Learning R and dogs – An experience from a hackathon

Yes, you read it correctly. In what wonderful universe are dogs associated with learning the programming language R? I am sure we have all experienced some degree of despair when learning or working with R and would have loved to cuddle a dog for that feeling of unconditional support and affection. Unfortunately, our everyday working is usually far removed from our dear four-legged friends. Luckily for me, my experience last year brought these two unlikely elements together when I attended a one-day hackathon hosted by the Dogs Trust charity. Yes, I worked with individual dog-level data.

Founded in 1891, Dogs Trust is the longest-standing and largest dog welfare charity in the UK. Over the years, they have collected a wealth of data on all the dogs that entered and passed through their 21 shelters. Usually, a hint of what this data looks like can be found in their annual reports, but for one day only we had access to it all, which included information like breed, gender, size, colour, entry and exit dates, age, and…wait for it…names! GDPR does not apply to these furry beings.

The aim of the day was to help the charity learn from their data. They wanted to know whether the data could provide insights that could be used to improve their rehoming strategies. My group used a combination of mapping and epidemiological methods to identify patterns in breed popularity. We identified drastic changes in the last two decades: a rise in preference for small dogs like Terriers and Chihuahuas (should we name it Paris Hilton effect?) over traditionally popular breeds like the Border Collie and an explosion of Huskies in recent years. Not surprisingly, size and breed were associated with time taken to adoption – Huskies were one of the fastest breeds to be adopted but, unfortunately, also the fastest to be returned.

These findings affected me on a personal level and had an impact on the way I looked at dogs – cute commodities characterised by features distinct to each breed. In numbers and data, I saw that our preferences, influenced by social media and other surroundings, may have a direct effect on the types of breeds of dogs that come into care. Regrettably, many of these dogs are subsequently put down due to the overwhelming demand for shelters. For those working at Dogs Trust, however, the results from the hackathon provided them with various ideas regarding more effective ways for re-homing dogs .

Overall, how much R coding did I learn from this experience? I had just embarked on my R journey (3 months or so – a novice at best) and had never experienced a hackathon. I really did not know how I could contribute or what I would get out from it. Unsurprisingly, my R skills did not suddenly expand, but what I did learn was how to work as a team under extreme time pressure to produce a presentable, albeit work-in-progress, output. Teamwork, communication, and leadership, cliché as these skills are, are paramount to a hackathon – and can be applied to life in academia too. Additionally, it was interesting to see how other teams tackled the same question through completely different angles. For these reasons, I thoroughly recommend attending a hackathon, if only to improve your collaboration skills. Who knows what kind of experience you will end up having?

Eoin Kelleher, illustrator, is an anaesthesia trainee from Dublin, Ireland. He is currently enrolled on the MSc Global Health Science & Epidemiology programme. In his spare time, he is an amateur illustrator. He has just published an illustrated book of the human body, Journey Through the Body.