New Python guidelines: ACCRE-optimized Python 3 builds with virtual environments are strongly encouraged
We have significantly revised our documentation on using Python on the ACCRE cluster to address some of the most common issues that are brought to our attention and improve the overall cluster user experience. Here is a brief overview of some changes:
- Although Anaconda provides an easy way to use a large selection of software, we strongly recommend using the Python interpreters installed in our software stack. In this way you can use all the Python packages we provide that have been optimized to ensure performance and compatibility.
- When installing new Python or Anaconda packages locally, we strongly recommend using virtual environments instead of installing them locally. This allows much better control over the installed packages and, when paired with modules collections, strongly prevents incompatibilities that affect many users.
- We highly encourage transitioning to Python 3, as Python 2 will be deprecated at the end of 2019.
- We encourage using the ACCRE Visualization Portal to create Jupyter notebooks, as opposed to the SSH tunneling instructions we have provided in the past.
- We have added instructions for using Python on GPU nodes.
The new instructions are available here.