Kun Chen, Linglong Kong, Sam Tyner-Menroe, Jun Yan, and Hua Zhou

The ASA Section on Statistical Computing held a half-day online symposium, titled “Statistical Computing in Action,” November 12, 2022. The event attracted more than 110 participants, about half of whom were students. Held as part of the section’s membership drive, the symposium was designed to showcase the power and beauty of statistical computing to students, help practitioners sharpen their statistical computing skills, and foster community across different computing languages.

The symposium started with a brief background introduction of the event by the section’s 2022 chair, Jun Yan. ASA Executive Director Ron Wasserstein then delivered opening remarks on behalf of the ASA leadership and reminded the audience that 2022 marked the 50th anniversary of the Section on Statistical Computing. “The organizers have ambitious goals and assembled a fabulous team of presenters for a real treat,” said Wasserstein.

The symposium featured a keynote talk, data jamboree, and panel discussion. Douglas Bates, professor emeritus of the University of Wisconsin-Madison, gave the keynote, titled “Cross-Language Technologies for Statistics and Data Science.” He shared the following three tools he has used with R, Python, and Julia that make the transition between languages easier:

1. Arrow storage format, a language-neutral format for storing and easily reading tabular data
2. Quarto, which is an RMarkdown format and processor for literate programming
3. VS Code editor, which provides editing and code evaluation for all three languages and Jupyter notebooks

The data jamboree, chaired by Lucy D’Agostino McGowan of Wake Forest University, was a party of computing tools for solving the same data science tasks. Josh Day of Julia Computing led the Julia workshop, Dan Chen of the University of British Columbia led the Python workshop, and Sam Tyner-Monroe of Tritura led the R workshop. All three workshop leaders independently uncovered the same problems in the data, and each showed a unique perspective for how they approached the analysis. There was strong agreement among all presenters on the ‘trust but verify’ principle in data processing and modeling. The main data set was a subset of the New York City motor vehicle collisions data from NYC open data. The tasks were adapted from an introduction to data science course taught by Jun Yan at the University of Connecticut.

Kun Chen of the University of Connecticut moderated the panel discussion, titled “Frontliners and Next Frontiers of Statistical Computing in Data Science.” The panelists were Hannah Frick of Posit, Haoda Fu of Eli Lilly, Eric Kolaczyk of McGill University, and Teresa Sönmez of 23andMe. They discussed career development and the future of statistical computing and shared their experiences as statisticians and data scientists, including their work and research; skills for career success; differences in different career paths; diversity, equity, and inclusion; and specific suggestions for job searches and interviews. The panelists also shared their predictions and hopes for important research areas, available career opportunities, and computing tools for statistical computing in the next 5–10 years.

The symposium generated many volunteer opportunities for section members and served as an interface for members of different computing communities to exchange ideas.

Videos of the section’s events are available on the section’s YouTube channel.