Ever since starting work, I hadn’t imagined having another chance to take part in a hackathon/datathon again. As part of my engineering orientation week, however, I found myself grouped with four other campus hires and tasked with delivering new insights on the coronavirus.

COVID Data, Again??

At first, I didn’t feel especially excited, because I felt like COVID was already so overly analysed, visualised, and the whole data science world had already gone in to pick all the low-hanging fruits. By now, there must be some thousands of prediction models for the spread of the coronavirus, especially with Kaggle’s COVID-19 Open Research Dataset Challenge.

Maybe that’s an exaggeration, but it sure feels that way if you just do a quick Google search. What’s my team going to do, then?

It’s in the title

After multiple brainstorming sessions, we split into pairs, investigating the effects of COVID-19 on human behaviours and our climate, respectively.

Human & Nature

Across the world, governments have implemented traveling restrictions and many countries are still under lockdown. People’s lifestyles have changed during these past months. Our team was interested in studying the impacts of COVID-19 on our planet, in terms of human behavior and climate changes. We hypothesized that the pandemic would have effects on various aspects of societal life such as daily commute, consumer purchases all around the world, and as a consequence, also an effect on the global climate.

Datasets

• World Health Organization Data from GS Quant
• Google Community Mobility Reports from Google
• Retail Sales Index (Singapore) from Department of Statistics, Singapore
• Tourism Sector Performance Report (Singapore) from Department of Statistics, Singapore
• PM2.5 Data (Major Chinese Cities) from Air Quality Open Data Platform
• Temperature Data (Major Chinese Cities) from Air Quality Open Data Platform

Results

I wanted to go into more detail on some of our findings, but unfortunately I’m unable to include a screenshot of our data visualisation dashboard or our charts. You can see why that makes things really difficult. Hopefully, the information can be unclassified…

<Watch this space!>

Conclusion

Our investigation ultimately drew a strong relationship between COVID-19 and mobility across the globe. In Singapore, we saw a decline in sales for non-essential goods like jewellery, clothings, and apparels when approaching the lockdown period. Unsurprisingly, the sales for supermarkets, hypermarkets, mini-marts, and convenience stores rose in contrast to the latter as people stocked up more essential goods in fear of a complete lockdown. There was also a noticeable effect on air quality as measured by PM2.5, supporting our postulation that reduced human activty had a significant impact on the environment.

To be honest, I was really impressed with some of the other teams from APAC, so it came as quite a surprise to have passed both rounds of judging to represent APAC before a panel of executives next week.