On Thursday 21st January rstudio::global kicked off a 24-hour long virtual data science conference. Over 16,000 people registered to join what was sure to be one of the highlights of the data science calendar in 2021.
As a RStudio Full Certified Partner, Mango were proud to be a sponsor of the conference and enjoyed talking to attendees in our sponsorship gallery. If you would like to hear the highlights, please register to join the next LondonR – where Rachael Dempsey from RStudio will be re-capping the conference on 2nd February.
Conference highlights and themes
rstudio::global opened with a keynote from their Chief Scientist, Hadley Wickham. Hadley shared with the audience how the tidyverse has evolved, the project’s greatest successes and how the team learnt from failures. He also discussed how the thought processes of CBT have helped him be more productive – something which resonated with the audience, with many people sharing on social how much they related to these kinds of thoughts.
Most importantly, during the Q&A section, Hadley revealed it takes him 20-30 minutes to perfect his famous bowtie!
Fellow keynote speaker John Burn-Murdoch is a senior data visualisation journalist at the Financial Times. Whilst already well known to the EARL Conference audiences from his past presentations at our London conference, John’s fame spread has worldwide as a result of his influential charts visualising the spread of COVID-19. His graphs had to communicate complicated and evolving data to a mass audience and John spoke about what he had learned along the way. His number one point was that text is critical. In studies showing where viewer’s eyes tracked to when faced with a chart, the first and most important place was the title. “Text is critical” he said, “words feature predominantly so text and annotation add enormous value, particularly to ‘non chart’ people”. The theme of storytelling was also important as readers want to be able to see and understand the story and animation can help this process. A further key take away was not to “publish and vanish” but to stick around and interact with viewers. John noted that feedback on his charts via Twitter was enormously helpful as was using this medium to be able to debate and discuss his approach and methodology. It was important to remember, he said, not to be “too smart” and that charts need to be “accessible” for people not familiar with viewing charts. John noted that there were plenty of resources available to the dataviz community around creating charts, but little around communicating the message. This was a thoroughly engaging talk with lots of useful tips for anyone who creates charts to visualise their data.
A popular theme of the conference was the pairing of R and Python. Sean Lopp’s talk ‘R & Python: Going Steady’, aimed to debunk the myth that your team can only use one tool and inform data scientists how they can use both languages successfully.
“Optimise for people, not for tools” – Sean Lopp
Data for good was another theme that stood out at rstudio::global. The virtual nature of the conference meant that speakers could present from across the globe, one such speaker was Shelmith Kariuki based in Nairobi, Kenya. Shelmith is currently working as a Data Analytics Consultant with UN DESA and her talk covered the ‘rKeynaCensus’ package, which contains the results of the 2019 Keyan population census – the first paperless census to be conducted in Kenya.
If you enjoyed rstudio::global along with the other 16,000+ attendees, then join our next LondonR session where we will share further highlights.