Support for inline matplotlib in Jupyter Notebook extension

Omniverse Isaac Sim
1

In the documentation it is mentioned that matplotlib is not supported when using the Jupyter Notebook extension. I can understand that opening of plot windows would not be supported, but how about supporting inline plotting:

%matplotlib inline

I see the following remark in exts/isaacsim.code_editor.jupyter/data/launchers/ipykernel_launcher.py:

        # magic commands
        if code.startswith("%"):
            # TODO: process magic commands
            pass

What is the timeline to add support for these specifiers such as %matplotlib inline?

2

Right now there is no committed timeline that can be shared for adding support for these magics or for inline plotting in the embedded notebook UI. The request has been passed on to the team as a feature request, though.

3

Hi.
Any update here?
Will the upcoming release of Isaac Sim (6.0.0) allow matplotlib and OpenCV usage either as inline plots in the notebook or as pop-up windows within Ssaac Sim?

4

No, Isaac Sim 6.0.0 will not support inline matplotlib or OpenCV windows. The feature hasn’t been worked on — the TODO is still sitting there from the original implementation.

5

Hi @mindthomas,

To close the loop on this: inline matplotlib (and OpenCV window) support in the embedded Jupyter Notebook extension is not currently planned. The magic command handling in ipykernel_launcher.py remains a placeholder, and there are no plans to implement it in the foreseeable future.

As noted in the Jupyter Notebook documentation:

Under Limitations:

  • IPython magic commands are not available
  • Matplotlib plotting is not available in notebooks

The embedded notebook (isaacsim.code_editor.jupyter) is designed for lightweight interactive scripting within a running Isaac Sim instance.

For workflows that require plotting or visualization, we recommend using the standalone Jupyter Notebook workflow instead. You can launch it via ./jupyter_notebook.sh with the Isaac Sim Python 3 kernel. This runs as a standard IPython kernel and fully supports %matplotlib inline, OpenCV, and other standard features.

For setup details (Linux only), see:

Closing this topic. Thank you for the feedback.