What makes a great R package? – Joseph Rickert

February 27, 2018

Abstract

With well over 11,000 packages on CRAN, it is clear that countless hours have been devoted to the R project by some of the finest minds engaged in statistical computing. Collectively, the R package system has been an unqualified success. But not all packages are equal. Presumably most packages are at least marginally useful, but some appear to have been forgotten. Some appeal to a wide range of users, while others are brilliant but unfathomable to anyone outside of an arcane area of science or statistics . In this talk, I will explore the qualities that have elevated some package beyond mere utility and made them valuable to large numbers of users. I will offer some examples of what I consider to be great R packages, argue my case, and offer some advice to package authors and discerning package users.


About the speaker

Joseph Rickert
RStudio Community Ambassador

Joseph is RStudio’s “Ambassador at Large” for all things R, editor of the R Views blog, and RStudio’s representative on the R Consortium’s board of directors. Joseph came to RStudio via Revolution Analytics and Microsoft where he was a data scientist, blogger and community manager. Before joining Revolution Analytics Joseph worked as a statistician for a small healthcare economics consulting firm, and before that he held a variety of technical, marketing and sales positions while working for startups that spanned multiple industries including government contracting, local area networks, disk drives and test equipment. Joseph studied Classics and Mathematics as an undergraduate at Franklin & Marshall College and earned an M.A. in Humanities and an M.S. in Statistics from the California State University. You can follow him on Twitter as @RStudioJoe.

Previous Video
You can make a package in 20 minutes – Jim Hester
You can make a package in 20 minutes – Jim Hester

No More Videos