Dear lenstronomers,
I would like to give a perspective in the near and intermediate future of the development of lenstronomy.
Over the passt few years, lenstronomy has gained popularity in the strong lensing community and beyond. Soon we are approaching 100 research publications making use of lenstronomy, we have 21 direct contributors to the source code and about 100 active users at various stages ranging from educational to science analyses.
The last two years have seen a consolidation of the features, documentation and we have published a second lenstronomy paper acknowledging the contributions from a broader community. The next steps in the lenstronomy project is to ensure long term support and development.
The following key organizational tasks and developments I identified as the key goals for the next year:
- moving lenstronomy to a dedicated project repository: A dedicated project repository can enhance the visibility of the software and highlights the fact that there are more than one maintainer and developer associated with the software. Github makes this process rather smooth and for you as a user, nothing changes, and as a developer, only the project repository route needs to be changed (perhaps not even that). More to come in the next few months on that.
- establishing a base of maintainers and co-developers: My own capacity for active coding and maintaining is expected to drop to a lower level. In the intermediate term, I would like to establish a co-equal group of maintainers and developers. If you are interested in some capacity and have also some vision in what direction lenstronomy may evolve in the future, please let me know!
- improved Jupyter notebooks: Documentation and use examples can always improve. The intermediate plan is to also move the lenstronomy_extension notebooks to the project repository and enable all of them to be on google colab and with html rendering.
- logo: We need a recognizable logo :) I will set up a competition with a small prize to it in the coming months.
- Distribution: enable direct conda installation. Many of you are working in conda environments and having a conda-forge installation possible might facilitate setting up conda environments
- Visualization: Visualization is important for the user and to present results to colleagues and the public. I plan to improve the integration of the light cone rendering visualization tool lightcurve with lenstronomy and there is much more room for creativity and improvements
- Funding: A first step is applying for NumFocus affiliation. In the longer term, if there is perhaps even some funding for development and maintaining of the code basis that can be useful as well. Long-term support is best guaranteed with long-term funding and resources.
- Speedup: There are exciting developments in using GPU-compatible coding, JAX, and more broadly auto-differentiating tools. The goal is to have partial support for some external code bases that can take on such task, and/or a lightweight lenstronomy version that can fully support it. I hope that shifting the structure of the lenstronomy maintaining
If you have ideas and input about the prioritization of effort, please let me know! I will update this list on the individual tasks in the months to come.
Thank you very much for your contributions, usage, feedback etc to making lenstronomy what it is now!
Simon