How hard is it to create a simple Covid-19 tracker? This question came to my mind this week-end. Would it be possible to create a Covid-19 tracker just in a couple of hours? So I started a coding experiment, started up a Jupyter notebook and looked for some publicly available data. I found this git repository with data from Johns Hopkins University:

https://github.com/CSSEGISandData/COVID-19

This repository has aggregated Covid-19 stats data from almost all countries world wide. Much of the data in this Git repository is very often not corrected or interpolated so you might have data that varies in nonsensical ways like in this graph below:

Or sometimes you find countries that have no data at all or simply very little data too:

But ok, there are also countries that have very sound data in this repository like Germany:

All in all it is possible to create a script which gets the data from John Hopkins University, reads it, concatenates it into a dataset (there are multiple files CSV files per country in the repository) then generates a simple webpage with charts for each country displaying like you can see above the confirmed cases, deaths, recovered cases and active cases.

The end product of this script (or better of this Jupyter notebook) is a simple webpage that looks like this:

https://www.onepointltd.com/onepoint-lab/covid19/tracker.html

This is the whole code that generates the simple Covid-19 tracker:

https://github.com/onepointconsulting/datascience-experiments/blob/master/covid19/john_hopkins_dataset_creation.ipynb

The video below explains the code used to generate this dashboard. If you are interested in coding then the video is for you.