Short-Term Projects¶

We are currently working on xWRF, a prototype of an Pangeo-friendly analysis plug-in for Xarray. This prototypical effort is exploring how functionality that currently exists in the wrf-python package can be implemented in a way that is compatible with the Pangeo stack.

We are also in the process of completing Project Funnel, which developed a caching and workflow automation framework for climate and weather analysis based on Prefect. As part of this project, we developed the xpersist package for caching steps in the workflow that return Xarray objects.

Selecting Future Projects¶

Our current mechanism for selecting short-term projects for Xdev’s effort is internal. We have assessed the pain points that the community has based on Q/ A through either our Zulip platform or through our weekly Office Hours and based on direct interactions with scientists. A record of questions asked in the Xdev Office Hours is made, and Zulip records are held indefinitely. We collaborate with individuals who reach out to us individually.

Moving forward, however, we believe that we need to adopt a more equitable approach to assessing the most important activities for Xdev and what activities will have the greatest impact. With that in mind, it is my belief that we should begin conducting surveys of the community to ask them specifically for their pain points and their thoughts on what we should be working on. As more of the community adopts Pangeo technology and Python, we expect that a formal project proposal process can be created, whereby individuals or groups with a desire for software engineering experience in the Python realm can reach out for help on specific problems. In this capacity, Xdev could select proposals that can be generalized to impact a broader range of users than just the proposing scientists. Through this process, we believe that Xdev can have a strong impact on changing best practices and the development of scientific software.