This is a collection of Jupyter Notebooks that are templatized and meant to be run with the papermill library. It is automatically run each day via a CRON job on GitHub actions, and generates a website with reports for various communities in the Jupyter ecosystem at https://predictablynoisy.com/jupyter-activity-snapshot/.
monthly_update/run_template.ipynb
- a "controller" notebook that manages the execution and report generation of template notebooks. Start here.monthly_update/templates
- contains template notebooks that generate reports about activity over a few months of time (by default, 3 months).monthly_updates/data/
- GitHub activity data that is stored as a part of executing notebook data "fetch" operations. This data is then re-used by report notebooks.monthly_updates/generated/
- Outputs from running papermill on the template notebooks, this is bundled up in a Jupyter Book for display purposes.
Other folders are out of date, and left-over from prototyping different kinds of reports. Feel free to look at them for inspiration but don't expect them to work!
The reports are best viewed on nbviewer. You'll see a few folders representing activity in different moments in time. Click on the folder and then on the report that you'd like to see.
A caveat: The data in these reports is just what could be (relatively) easily scraped from the GitHub and Discourse APIs. That means we are missing a lot of really important activity that is difficult to scrape automatically. Don't treat the numbers in these reports as the total truth - they are just one view on the complex web of activity that is the Jupyter community.
Note, if the information on a page looks outdated, try flushing the nbviewer cache with ?flush_cache=true
in the URL.