My passion for data visualization was piqued by the New York Times’s coverage of the 2012 election, which incorporated data visualization in new, industry-defining ways. Around the same time, I was taking my first statistics course in high school, and struggled to understand statistics through numbers alone. I found that by visualizing the distributions at first with my graphing calculator and later with Python visualization libraries, it was much easier to wrap my head around the material.
Since then, I’ve visualized data for many reasons, whether it be as a supplement to some data analysis task, or to better explore some metric in a project I’m building. I began with tools like Tableau, and slowly ventured into matplotlib, ggplot, and D3. In one particular case, at a hackathon, I wrote a real-time computer vision application that would convert a hand drawn chart into one expressed in D3.
During my time at Midnight Labs, I used D3 to create new, data-driven front-end experiences for various marketing pages. That summer, my familiarity with the intricacies of D3 really flourished. I learned about the data join model and learned how to use D3 to make living designs powered by data, instead of the more conventional chart. This was the first time where I worked directly with engineers who were masters of the craft and gave me mentorship.
I dedicated myself to learning more, and dove into Tufte’s famous literature and involved myself with the UW’s Interactive Data Lab, which I’ve been a member of ever since. There I began helping graduate students Kanit Wongsuphasawat and Dominik Moritz build Vega-Lite, a succinct but expressive grammar on top of Vega. I believe the Vega and Vega-Lite grammars are will change the data visualization game by making the creation of charts more declarative, more portable, and more generatable. It’s easily one of the most exciting projects I’ve ever worked on.
My data visualization experience as a research assistant in the IDL helped me land an internship at the Harvard Kennedy School’s Center for International Development where I spent three months exploring the capabilities of data storytelling as a medium for explaining conventional economic research. We worked to take an important research conclusion from a journal paper and turn it into an explorable tactile experience, in the hopes of reaching a broader audience.
As I mentioned previously, my interest in data visualization was sparked by the 2012 New York Times election coverage, particularly that of FiveThirtyEight–then a sub-blog of the Times. I continued to follow FiveThiryEight as it grew to be an independent powerhouse in data-driven political coverage, and eventually came full circle in my data visualization story when I interned there during the 2016 presidential election as a visual journalist. From that experience, I learned how to use data visualization to take otherwise unrelatable, unparseable data and make something people can explore, learn from, and empathize with.