Jupyter Notebook is a web application for creating and sharing computational Python documents.
We provide an interactive app for Jupyter Notebook using the Open OnDemand HPC web portal.
Run Jupyter Notebooks using Open OnDemand#
From the Open OnDemand dashboard, select Interactive Apps → Jupyter Notebook. You will see something similar to the following screen:
Set the desired parameters, and then click the Launch button.
If you are building a new environment, you should specify a value of at least 8 for the number of cores. It will take a few moments to build the new environment (up to 3 to 10 minutes), during which the job dialog will indicate Starting:
If you are connecting to an already existing Jupyter Notebook environment, your environment will load immediately after your job starts.
Manually configuring a Jupyter Notebook environment#
You have the option of manually building and configuring a Jupyter Notebook environment, rather than using Open OnDemand to automatically build it:
To run your newly created notebook on Open OnDemand (see above), take note of the path that the
environment location: that was printed during the
conda create output:
Enter this path into the Path to Jupyter Virtual Environment field in Open OnDemand:
Using R with Jupyter Notebooks#
Our Jupyter Notebook Open OnDemand app automatically loads "R" and installs the necessary dependencies to use R with Jupyter Notebooks.
Select the version of R that you would like to use on the job submission screen:
Fill out the remainder of the form, and then follow the instructions above to submit your job. When the job starts, press the "New" button at the upper-right hand corner of the Jupyter notebook screen, and select "R":
Manually install the "R" kernel in Jupyter#
If you need further customization, you can manually install the "R" kernel in your Jupyter Notebook environment.
You can now run "R" in your Jupyter Notebook session.