What is Nextjournal?
Nextjournal is a comprehensive notebook platform designed specifically for reproducible research across various domains including machine learning, data science, and scientific publishing. The platform runs anything that can be placed into a Docker container, offering polyglot notebooks that support multiple programming languages within a single environment. It features automatic versioning and real-time collaboration capabilities that allow multiple users to work simultaneously on the same notebook.
The platform provides on-demand provisioning of computational resources, including GPU support for machine learning tasks, with automatic shutdown to optimize costs. Users can connect external data sources securely, utilize familiar visualization libraries, and manage secrets through a secure storage system. Notebooks can be shared privately or published publicly with permanent URLs, and the remix functionality enables building upon previously published work while maintaining immutability.
Features
- Polyglot Notebooks: Support for multiple programming languages including Python, R, Julia, Clojure, Tensorflow, Keras, Pytorch and Bash within a single notebook
- Real-Time Collaboration: Multiple users can edit notebooks simultaneously with commit-less synchronization across connected clients
- Automatic Provisioning: On-demand compute instances with GPU support and automatic shutdown to prevent idle costs
- Data & Secrets Management: Secure connection to external data sources (S3, Google Cloud Storage) and private secrets storage using Vault
- Remix Functionality: Build upon previously published notebooks while maintaining immutability of original content
Use Cases
- Machine learning researchers training models on GPUs with minimal setup time
- Data scientists collaborating on analyses with real-time editing and visualization libraries
- Library authors publishing runnable documentation samples alongside text explanations
- Professors providing coursework with executable exercises to students
- Scientists creating reproducible analyses with all raw data for peer verification
FAQs
-
What programming languages does Nextjournal support?
Nextjournal supports multiple programming languages including Python, R, Julia, Clojure, Tensorflow, Keras, Pytorch and Bash within polyglot notebooks. -
How does the real-time collaboration feature work?
All edits to a Nextjournal notebook are synchronized in real-time among connected clients, allowing multiple authors to edit simultaneously without traditional commits. -
What happens to compute instances when computations are complete?
Compute instances are automatically shut down when computations are finished to prevent additional costs from idling machines. -
Can I import my existing notebooks into Nextjournal?
Yes, you can import existing Jupyter, IPython and RMarkdown notebooks into the Nextjournal platform. -
What external data sources can I connect to?
Nextjournal notebooks can connect to private data from S3 and Google Cloud Storage, clone private GitHub repositories, and import images from Docker Hub.
