You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Jupyter notebooks autosave to .ipynb_checkpoints, however, it seems to default to whatever the last directory you clicked on is. I'm not entirely clear on the expected behavior, but it creates issues when I'm browsing mounted data and jupyter puts an .ipynb_checkpoints directory in a data directory. Later, if I run code that tries to walk through that data directory, it finds .ipynb_checkpoints, which causes an error and forces me to manually go and delete that file. It looks like you can customize the checkpoint directory (jupyterlab/jupyterlab#5809) when starting jupyter, so maybe that's something we want to consider adding as part of starting jupyter lab?
The text was updated successfully, but these errors were encountered:
It looks like any time you try to do anything in Jupyterlab (terminal, text editor, notebook, etc), it creates that checkpoints hidden directory. If you set the directory location, there's a bug where it will overwrite checkpoints when you have several notebooks/files with the same name: jupyterlab/jupyterlab#5809 (comment)
Something to keep an eye on, but I'm hesitant to hardcoding the checkpoint directory ourselves because of that bug.
Here's the issue tracking this problem: jupyter/notebook#4645. Hasn't been fixed yet.
Jupyter notebooks autosave to
.ipynb_checkpoints
, however, it seems to default to whatever the last directory you clicked on is. I'm not entirely clear on the expected behavior, but it creates issues when I'm browsing mounted data and jupyter puts an.ipynb_checkpoints
directory in a data directory. Later, if I run code that tries to walk through that data directory, it finds.ipynb_checkpoints
, which causes an error and forces me to manually go and delete that file. It looks like you can customize the checkpoint directory (jupyterlab/jupyterlab#5809) when starting jupyter, so maybe that's something we want to consider adding as part of starting jupyter lab?The text was updated successfully, but these errors were encountered: