Jupyter Notebooks on Mosaic

Jupyter notebook is an open-source web application that enables users to create and share not only code but also analyses and interactive visualizations. In order to learn more about the different ways you can use Jupyter Notebooks in your work, please visit http://jupyter.org/ .

Creating a new Jupyter notebook

Any user with "CONTRIBUTE" or "ADMINISTRATOR" level membership in the workspace can create a Jupyter Notebook by clicking on the "New notebook" button. To create a Jupyter Notebook, the user needs to define the configuration of the virtual environment which will host the notebook, and select the files that to which the notebook will have access.

The only required configuration input is the "Notebook Name".

Currently, there are 3 Instance Types that are available for Jupyter Notebooks:

Instance Type Name # of CPU cores Memory Hard Disk Storage
Baseline-2 2 3.8 GB 32 GB
Baseline-4 4 7.5 GB 80 GB
Baseline-8 8 15 GB 160 GB

Additionally, in order to limit inadvertent charges, workers used for hosting Jupyter Notebooks are time limited. By default, a notebook will terminate after two (2) hours, but you also have the option to set this time limit to either one (1) or four (4) hours.

You can select which files are downloaded to the worker running the Jupyter Notebook by clicking on the "Select File" button. In order to ensure all file availability for anyone who launches a notebook, only files within the working workspace can be selected.

Launching a Jupyter notebook

A Jupyter notebook has to be launched before it can be used. This is done by clicking on the ‘Launch Notebook’ option within ‘Actions’ menu.

A notebook can also be launched by clicking on the ‘Launch Notebook’ button in the notebook information page

Once a notebook is launched, it takes anywhere between two (2) to five (5) minutes before the notebook is ready. While the notebook is being readied, the status of the notebook changes from ‘Not initiated’ to ‘Initiating’.Clicking on the ‘Refresh’ button refreshes the page and loads the updated state of a launched notebook. Once the state of a notebook changes to ‘Ready’, the notebook is ready to use.

Accessing a Jupyter notebook

Jupyter Notebooks can be accessed within workspaces through the Notebooks tab. Only a notebook in ‘Ready’ state can be accessed. To access a notebook, click on the Actions menu and select the ‘Go to Notebook’ option. The notebook opens in a new browser tab.

Terminating a Jupyter notebook

A notebook in ‘Ready’ state can be stopped by clicking on the ‘Terminate’ option within the Actions menu.

Included Python modules

The following Python libraries come pre-installed on every notebook created in Mosaic.

  • Numpy
  • Scipy
  • Pandas
  • Matplotlib
  • Seaborn
  • Statsmodels
  • Sklearn
Installing new Python modules

Apart from the pre-installed modules, additional Python modules can be installed from within a notebook using the ‘pip install’ command

Examples :
  • !sudo pip install -q scikit-bio (installs ‘scikit-bio’ module on Python 2 kernel)
  • !sudo pip3 install -q scikit-bio(installs ‘scikit-bio’ module on Python 3 kernel)
Supported kernels

The following kernels are supported on Mosaic Jupyter notebook

  • Python 2
  • Python 3
  • R