Introduction

The data is from Johns Hopkins University(JHU) (see https://github.com/CSSEGISandData/COVID-19). We are here looking at just the top 30 countries as measured by the cumulative number of confirmed cases in the country on December 1, 2020. The 30 countries include a population of ~3.8 billion. We are interested in looking at the data via different metrics. These include looking at both confirmed cases and deaths, both cumulatively and daily, and also normalized by dividing by the populations for the various countries, as well as the impact of 7 day average smoothing of the data. The charts are created in Excel. using the Perl Excel:Writer package (https://metacpan.org/pod/Excel::Writer::XLSX). The Excel data and charts files can be found at https://www-iepm.slac.stanford.edu/pinger/excel

Results:

Cumulative Confirmed casesCumulative Confirmed cases per million population


Cumulative Confirmed cases per million populationCumulative Deaths per million population



Smoothed 7 day average Daily Confirmed cases per million populationSmoothed 7 day average Daily deaths per million population



Cumulative Ratio of deaths/confirmed case



Daily confirmed cases 7 day average smoothed daily confirmed cases

Observations:

  • There is a large difference in the cumulative confirmed cases (and deaths)  versus the cumulative confirmed cases (and deaths) per million population. In particular if one does not normalize the cases by the population then the worst-off countries in order are the US, India, Brazil, Spain and Russia. Once one normalizes the cases by the population then a different set of countries appear as leaders, in order: Belgium, Czech Republic, US, Israel, and Spain.
  • The 'Cumulative ratio' shows that the ratio of deaths/confirmed has dropped dramatically since the early days (April-May 2020*). Mexico stands out in that the ratio has only dropped by about 10% while other countries have dropped by up to a factor of 10 or more.
  • There is a big difference between the 'Daily confirmed cases' and the '7-day average smoothing daily confirmed cases', in particular removing the jitter caused by the irregular/uneven reporting of Covid-19 data (e.g. missing data at weekends that is added on the following days) makes the un-normalized data very spiky and harder to view for trends.

'* Note we do not show the ratios for the early days due to the limited statistics causing wild fluctuations in the daily ratios.







  • No labels