# JupyterLab

JupyterLab (opens new window) is the latest web-based interactive development environment for notebooks, code, and data. Its flexible interface allows users to configure and arrange workflows in data science, scientific computing, computational journalism, and machine learning.

Here we try to summarize basics of what you should know when starting with Jupyter.

We recommend official JupyterLab guides to get familiar with the interface:

Jupyter notebooks are documents that combine live runnable code with narrative text (Markdown), equations (LaTeX), images, interactive visualizations and other rich output.


# Working with Notebooks

You can find useful explanations and video guides in official documentation:

# Working with lab mates

Typically your lab has more users and you all are sharing compute (opens new window) and storage resources. For this to work efficiently you will need to know how to check your resource usage and free up resources if needed.

# Monitoring resources

One way to monitor resources is using Glances. We have included Glances proxy extension (opens new window) in JupyterLab installation.

You can also you Terminal app inside your Workbench and run standard commands like HTOP (opens new window) and others

For checking storage capacity and how much data are written on disk have a look at df -h --type=ext4 and ncdu.

# Managing resources

Read more in the official JupyterLab guide about Managing Kernels and Terminals (opens new window) to find out about Kernels in JupyterLab and how you can stop them to free up RAM (operational memory) once you do not use them.

# Long running notebooks

If your Jupyter Notebook takes too long to execute (more than VPN session), you should consider running it from terminal. You will also need to use terminal multiplexers (opens new window) to assure that you can always reconnect to your session.

Activate your Conda environment first:

conda activate py310

Then run your notebook:

jupyter run your_notebook.ipynb
Last Updated: 9/27/2023