The inspiration for this project came from Graham Dodge, Co-Founder and CEO of SickWeather, who presented a pitch that would add demographic information to geotagged data points provided through the SickWeather API.

What it does

After combining census based demographics data with social media based illness reports, increased infectious disease activity, when compared to previous years, is represented as expanding circles of influence, as well as a Cold and Flu Index value. This data is represented in a movable and zoomable map page and data is constrained the state of Maryland.

How we built it

This project is built around the R programming language, which was used to organize data sets, map out layers of data, and combine these two elements into visual interpretation of possible vectors for transmission of infectious agents. Demographics data was gathered from publicly available Census data and the SickWeather API data was called and parsed through Python. The files and source code are hosted through AWS.

Challenges we ran into

A major challenge we ran into was our inability to create a dynamic front-end client and back-end server to hold and represent our project. Given our collective lack of web development experience, we resorted to a static representation of data collected and analyzed. This is not significantly detrimental to the functionality of our project, since census data consists of reports from the last 5 years, and is not a rapidly changing data set. Data collected from the SickWeather API includes a couple weeks back of data, and relies on the aggregration of many individual markers.

What's next for SickHacks

The next step for this project would be to combine the information and layout displayed here with the released SickWeather App. Other considerations include increasing data input by exploring different sources of information, such as clinical data.


Special thanks to Erich Walker, Director of Information Technology at PPMS, who provided us with technical expertise and a familiarity with the SickWeather App.

Built With

Share this project: