Interactive window with Matplotlib and notebook option

[PierreMarchand20]

Environment data

• VS Code version: 1.30.1
• Extension version (available under the Extensions sidebar): XXX
• OS and version: macOS Mojave
• Python version (& distribution if applicable, e.g. Anaconda): 3.7.2
• Type of virtual environment used (N/A | venv | virtualenv | conda | ...): N/A
• Relevant/affected Python packages and their versions: Matplotlib

Is %matplotlib notebook supposed to work in vscode ? In my case, it does not:

ReferenceError: IPython is not defined
    at eval (eval at runCodeHere (webpack:///./node_modules/@nteract/transforms/lib/javascript.js?), <anonymous>:767:1)
    at runCodeHere (webpack:///./node_modules/@nteract/transforms/lib/javascript.js?:32:12)
    at JavaScript.componentDidMount (webpack:///./node_modules/@nteract/transforms/lib/javascript.js?:60:7)
    at rh (webpack:///./node_modules/react-dom/cjs/react-dom.production.min.js?:213:482)
    at Xg (webpack:///./node_modules/react-dom/cjs/react-dom.production.min.js?:205:159)
    at Yg (webpack:///./node_modules/react-dom/cjs/react-dom.production.min.js?:204:174)
    at If (webpack:///./node_modules/react-dom/cjs/react-dom.production.min.js?:200:307)
    at Object.enqueueForceUpdate (webpack:///./node_modules/react-dom/cjs/react-dom.production.min.js?:131:37)
    at MainPanel.E.forceUpdate (webpack:///./node_modules/react/cjs/react.production.min.js?:13:311)
    at MainPanel._super.MainPanel._this.updateOrAdd (webpack:///./src/datascience-ui/history-react/MainPanel.tsx?:341:23)

[rchiodo]

@PierreMarchand20 what's your extension version?

[PierreMarchand20]

Oh sorry, it is 2018.12.1. But I am actually starting to think it may not be a bug from vscode. I installed Jupyter and jupyterlab with pip, may be it needs something more to use this option.

[PierreMarchand20]

I tried with %matplotlib ipyml following https://matplotlib.org/tutorials/introductory/usage.html#ipympl but it did not work either

[PierreMarchand20]

Ok following https://github.com/matplotlib/jupyter-matplotlib , I got FigureCanvasNbAgg() instead

[rchiodo]

What does your cell look like? Is it just a matplotlib?

[rchiodo]

The code that's failing is trying to load some javascript returned by jupyter when executing your cell. We send the cell contents to Jupyter and then stick its output js into a react control.

[PierreMarchand20]

It is just Matplotlib, and it seems that it calls the backend nbagg (https://matplotlib.org/users/prev_whats_new/whats_new_1.4.html#the-nbagg-backend) which calls some javascript (according to https://www.reddit.com/r/IPython/comments/36p360/try_matplotlib_notebook_for_interactive_plots/ but it is really hard to find more information).

[PierreMarchand20]

Regarding ipympl, since it is a jupyter widget, it may be related to this issue microsoft/vscode-python#3429

[rchiodo]

Sorry I wasn't reading your original entry that carefully. %matplotlib notebook changes matplotlib into an interactive widget. That explains why it doesn't work.

Yes I think you're correct. This is likely the same as microsoft/vscode-python#3429. We don't support jupyter extensions yet.

Thanks, this should be a good reference point to figure out how to get them working.

[franva]

how about now?

Does VSCode support jupyter extensions?

[DonJayamanne]

%matplotlib widget works today.
Please try that.

[franva]

thanks @DonJayamanne
just tried and it is almost perfect with 1 glitch.

The animation doesn't start by itself(although the sensory data has been streaming through),
it only starts animating once I resized its plot by dragging a small triangle at the right bottom corner.

I have uploaded a screen recording for this glitch.

nb.mp4

[DonJayamanne]

The animation doesn't start by itself(although the sensory data has been streaming through),

Please could you file a separate bug with a minimal code sample (python packages to be installed) to replicate this issue