 278     self.manager.handle_json(content)

File [~/miniforge3/lib/python3.12/site-packages/matplotlib/backends/backend_webagg_core.py:474](http://localhost:8888/home/nvaytet/miniforge3/lib/python3.12/site-packages/matplotlib/backends/backend_webagg_core.py#line=473), in FigureManagerWebAgg.handle_json(self, content)
    473 def handle_json(self, content):
--> 474     self.canvas.handle_event(content)

File [~/miniforge3/lib/python3.12/site-packages/matplotlib/backends/backend_webagg_core.py:264](http://localhost:8888/home/nvaytet/miniforge3/lib/python3.12/site-packages/matplotlib/backends/backend_webagg_core.py#line=263), in FigureCanvasWebAggCore.handle_event(self, event)
    261 e_type = event['type']
    262 handler = getattr(self, f'handle_{e_type}',
    263                   self.handle_unknown_event)
--> 264 return handler(event)

File [~/miniforge3/lib/python3.12/site-packages/matplotlib/backends/backend_webagg_core.py:295](http://localhost:8888/home/nvaytet/miniforge3/lib/python3.12/site-packages/matplotlib/backends/backend_webagg_core.py#line=294), in FigureCanvasWebAggCore._handle_mouse(self, event)
    286 e_type = event['type']
    287 button = event['button'] + 1  # JS numbers off by 1 compared to mpl.
    288 buttons = {  # JS ordering different compared to mpl.
    289     button for button, mask in [
    290         (MouseButton.LEFT, 1),
    291         (MouseButton.RIGHT, 2),
    292         (MouseButton.MIDDLE, 4),
    293         (MouseButton.BACK, 8),
    294         (MouseButton.FORWARD, 16),
--> 295     ] if event['buttons'] & mask  # State *after* press[/release.](http://localhost:8888/release.)
    296 }
    297 modifiers = event['modifiers']
    298 guiEvent = event.get('guiEvent')

KeyError: 'buttons'

Note: need to set level to 'Debug' in log console to see it.

The figure is basically inactive.

Could be related to #592 but I did install 0.9.7 and it did not fix the issue.

Downgrading to matplotlib=3.9.4 fixes the issue.

Versions

 3.12.7 | packaged by conda-forge | (main, Oct  4 2024, 16:05:46) [GCC 13.3.0]
ipympl version: 0.9.7
Selected Jupyter core packages...
IPython          : 8.30.0
ipykernel        : 6.29.5
ipywidgets       : 8.1.5
jupyter_client   : 8.6.3
jupyter_core     : 5.7.2
jupyter_server   : 2.14.2
jupyterlab       : 4.3.1
nbclient         : 0.10.1
nbconvert        : 7.16.4
nbformat         : 5.10.4
notebook         : not installed
qtconsole        : not installed
traitlets        : 5.14.3

[ianthomas23]

The symptoms are indeed the same as #592, and this has been fixed. I cannot reproduce with the same packages in a new environment:

ipympl                         0.9.7           pyhd8ed1ab_1          conda-forge
matplotlib                     3.10.1          py312h1f38498_0       conda-forge
Selected Jupyter core packages...
IPython          : 8.30.0
ipykernel        : 6.29.5
ipywidgets       : 8.1.5
jupyter_client   : 8.6.3
jupyter_core     : 5.7.2
jupyter_server   : 2.14.2
jupyterlab       : 4.3.1
nbclient         : 0.10.1
nbconvert        : 7.16.4
nbformat         : 5.10.4
notebook         : not installed
qtconsole        : not installed
traitlets        : 5.14.3

It looks you aren't actually using ipympl 0.9.7 at runtime, even though that version is installed.

One report of this issue at matplotlib/matplotlib#29654 was that it was a caching issue which was solved by restarting the kernel a number of times. Please can you try that?

[T-Wainwright]

Getting the same issue with matplotlib 3.10.1, and ipympl 0.9.7, also setting the following behind the scenes to handle interactive figures:

matplotlib.use("module://ipympl.backend_nbagg")
plt.ion()

tried restarting a number of times, and reinstalling ipympl with --no-cache-dir, but no luck there.

No interactive features of interactive plots work with the error. Fixed again by downgrading matplotlib to 3.9.4 for now

[nvaytet]

It looks you aren't actually using ipympl 0.9.7 at runtime, even though that version is installed.

Doing

import ipympl

ipympl.__version__

in the same notebook does yield 0.9.7... 🤔

[nvaytet]

I cannot reproduce with the same packages in a new environment:

Can it be the Python version that is different? I can't think of anything else...

[nvaytet]

I have now tried this on a different system and I am still getting the same issue: ipympl==0.9.7 continues to raise the error about buttons with matplotlib==3.10.3.

[HojinCho]

Reporting the same issue. Python 3.12.6, matplotlib 3.10.1, and ipympl 0.9.7 from conda-forge, linux; x86-64. Downgrading matplotlib and matplotlib-base to 3.9.4 without changing other packages did resolve the issue.

Older environment with Python 3.12.8, matplotlib 3.9.3, and ipympl 0.9.4 do not experience this issue.

[ianhi]

@nvaytet what OS were you on? And did you also install via conda?

@HojinCho can you please try a debugging step for me by installing with pip instead of conda

[ianhi]

I tried to replicate your envinronment with conda on macos and was unable to replicate

[HojinCho]

@nvaytet what OS were you on? And did you also install via conda?

@HojinCho can you please try a debugging step for me by installing with pip instead of conda

Can confirm, matplotlib 3.10.1 and ipympl 0.9.7 installed from pip, under Python 3.12.6 from conda-forge, works without the error (linux x86-64). Seems like the conda-forge distribution has a bug.

[ianhi]

Thank you @HojinCho! I another debugging request. Could you try installing directly from the tar.gz on pypi (which is how the conda forge build does it).

 pip install https://files.pythonhosted.org/packages/39/1b/acd3d481869d614a7acbb831bd9304411eae47b05d7607db989613028788/ipympl-0.9.7.tar.gz

and let me know if that works?

I want to determine if its an issue with the uploaded source, or with the actual build process

[HojinCho]

Thank you @HojinCho! I another debugging request. Could you try installing directly from the tar.gz on pypi (which is how the conda forge build does it).

 pip install https://files.pythonhosted.org/packages/39/1b/acd3d481869d614a7acbb831bd9304411eae47b05d7607db989613028788/ipympl-0.9.7.tar.gz

and let me know if that works?

I want to determine if its an issue with the uploaded source, or with the actual build process

I did following. From the same environment I tested previously, I ran pip uninstall ipympl, which removed (obscured the system path with ${ENV} just to be safe)

${ENV}/etc/jupyter/nbconfig/notebook.d/jupyter-matplotlib.json
${ENV}/lib/python3.12/site-packages/ipympl-0.9.7.dist-info/*
${ENV}/lib/python3.12/site-packages/ipympl/*
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/package.json
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/480.61003b90daa033903fc2.js
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/486.73b0487516c94a05d0c2.js
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/486.73b0487516c94a05d0c2.js.LICENSE.txt
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/643.9489cdb90c0b68743dc8.js
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/688.89f91b958e83eda58752.js
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/remoteEntry.1cd74fa323152bfea195.js
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/style.js
${ENV}/share/jupyter/labextensions/jupyter-matplotlib/static/third-party-licenses.json
${ENV}/share/jupyter/nbextensions/jupyter-matplotlib/extension.js
${ENV}/share/jupyter/nbextensions/jupyter-matplotlib/index.js
${ENV}/share/jupyter/nbextensions/jupyter-matplotlib/index.js.LICENSE.txt
${ENV}/share/jupyter/nbextensions/jupyter-matplotlib/index.js.map

and then I ran the command you posted. It produced following output regarding Node.js and npm.

  × Preparing metadata (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [43 lines of output]
      INFO:hatch_jupyter_builder.utils:Running jupyter-builder
      INFO:hatch_jupyter_builder.utils:Building with hatch_jupyter_builder.npm_builder
      INFO:hatch_jupyter_builder.utils:With kwargs: {'path': '.', 'build_cmd': 'build:prod', 'npm': ['jlpm']}
      INFO:hatch_jupyter_builder.utils:Installing build dependencies with npm.  This may take a while...
      INFO:hatch_jupyter_builder.utils:> /tmp/pip-build-env-n6eovjgp/overlay/bin/jlpm install
      Traceback (most recent call last):
        File "/tmp/pip-build-env-n6eovjgp/overlay/bin/jlpm", line 8, in <module>
          sys.exit(main())
                   ^^^^^^
        File "/tmp/pip-build-env-n6eovjgp/overlay/lib/python3.12/site-packages/jupyterlab/jlpmapp.py", line 43, in main
          execvp("node", ["node", YARN_PATH, *argv])
        File "/tmp/pip-build-env-n6eovjgp/overlay/lib/python3.12/site-packages/jupyterlab/jlpmapp.py", line 24, in execvp
          cmd = which(cmd)
                ^^^^^^^^^^
        File "/tmp/pip-build-env-n6eovjgp/overlay/lib/python3.12/site-packages/jupyterlab_server/process.py", line 61, in which
          raise ValueError(msg)
      ValueError: Please install Node.js and npm before continuing installation. You may be able to install Node.js from your package manager, from conda, or directly from the Node.js website (https://nodejs.org).
      Traceback (most recent call last):
        File "${ENV}/lib/python3.12/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 389, in <module>
          main()
        File "${ENV}/lib/python3.12/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 373, in main
          json_out["return_val"] = hook(**hook_input["kwargs"])
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "${ENV}/lib/python3.12/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py", line 178, in prepare_metadata_for_build_wheel
          whl_basename = backend.build_wheel(metadata_directory, config_settings)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/tmp/pip-build-env-n6eovjgp/overlay/lib/python3.12/site-packages/hatchling/build.py", line 58, in build_wheel
          return os.path.basename(next(builder.build(directory=wheel_directory, versions=['standard'])))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/tmp/pip-build-env-n6eovjgp/overlay/lib/python3.12/site-packages/hatchling/builders/plugin/interface.py", line 147, in build
          build_hook.initialize(version, build_data)
        File "/tmp/pip-build-env-n6eovjgp/normal/lib/python3.12/site-packages/hatch_jupyter_builder/plugin.py", line 94, in initialize
          raise e
        File "/tmp/pip-build-env-n6eovjgp/normal/lib/python3.12/site-packages/hatch_jupyter_builder/plugin.py", line 89, in initialize
          build_func(self.target_name, version, **build_kwargs)
        File "/tmp/pip-build-env-n6eovjgp/normal/lib/python3.12/site-packages/hatch_jupyter_builder/utils.py", line 115, in npm_builder
          run([*npm_cmd, "install"], cwd=str(abs_path))
        File "/tmp/pip-build-env-n6eovjgp/normal/lib/python3.12/site-packages/hatch_jupyter_builder/utils.py", line 231, in run
          return subprocess.check_call(cmd, **kwargs)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "${ENV}/lib/python3.12/subprocess.py", line 413, in check_call
          raise CalledProcessError(retcode, cmd)
      subprocess.CalledProcessError: Command '['/tmp/pip-build-env-n6eovjgp/overlay/bin/jlpm', 'install']' returned non-zero exit status 1.
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed

× Encountered error while generating package metadata.
╰─> See above for output.

note: This is an issue with the package mentioned above, not pip.
hint: See above for details.

I've never seen this output, so I did not try any further and post it here. Should I try to install additional dependencies and proceed?

[HojinCho]

Additionally, I tested with another clean environment, with Python 3.12.6 from conda-forge and matplotlib 3.10.1 from PyPI. Running

pip install https://files.pythonhosted.org/packages/39/1b/acd3d481869d614a7acbb831bd9304411eae47b05d7607db989613028788/ipympl-0.9.7.tar.gz

produced the same error with Node.js and npm.

[nvaytet]

I'm running kubuntu 24.04 and installed everything via conda forge

[ianhi]

Ok! Thank you both for the info. I think it seems that the conda forge build is broken (at least for Linux) I can't test that easily while traveling. But I'll leave myself a reminder to check this on my linux computer when I'm home in a few weeks. In the meantime it seems installing from pip is a workaround, and hopefully anyone with the power to regenerate a conda build (@martinRenou ?) might be able to fix this