JupyterLab

What JupyterLab is

Project Jupyter describes JupyterLab as its "latest web-based interactive development environment for notebooks, code, and data". It runs in your browser but works on your own machine: you start it from a terminal and it opens a workspace with a file browser, notebooks, a text editor, terminals and code consoles side by side. The classic Jupyter Notebook still exists as a simpler, document-focused interface, and both open the same .ipynb files.

A notebook is an open document format based on JSON that keeps your code, the output it produced, Markdown notes and equations together. That is why notebooks became the default way to explore a dataset, try a model and show someone else what you found.

Why it matters for a data science learner

Data work is iterative. You load a CSV, look at the first rows, notice a column full of blanks, clean it, plot it, and change your mind. Running one cell at a time and seeing the table or chart immediately makes that loop fast. Almost every tutorial you will meet for pandas, NumPy, seaborn, statsmodels and scikit-learn is written as a notebook, so being comfortable here removes friction from everything else.

Useful features once you are past the basics:

• Code consoles: scratchpads for trying a line without cluttering your notebook.
• Viewers for data files: CSV, JSON, images and PDFs open in their own tabs.
• Output in its own tab: pin a chart while you keep editing the cells that produce it.
• Extensions: themes, a debugger, Git integration and more.

Getting started and time needed

Install with pip install jupyterlab, then run jupyter lab in the folder that holds your project. Expect ten minutes to install and an afternoon to feel at home with cells, kernels and keyboard shortcuts. If you cannot install anything yet, the Try page on jupyter.org lets you use it in the browser.

Prerequisites

Basic Python and a terminal. It helps to use a virtual environment per project so that each notebook has the libraries it expects.

Is it free?

Yes. It is open source and there is nothing to buy. Hosted services built on Jupyter exist, but you do not need any of them to learn.

Common questions

JupyterLab or the classic Notebook? JupyterLab, unless you want the most minimal interface. The files are the same.

Are notebooks good for production code? Not on their own. Explore in a notebook, then move code you reuse into plain .py modules and tests. Restart the kernel and run all cells before sharing, because cells run out of order are the most common source of "it worked on my machine".

Other languages? Jupyter supports over 40 languages, including R and Julia, through separate kernels.

Best for: anyone learning or doing data analysis in Python who wants code, results and notes in one place.