useR! 2025

ellmer

Call LLM APIs from R

positron

Positron, a next-generation data science IDE

Shiny

Small boosts here and there - Simon Couch

Rather than writing an entire R package or carrying out a data analysis in one fell swoop, I’m interested in large language models (LLMs) doing things for me that I don’t like to do: tedious little refactors, transitioning from deprecated APIs, and templating out boilerplate. Chores, if you will. This talk introduces chores, an R package implementing an extensible library of LLM assistants to help with repetitive but hard-to-automate tasks. I’ll demonstrate that LLMs are quite good at turning some 45-second tasks into 5-second ones and show you how to start automating drudgery from your work with a markdown file and a dollar on an API key

Simon Couch

The Future of Asynchronous Programming in R - Charlie Gao

Asynchronous programming can be a powerful paradigm, whereby computations are allowed to run concurrently without blocking the main session. It is an opportune time to survey the current landscape, as R infrastructure in this respect has matured significantly over recent years. Instead of running a script sequentially from top to bottom, logic that takes a long or unpredictable amount of time to complete may be offloaded to different R processes, possibly on other computers or in the cloud. In the meantime, the main session may be running constantly and non-interactively, performing operations in real time, synchronizing with these tasks only when necessary. This style of programming requires a very specific set of tooling. At the very base, there is an infrastructure layer involving key enabling packages such as later and mirai. It will be explained at a high level why these two packages together currently offer the most complete and efficient implementation of async for the R language. There are further tools which expand async functionality to cover specific needs, such as the watcher package for filesystem monitoring. There are then a range of tools built on top of these, bringing async capabilities to the end-user, such as the httr2 package for querying APIs and the ellmer package for interacting with LLMs. In addition to these existing tools, exciting developments in asynchronous programming are just around the corner. These will be previewed, together with speculation on what might be possible at some point in the future

Charlie Gao

A first look at Positron - Julia Silge

Description Positron is a next generation data science IDE built by the creators of RStudio. It has been available for beta testing for a number of months, and R users may have wondered if they should try it or if it will be a good fit for them. This new IDE is an extensible tool built to facilitate exploratory data analysis, reproducible authoring, and publishing data artifacts, and it is an IDE that supports but is not built only for R. How should an R user think about Positron, compared to the other options out there? In this talk, learn about how and why Positron is designed the way it is, what will feel familiar or new coming from other IDEs such as RStudio, and when (or if) people who use R should consider giving it a try. You’ll hear about different choices when it comes to defaults and ways of working, such as how to think about your projects or folders and how to manage multiple versions of R. You will also learn about new functionality for R users and package developers that we have never had before, like new approaches for managing R package tests and the ability to customize an IDE using extensions. If you are curious about Positron and how it fits into the R ecosystem, you’ll come away from this talk with more details about its capabilities and more clarity about whether it may be a good choice for you. Additional Material or Paper *Visit https://positron.posit.co for documentation and installers *Find us on GitHub at https://github.com/posit-dev/positron *Positron is currently available on Posit Workbench in preview

Julia Silge

Keynote: I Wrote This Talk with an LLM - Hadley Wickham

In this keynote, I’ll explore the evolving relationship between data scientists, statisticians, and large language models through a unique experiment: this entire talk was created in collaboration with an LLM. From outline to slides, from code examples to key insights, I’ll share the practical realities of using AI as a thought partner in the R ecosystem. Drawing on my experience developing tidyverse packages and teaching data science, I’ll demonstrate how LLMs can augment (rather than replace) the R user’s workflow. We’ll examine specific examples where AI assistance shines—rapid prototyping, documentation generation, and creative ideation—alongside areas where human expertise remains irreplaceable. Most importantly, I’ll reflect on what this experiment reveals about the future of our community: How might AI change the way we teach R? What new skills should we prioritize? And how can we ensure that the tools we build remain accessible and empowering for all users? Join me for this meta-exploration of AI’s role in our work, with honest reflections on both the promise and limitations of these new collaborators in our statistical computing journey. This abstract was generated by Claude Sonnet 3.7 and lightly edited by me. I used the prompt: I am Hadley Wickham, chief scientist at RStudio/Posit and I’ve been invited to give a keynote on AI at the useR conference. Please write a talk abstract for a talk entitled ‘I wrote this talk with an LLM’ https://hadley.nz/

Hadley Wickham

Leveraging LLMs for student feedback - Mine Centinkaya Rundel

A considerable recent challenge for learners and teachers of data science courses is the proliferation of the use of LLM-based tools in generating answers. In this talk, I will introduce an R package that leverages LLMs to produce immediate feedback on student work to motivate them to give it a try themselves first. I will discuss technical details of augmenting models with course materials, backend and user interface decisions, challenges around evaluations that are not done correctly by the LLM, and student feedback from the first set of users. Finally, I will touch on incorporating this tool into low-stakes assessment and ethical considerations for the formal assessment structure of the course relying on LLMs