It merges the following datasets for each country into one file:
- Covid-19 tests, cases, and deaths data from https://github.com/owid/covid-19-data/tree/master/public/data
- A Covid-19 "age adjusted risk" factor, as calculated by using population pyramids from https://www.populationpyramid.net/ in conjunction with data on the spread of Coronavirus deaths by 10-year age group.
- Percentage of people that are obese (BMI >30) per country, as obtained from the WHO website
- GDP per capita (PPP) data from the IMF
- Population density, and an inequality of density distribution measure* as calculated by Differential Capital (country_density_dat.csv)
- Government response to Covid-19 data from https://www.bsg.ox.ac.uk/research/research-projects/coronavirus-government-response-tracker
And creates a number of new fields, based on the above data:
- days_since_outbreak - days since first case was registered in the given country
- days_since_deaths_crossed_0.1_per_million
- days_since_10_deaths
- gov_response_at_0.1_deaths_per_million - StringencyIndex at the point where Deaths-per-million crossed 0.1
- deaths_5_day_averages, infections_5_day_averages - Average of deaths per day and average of cases per day over previous 4 days and current day.
- peak_deaths_already_occurred_indicator (& peak_cases_already_occurred_indicator) - 1 if deaths_5_day_averages (& infections_5_day_averages) had its maximum at least 4 days before last day of data, 0 otherwise.
If you wish to update the data yourself to include new data from sources (1) and (6), simply run the first 3 scripts in this repository (Data Retrieval, Data Cleaning, Identifying Stages of Pandemic by Country). This will fetch the latest data from those sources and recalculate (7) - (12).
*This inequality measure (ineq) ranges from 0 to 1 in country_density_dat.csv, with 0 being completely equal spread and 1 being completely unequal spread. In full_dataset_for_modelling.csv this measure has been transformed to be 1 / (1 - ineq).
This script shows the results of some linear models that attempt to explain the cumulative number of deaths at the time of the peak for a given country, using some combination of fields from our dataset.
This is up to date as of 19 May at 11:30am
This file compares the current total number of deaths in each country to the final number of deaths expected based on data from the Diamond Princess cruise ship.
If you wish to update this file to include the latest data, you must run scripts "01 - Data Retrieval.rmd", "02 - Data Cleaning.rmd", and then "0x - Diamond Princess.rmd".