Closed ncclementi closed 1 year ago
I've cleaned things up a little and (force) pushed. This isn't doing interactive rendering for me just yet though. Was that working for you? Any thoughts?
Yes it was doing it perfectly, with no problem. I tried it even before pushing it.
Can I ask you to try with this one? I'm curious if this is an environment issue.
Just making sure: did you click on the magnifier symbol on the right.?
With my environment, I was able to run your updated notebook and it interacts
I removed some things from the environment, like JLab. Are you running with that environment file or with the previous one that you made?
I'm running it with the one I had, were the only difference is Jlab and dask-lab-extension. Not sure why removing JLab from the env when you need it to launch the notebook
name: datashader-demo
channels:
- conda-forge
dependencies:
- python==3.10
- dask
- coiled
- pyarrow
- ipykernel
- dask-labextension==6.1.0
- jupyterlab
- s3fs
- datashader
- holoviews
- hvplot
Not sure why removing JLab from the env when you need it to launch the notebook
I'm running JupyterLab from my base environment, and using the kernel switcher to select different kernels. This is useful, for example, if I'm running different notebooks with different environments (such as datashader kind of forces us to do). I would prefer not to have multiple notebook servers running (although that's ok too).
Maybe this helps a little bit with package sync times too.
Nat, Gabe, James, among other folks, tested the times of spinning up a cluster with and without JLab and there was pretty much no impact, I wouldn't worry about that.
OK, I can get this to run, but only if we run JupyterLab from the exact same environment. This doesn't seem to work if we're using ipykernel . Merging anyway.
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB