Mini project: web scrapping and analysing COVID-19 data
A short web scrapping and model building project
With a second, more severe wave of COVID-19 hitting my country (Greece) I was left wondering how the pandemic will unfold. Our National Public Health Organisation (EODY) daily publishes the number of confirmed COVID-19 cases. This number increased dramatically the last month along with the number of COVID-19 patients in ventilators and deaths.
The first thing that one might have is to use the time series of confirmed coronavirus cases to predict the future. The problem with this approach is that this number is dependent on the number of COVID tests that are performed. A number that varies significantly from day to day and that is not published daily. On the other hand, the number of COVID patients in ventilators and of course the number of deaths is published daily. Hence, I decided to use the first one to try to predict the second.
The trouble is that what is being published is the actual announcement of the National Public Health Organisation (EODY) on their web page. There are no APIs, excel files, or whatever that would be easy to process (Johns Hopkins University gathers COVID-19 data from many countries/areas including Greece but has only the number of confirmed cases, number of deaths and number of recovered confirmed cases). My solution was to…
