This guide contains tips and FAQs which help to avoid common issues that arise when using JupyterLab.
- Open multiple notebook tabs
- Check solutions or feedback
- Check memory usage
- Get a fresh copy of a notebook
- View hidden files
- Clone a GitHub repo
- Get a copy of everything
- Why did I get a zero/low-mark when I submitted everything on time?
- Why are none of my cells running?
- Why does my kernel keep dying?
- Why did I get a zero on a question when all the tests passed?
- Why did I get a low mark on a question when all the tests passed?
- I made a typo in my code, can I still get marks for the question?
- I am getting a 400 (Bad Request) or 404 (Not Found) error, how do I fix it?
- How do I get a blank worksheet or tutorial for practice?
When working on a notebook, you might want to have multiple tabs of the same notebook open to easily reference previous questions.
DO NOT open another browser tab, create another instance of JupyterLab, and open the same notebook. Doing so may overwrite the notebook you are working on with an empty or half-completed version.
Instead, use JupyterLab's built in tab functionality to create another view for the notebook.
To do so in JupyterLab, right click the tab that contains the notebook in question.
On the dropdown menu, click New View for Notebook:
This will create a split screen, showing a copy of the notebook on half of the screen.
Solutions and feedback are uploaded to your home directory. This is a different directory to where assignments are stored.
To get to your home directory, open JupyterLab and then click the small grey folder icon on top of the menu to the left that shows files and folders.
Then that same menu should show assignment solutions and feedback. Solutions are typically released a few days after the due date, feedback is released after assignments are graded (this may take a couple weeks as tutorials are manually graded).
Student JupyterLab instances are limited to 2GB (2048MB) of memory usage, if that limit is reached by running a notebook then the kernel will be killed and cells will stop running.
To check how much memory JupyterLab is using, there is a indicator on the bottom left of the window:
The maximum amount of memory usage a notebook should use is around 1GB (1024MB). If you find that JupyterLab is hitting the limit very often, here are some steps you can take:
On the top left, go to File -> Hub Control Panel and it will open a window with a red button with text Stop My Server, click the button and then click the Log Out button on the top right.
Close the browser tab and then open JupyterLab again from Canvas.
If restarting your server does not work, then there may be an issue with your code.
Run your notebook one cell at a time. If you hit a cell that causes memory usage to hit the 2GB limit, examine the code in the cell to make sure there aren't any errors e.g. running a function on unfiltered data.
If you need a new copy of a notebook, for example if you accidentally deleted a grading cell or want one for practice, first rename the notebook you have in your account currently.
For example if you need a new copy of a worksheet_wrangling.ipynb, right-click it in the file menu and choose the Rename option and rename it to worksheet_wrangling_old.ipynb.
Then close JupyterLab and then open the assignment again on Canvas. A new copy of the assignment should be created in place of the renamed assignment.
View hidden files
Warning: Only delete hidden files that you have created, other hidden files may be used by the server
To view hidden files, on the top left go to View and on the dropdown click Show Hidden Files
To clone a GitHub repo for collaborative work, follow these steps:
- Find the HTTPS repo link from GitHub and copy it to the clipboard
- Go to your JupyterLab and make sure you're in the home directory by clicking the small grey folder icon on the top left of the menu that shows your files and folders
- Go to the JupyterLab git extension menu by clicking its tab on the left of JupyterLab:
- Click "Clone a Repository" and then paste the repo link from step 1 into the box and click Clone.
The repo should now be in your home directory in a folder titled with the name of your repo.
To get to your cloned repo, remember to navigate to the home directory as shown in step 2.
To get a zipped copy of all of your work at the end of term:
- Open a terminal by clicking the terminal icon when you first open JupyterLab from the Canvas homepage or in a new JupyterLab tab.
- Enter the following command into the terminal:
cd /home/jovyan/work/ && zip -r backupjupyterlab.zip *
If you received a Disk quota exceeded error, then there are too many files. Delete any files or folders that you don't need (especially large .csv files) and try the command again.
- Go to your home folder as described here and there should be a zip file. Right-click the file and click Download to download a copy onto your device.
If you are aware of the above, then make sure that you are opening another notebook tab in the correct way. If you opened another browser tab with JupyterLab, then there is a chance your notebook was overwritten.
Lastly, do not rename the assignment notebook file or any folders. The autograder relies on those names to collect submissions.
Before anything else, try restarting JupyterHub as described here.
Make sure you're using a compatible browser. We recommend Chrome, Edge, or Firefox. Also update your browser to the latest version.
Update your device's operating system to the latest version and reboot your device.
Disable any extensions to check if they are affecting JupyterLab.
Try opening JupyterLab in an incognito browser session.
Please read Check memory usage.
You may have deleted a grading cell accidentally (grading cells are cells that contain test functions e.g. test_3.2()).
The autograder relies on these cells to give students marks. Creating a new cell and then copying the test function into that cell will not fix the issue.
If you do accidentally delete a grading cell, you should get a fresh copy of the assignment notebook. Then, copy your work into the fresh notebook.
Some questions have hidden tests that are not shown in the student version of the assignment.
Your answer may have passed the visible tests but failed the hidden tests when it was being marked.
Examine the feedback for the question carefully to see if hidden tests were run on your code. Instructions for accessing feedback are here.
We will not be making exceptions for typos in autograded coding questions.
Read the questions carefully as some questions request answers of a certain type (e.g. String).
The questions include tests that give instant feedback on your code. Run those tests frequently to ensure that you haven't made any typos.
Before continuing with the instructions below, try JupyterLab again after restarting your browser.
The errors could be due to your browser's cache and it may need to be cleared.
Instructions:
After clearing your browser's cache, restart the browser and try opening JupyterLab again.
If it still has the same error, try using Incognito or Private Browsing on your browser.
Lastly, disable any browser extensions as they could interfere with JupyterLab. Also check that your browser is updated to its latest version.
Follow the instructions here to get a blank copy of a tutorial or worksheet.