Messing around charting coronavirus data (mainly global and UK case counts) and some other things.

Selected output at https://timday.github.io/coronavirus/, updated most days.

Contents

Processing relevant to each area:

• global
• uk
• usa

In each folder, ./fetch.sh updates daily-updated data (see code for the origins of other static csv files also in the repo), ./plot-*.py generate various charts (saved to ./output/ as well as displayed), ./publish.sh updates them to the GitHub Pages' docs directory.

• docs - content for GitHub Pages website at https://timday.github.io/coronavirus/

• experiments - Mainly playing around with DDEs for (largely abandoned) model-fitting projections.

TODO

• Move "active cases" to a front-loaded sawtooth distribution, 14-28 days duration.
• "Active cases" for UK UTLAs.
• Do something with Oxford stringency index data (CSV timeseries available)? Interested in e.g impact on death rate, time from onset to peak active cases.

Note to self

https repo access: Avoid username prompt on pushes with git remote set-url origin https://<username>@github.com/timday/coronavirus.git