Jupyter (software)

Jupyter (software)
Name	Jupyter
Developer	Project Jupyter
Released	2014
Programming language	Python, JavaScript
Operating system	Cross-platform
Genre	Interactive computing, Notebook interface
License	BSD

Jupyter (software) is an open-source interactive computing platform originally spun out of the IPython project and developed by Project Jupyter. It provides a web-based notebook interface combining executable code, rich text, mathematics, and visualizations for reproducible workflows used across science and industry. The project intersects with major institutions and initiatives such as the NumFOCUS foundation, Berkeley Laboratory, Microsoft, Google, Amazon Web Services, and Mozilla.

Overview

Jupyter offers a document format and application for interactive computing that integrates code execution, narrative, and output within a single file model. The project connects to language kernels such as Python (programming language), R (programming language), Julia (programming language), and Scala (programming language), enabling multi-language support across environments like Anaconda (distribution), Docker, Kubernetes, and OpenStack. Users work through a browser-based frontend served by the Jupyter Notebook server or the modern JupyterLab interface, producing artifacts consumed by tools including GitHub, GitLab, Binder (service), and nbviewer.

History and Development

Jupyter originated from the evolution of IPython begun by Fernando Pérez and collaborators, formalized with the creation of Project Jupyter in 2014 to reflect language-agnostic ambitions. Development received support and contributions from organizations such as Cal Poly, UC Berkeley, MIT, Harvard University, and corporate partners like Microsoft Research and Google Research. Key milestones include the separation of the kernel protocol from the IPython kernel, the specification of the notebook JSON format, and the emergence of JupyterLab as the next-generation interface, influenced by standards work in communities like W3C and practices from Apache Software Foundation projects.

Architecture and Components

The architecture separates a browser-based frontend, language-specific kernels, and server-side services. Core components include the Jupyter Notebook application, the JupyterLab environment, the Jupyter Notebook format (JSON-based .ipynb), and the Jupyter Kernel Protocol for communicating with kernels such as the IPython kernel or third-party kernels like IRkernel, IJulia, and almond (Scala kernel). Supporting components and tools include the nbformat library, nbconvert for static export to formats used by LaTeX, HTML5, and PDF, and servers enabling deployment on infrastructure such as Heroku, Google Colab, Azure Notebooks, and AWS SageMaker.

Use Cases and Applications

Jupyter notebooks serve a variety of use cases across research, education, and product development. In scientific computing they are employed by groups at CERN, NASA, European Space Agency, and Los Alamos National Laboratory for data analysis, visualization, and reproducibility. In machine learning and data science, practitioners use notebooks with frameworks and libraries like TensorFlow, PyTorch, scikit-learn, pandas, and NumPy to prototype models and report experiments. In education, institutions such as Massachusetts Institute of Technology, Stanford University, University of Oxford, and University of Cambridge adopt notebooks for hands-on instruction, while publishers like O’Reilly Media and Springer integrate notebook examples into textbooks. Industry users include teams at IBM, Facebook, Netflix, Uber, and Spotify for exploratory analytics and reproducible reports.

Ecosystem and Extensions

An extensive ecosystem surrounds the core project, including extensions, visualization libraries, and hosting services. Notable extensions and projects include JupyterHub for multi-user deployments, Binder (service) for reproducible execution environments, Voila for converting notebooks into standalone dashboards, and visualization integrations with Bokeh, Plotly, Matplotlib, and Altair. Packaging and interoperability are aided by standards and package managers like Conda (package manager), pip, npm, and build tools such as Make (software). Integration with version control and continuous integration workflows leverages platforms like GitHub Actions, Travis CI, CircleCI, and Jenkins.

Adoption and Community

Adoption spans academia, government labs, and commercial ecosystems, with community governance and contribution models involving individuals and organizations including NumFOCUS, Project Jupyter Steering Council, Google Summer of Code, and numerous university research groups. The developer and user community converges at conferences and events like SciPy, PyCon, DataFest, OSCON, and FOSDEM, while mailing lists, chatrooms on Slack, and forums like Stack Overflow and Discourse support collaboration. Major contributors over time include researchers and engineers affiliated with UC Berkeley AMPLab, Lawrence Berkeley National Laboratory, and corporate R&D teams at Microsoft and Google.

Security and Governance

Security considerations address execution isolation, kernel privilege, and notebook provenance when deploying in shared environments. Solutions and best practices reference containerization with Docker, orchestration with Kubernetes, authentication via OAuth 2.0, authorization patterns used by LDAP, and secrets management techniques from HashiCorp Vault. Governance of the project is balanced between community contributors, the Project Jupyter Steering Council, and fiscal sponsorship by NumFOCUS, with licensing and intellectual property managed under permissive BSD-style terms and collaboration agreements influenced by norms at Apache Software Foundation and other open-source bodies.

Category:Free software Category:Open-source software Category:Educational software