Sarah Gibson - Sharing Reproducible Python Environments with Binder

Python

Learn how to share reproducible Python environments with Binder, a tool that uses Docker to create containers with software dependencies, suitable for researchers, students, and institutions worldwide.

Key takeaways

Reproducible research is important for efficiency and success
Binder is a tool that makes it easy to share reproducible computational environments
It uses Docker to create containers with software dependencies
Binder is not a tool for analysis, but for sharing results
It supports multiple programming languages, including Python
The project has grown to hosting over 140,000 user sessions per week
The main use case for Binder is sharing reproducible computational environments
The service can be used by anyone, including researchers, students, and institutions
It is hosted on multiple clusters around the world, including Google Cloud, Azure, and OVH
The project is open source and built modularly using other open source tools
It provides a repeatable analysis by capturing the software dependencies and high-level architecture
The service can be used to run Jupyter notebooks and provide interactive computing environments
It does not support GPU services, but can be configured to only allow sharing between specific teams
The project has been awarded a Moore Foundation grant to run its operations
The funding streams come from institutions and cloud providers

Sarah Gibson - Sharing Reproducible Python Environments with Binder

More talks