Covid-19 Simulator - Prediction

Inspiration

As the Fall 2021 term is starting, the Covid-19 pandemic has become more severe over the past few months. As a group of students, we are stressed about universities' reopening plans, and wish to predict the future trends of the pandemic in the most populated countries.

What it does

The Covid-19 Simulator fetches live data about active cases and vaccination records in each country at the begin of the execution, combining with pre-existing data on population density, elder ratio, number of ECMOs, hospital and ICU capacity, and mask enforcement about each country, to establish a sophisticated model on predicting Covid-19 trends in the future.

How we built it

C++ is used to design the body of the simulator (process/format input/output, simulate the pandemic). C, JavaScript, and Python are used to scrape data from websites on active cases and vaccination records.
Details on calculating new infections, recovers, and deaths are below:
Number of new infections per day =
(Basic Reproduction Number * Population Density ^ 1/4) / (420 * Mask Enforcement * Vaccinated Ratio)
Number of severe cases =
Active cases * Vaccinated Ratio * Protection Rate + Active Case * Unvaccinated Ratio * Possibility of Covid-19 Turning into Severe Cases
Number of people recovered per day =
Active Cases * Ratio of people that hospitals are available to / 21 + Active Cases * Ratio of people that hospitals are not available to / 42;
Number of deaths per day =
(Number of severe cases - Number of ICU - Number of ECMO) * Possibility of Deaths from Covid-19 + Number of ICU * Possibility of Deaths from Covid-19 in ICU + Number of ECMO * Possibility of Deaths with ECMO

Challenges we ran into

Web scraping is an unfamiliar concept to all of us. We spent time learning how to scrape data off websites using C, JavaScript, and Python, and process data into forms that can be fed directly into the simulator. Especially when Chinese is contained in the scraped data, the file will contain corrupted characters. We resolved the issue by feeding the data directly into the simulator without outputting to a file beforehand.

Accomplishments that we're proud of

We are able to make predictions about the Covid-19 pandemic and possibly advise the university to make changes about their reopening plans, such as strictly enforcing masks in classrooms.

What we learned

Which parameters affect Covid-19 spread and how to make an algorithm to predict a pandemic
How to use C, JavaScript, and Python to write web scrapers

What's next for Covid-19 Simulator

We will explore more potential parameters that are crucial about pandemic prediction, such as Covid variants and their different infectious rates. We will also conduct more research on designing algorithms, making them more comprehensive.