We highly encourage contributions to this project and welcome PRs.
Familiarity with fastpages is helpful. All you need to do is include the appropriate front matter in your notebook. See this guide.
A good pull request to emulate is this one - study the Jupyter notebook closely, especially the first cell that specifies the metadata, and how cells are hidden with #hide or #hide_input.
You should preview your dashboard locally by following the Development Guide, before submitting a PR.
- Your dashboard should not duplicate or present substantially the same information to what already exists in other dashboards.
- Any technical information and notes should go in an appendix.
- Limit the amount of information presented to one narrow, logical subject. We encourage hiding extraneous information as that can be viewed in the source notebook if necessary.
- Make the visualization as approachable as possible. Explain things in plain English. Font sizes should be easy to read.
- Put your main visualization at the top of your notebook. The methodology or "how to" should either be in footnotes or hidden from view.
- Dashboards that present a predictive model will attract additional scrutiny and review. If models are too complex to be understood, or not presented clearly - we may decide it is not appropriate for this site. Furthermore, analysis containing predictive models will be displayed with appropriate warnings to set expectations.