Introduction to JupyterLab

Dan Strong
2 min readDec 8, 2017

I was pleasantly surprised to find JupyterLab packaged with a new installation of Anaconda (Python 3). According to Boyan Angelov, JupyterLab is a “true IDE for interactive computing,” complete with file browser, terminal and code editor (in this case, Jupyter notebooks). The software is still in alpha preview, but appears to work quite well. To get started, you’ll need Jupyter Notebook version 4.3 or later, which you can check by typing jupyter notebook --version in your shell.

Probably the easiest way to get JupyterLab is with Anaconda, but you can also install it from the command line:

conda install -c conda-forge jupyterlab

…or with the Python package manager:

pip install jupyterlab jupyter serverextension enable --py jupyterlab --sys-prefix

There is another method of installing JupyterLab from Github, which is outlined out on the official markdown file.

In order to start JupyterLab, simply type jupyter lab in your shell, and a new session will open in your browser. JupyterLab uses CSS Variables for styling, so only the latest versions of Chrome, Safari and Firefox are supported.

Here’s what it looks like “in the flesh”:

JupyterLab IDE

You can open multiple tabs and have several Python or R kernels running side-by-side.

JupyterLab supports a number of different programming languages:

There’s a built-in python console

You can map keys to match your favorite text editor or IDE

Overall, I’m impressed with what JupyterLab can do, and so far I haven’t discovered many limitations with this new IDE. I’m looking forward to seeing JupyterLab improve and gain functionality. I’ll post any updates here on my blog.

Useful Links

Originally published at danstrong.tech.

--

--

Dan Strong

Research analyst with an interest in Python for data analysis and Ruby for web development