Mastering Conda Environments for Reproducible Research

Master the art of building fast, portable, and reproducible software environments on the Tufts HPC using modern Conda best practices.

This workshop covers the complete research workflow, from configuring your .condarc to move environments into group storage—avoiding the 30GB home quota—to utilizing the high-performance libmamba solverfor rapid dependency resolution.

You will learn to leverage community channels like conda-forge, manage project-specific environments, and integrate your setups as custom kernels for Jupyter on Open OnDemand.

You will leave with the skills to run efficient, isolated, and fully reproducible data analysis workflows on the cluster.

Requirements:

• This workshop will use the newly updated HPC. If you don't have access to the new cluster(https://ondemand-prod.pax.tufts.edu/), please submit a cluster account request at least 3 days prior to the workshop.
• 2FA is required
• Tufts VPN is needed for off-campus network

-- This workshop is provided and supported by Research Technology, TTS