covid19

covid-19 extracts and transforms covid-19 time-series data, with the goal of visualising it in an intuitive and performant way.

Data source: github.com/datasets/covid-19/data/time-series-19-covid-combined.csv

Target: JSON or Proto with time series data of covid stats per country optimised for size.

Prerequisites

• go,
• golangci-lint,
• GNU make
• node v14.7.0
• yarn v1.22.4

Development

• Build with make
• View build options with make help

Data visualisation notes

• simple time lines:
  • cumulative/new cases/deaths
  • rolling 3/5/7 day average
  • log or linear y-axes
  • absolute and per 1M/10K population number
  • choose country/ies (drop-down/search; keep pref in localStorage)
• bar charts, maybe stacked:
  • ranking of countries by monthly/weekly/rolling-month highest cases per capita
  • see https://twitter.com/RARohde/status/1289562024830173185?s=19
• geography:
  • selection on map
  • Austrialia: state, lga, suburb selection
• data:
  • compact JSON encoding: ~120K gzipped for all countries (too much)
  • try protos
  • try country pre-selection if needed