JupyterLab: The next generation user interface for Project Jupyter
https://github.com/jupyter/jupyterlab
It has been a collaboration between:
- Project Jupyter
- Bloomberg
- Anaconda
Use it to open different activities:
- Notebook
- Console
- Editor
- Terminal
- Open example notebooks to show that notebooks still work
- Collapse input/output
- Drag and drop cells
- File Browser (file operations, context menu, including drag and drop)
- Running
- Command Palette (fuzzy searching for 'new')
- Open
markdown_python.md
in the File Editor - View the rendered markdown, arrange side by side
- Attach a Kernel/Console and run the code by selecting blocks and pressing
Shift+Enter
The dock panel allows you to arrange the activites into an arbitrary layout.
Tabs and single document mode allow you to focus.
JupyterLab has a powerful and extensible architecture for handling a wide range of file formats:
- CSV
./data/iris.csv
(small)TCGA_Data
(small to medium)- Urban_Data_Challenge:
data/big.csv
- Images
data/hubble.png
- Vega-Lite
data/vega.vl.json
- Open DC museum GeoJSON file from OpenData DC:
data/Museums_in_DC.geojson
- Notebook demonstrating bqplot widgets:
notebooks/bqplot.ipynb
The genius of open-source is being able to shape your tools to your heart's content.
Just like Jupyter is built on top of building blocks of the protocol and message spec, you can build on this platform for your workflow.
- Everything in JupyterLab is an extension, including everything we have demoed
- Extensions are just
npm
packages with metadata - Anyone can create, package, ship plugins
- Extension can, for example:
- Add things to command palette, menu
- Add viewers for documents
- Expose other controls (e.g., manage a spark cluster?)
- Provide more capabilities to the system