Talk recordings and workshop materials from posit::conf(2023)

Want to suddenly overhaul that one package you wrote two years ago? It probably won’t work when your colleagues try to update it.

Your new teammates are data.table fans, but you prefer the tidyverse. Do you need to refactor? Are style choices, like indentation important when collaborating, or are you just being persnickety?

Keys to good 1-on-1s start with a small amount of prep and a running shared document of notes and takeaways. Another key is to rotate types of 1-on-1s. Possibilities include “heads down” on recent work, “heads up” looking further out, and career-focused sessions. After some tips on the right sort of questions and uncovering sneaky issues, I will also touch on effective feedback.

To address these challenges, we established the Pfizer R Center of Excellence (CoE) in early 2022. We focus our efforts on bringing together a rapidly growing community of colleagues, providing technical expertise, and offering best-practice guidance. A well-established, maintained and engaged R community promotes an inclusive and supportive learning environment that drives innovation within organizations. Our aim is to help colleagues thrive in their R journey, regardless of their expertise level.

In the life sciences, whether it’s pharma, biotech, research, or another type of organization, we are unique in that we blend scientific knowledge with technical skills to extract insights from large, complex datasets. In the cloud, we can architect solutions to help us scale, automate, and collaborate. Interestingly, the use of R and Python by bioinformatics, genomics, biostatistics, and data science teams can be challenging in a cloud-first world where all the data is somewhere other than your laptop (like a data lake).

In this talk, I will share best practices and lessons learned surrounding the use of R and Python by technical teams in the cloud. We’ll focus on the use of Posit Workbench and RStudio on various cloud services such as Azure ML and Databricks.

Unlock the potential of Shiny apps for your organization! Join Tiger as he shares insights from implementing Shiny apps at his workplace, handling over 160,000 internal requests. Discover a practical mindmap to find, build, and enhance Shiny app use cases, ensuring robustness and improved user engagement.

Data science sure has changed over the past few years! Everyone’s talking about production. RStudio is now Posit. Models are now tidy.

This talk is about embracing that change and updating existing models using the tidymodels framework. I recently completed this change, letting go of our in-production code and revisioning it with tidymodels. My team ended up with a faster, more scalable pipeline that enabled us to better automate our workflow and increase our scale while improving our stakeholders’ experiences.

I’ll share tips and tricks for adopting the tidymodels framework in existing products, best practices for learning and upskilling teams, and advice for using tidymodel packages to build more accessible data science tools.

In this talk, learn about model drift, the different ways the word “performance” is used with models, what you can monitor about a model, how feedback loops impact models, and how you can use vetiver to set yourself up for success with model maintenance. This talk will help practitioners who are already deploying models, but this is also useful knowledge for practitioners earlier in their MLOps journey; decisions made along the way can make the difference between resilient models that are easier to maintain and disappointing or misleading models.

This talk describes how to make distribution-free prediction intervals for regression models via the tidymodels framework.

By creating and deploying an interactive web application you can better share your data, code, and ideas easily with a broad audience. I plan to talk about several Python web application frameworks, and how you can use them to turn a class, function, or data set visualization into an interactive web page to share with the world. I plan to discuss building simple web applications with Plotly Dash, Streamlit, and Shiny for Python.

Materials:

There are a bunch of reasons why apps fail to deploy. They can be easy to fix, like Hardcoded secrets, fonts, or missing libraries. Or they can be intractable and super frustrating to deal with, like manifest mismatches, resource starvation, and missing libraries.

Ideas from home organization shows are surprisingly applicable to file management. Using a room divider to establish dedicated zones for different activities in a studio apartment is analogous to creating self-contained projects in RStudio. Likewise, swapping mismatched hangers with matching ones to tidy a closet resembles the adoption of a file naming convention to make a directory easier to navigate.

In this talk, I will share good practices in file management through the lens of home organization. We all know that clutter, whether it is in our physical space or on our machine, destroys our ability to focus. These practices will help R users of all levels create a serene, relaxing environment where they feel inspired to work with data.

I’ll share about documenting things, and how intentional approaches to documentation and onboarding are helping big distributed teams tackle hard challenges and change organizational culture. The goal is to provide concrete tips to help you document things effectively & hear stories of how putting a focus on documentation can be help teams be efficient, productive, and less lonely. I’ll give a short lightning talk (inspired by Jenny Bryan’s Naming Files talk) followed by stories from NASA Openscapes, NOAA Fisheries Openscapes & beyond.

Materials:

This is a talk about environment management, but not in the way you’re thinking. In many industries, domain-specific experts need enough understanding of data science to support their work and communicate with data scientists, but often have insufficient training in these skills, and limited time with which to obtain data science skills and practice them. This is particularly challenging for those who are interdisciplinary and have limited control over their workload, such as medics and field scientists. In this talk, an educational scientist describes the previous 10 years of supporting veterinary scientists to adopt open science practices surrounding data science. What worked, what failed miserably, and reflections on why it can be so hard to get a horse to drink.

Materials:

Everyone will have a different journey when becoming a data scientist. However, there are a few tips to consider to make the journey less daunting and more enjoyable. Listen, as I tell my story as a data scientist and offer resources and tips to build confidence for those who are new to their journey. The tools are available however, it is not always easy to find them.

keywords: openscience, The Carpentries, R programming language, GPS, data science journey, data science resources

Materials:

dplyr’s select helpers exemplify how the tidyverse uses opinionated design to push users into the pit of success. The ability to efficiently operate on names incentivizes good naming patterns and creates efficiency in data wrangling and validation.

However, in a polyglot world, users may find they must leave the pit when comparable syntactic sugar is not accessible in other languages like Python and SQL.

In this talk, I will explain how dplyr’s select helpers inspired my approach to ‘column name contracts,’ how good naming systems can help supercharge data management with packages like {dplyr} and {pointblank}, and my experience building the {dbtplyr} to port this functionality to dbt for building complex SQL-based data pipelines.

Materials:

DuckDB is an in-process analytical data management system. DuckDB supports complex SQL queries, has no external dependencies, and is deeply integrated into the R ecosystem. For example, DuckDB can run SQL queries directly on R data frames without any data transfer. DuckDB uses state-of-the-art query processing techniques like vectorised execution and automatic parallelism. DuckDB is out-of-core capable, meaning that it is possible to process datasets far bigger than main memory. DuckDB is free and open source software under the MIT license.

In this talk, we will describe the user values of DuckDB, and how it can be used to improve their day-to-day lives through automatic parallelisation, efficient operators, and out-of-core operations.

Materials:

duckdb is the new high-performance analytical database system that works great with R, Python, and other host systems. dplyr is the grammar of data manipulation in the tidyverse, tightly integrated with R, but it works best for small or medium-sized data. The former has been designed with large or big data in mind, but currently, you need to formulate your queries in SQL.

The new duckplyr package offers the best of both worlds. It transforms a dplyr pipe into a query object that duckdb can execute, using an optimized query plan. It is better than dbplyr because the interface is ““data frames in, data frames out””, and no intermediate SQL code is generated.

The talk first presents our results, a bit of the mechanics, and an outlook for this ambitious project.

The DataFrame Decision locks you into different methods, with subtly different behavior::

• different table methods (e.g. polars .with_columns() vs pandas .assign())
• different column methods (e.g. polars .map_dict() vs pandas .map())

In this talk, I’ll discuss how siuba (a dplyr port to python) combines with duckdb (a crazy powerful sql engine) to provide a unified, dplyr-like interface for analyzing a wide range of data sources‚ whether pandas and polars DataFrames, parquet files in a cloud bucket, or pins on Posit Connect.

How do you create brand cohesion across your large team when it comes to data viz? Inspired by the BBC’s bbplot, our team at the US Agency for International Development (USAID) developed a package on top of ggplot2 to create a common look and feel for our team’s products. This effort improved not just the cohesiveness of our work, but also trustworthiness. By creating this package, we reduced the reliance on using defaults and the time spent on each project customizing numerous graphic elements. More importantly, this package provided an easier on-ramp for new teammates to adopt R. We share our journey within a federal agency developing a style guide and aim to guide and inspire other organizations who could benefit from developing their own branding package and guidance.

Materials:

It’s easy to think that R users do not need HTML and CSS. After all, R is a language designed for data analysis, right? But the reality is that these web standards are everywhere, even in R. In fact, many great tools like {ggtext}, {gt}, {shiny}, and Quarto unlock their full potential when you know a little bit of HTML & CSS. In this talk, I will demonstrate specific examples where R users can benefit from HTML and CSS and show you how to get started with these two languages.

Materials:

Whether you want your slides to stand out from the crowd, or you need your documents to fit within your corporate style guide, being able to style Quarto documents is a valuable skill.

As we move further into the intricacies, Jeremy unpacks advanced strategies for model testing and optimization, utilizing tools like GPTQ and Hugging Face Transformers. He also explores the potential of specialized datasets like Orca and Platypus for fine-tuning and discusses cutting-edge trends in Retrieval Augmented Generation and information retrieval. Whether you’re new to the field or an established professional, this presentation offers a wealth of insights to help you navigate the ever-evolving world of language models.

(The above summary was, of course, created by an LLM!)

Because the field of data science is inherently task-oriented, it is no wonder that most people struggle to see the utility of generative art past the bounds of a casual hobby. This talk will invite the participant to learn about generative art while focusing on “why” people should create it and its potential place in data science. This talk is suitable for all disciplines and artistic abilities. Furthermore, this talk will aim to expand the participant’s perspective on generative art with the following concepts:

• What is generative art and how can it be created in R or Python
• Justifications for generative art within Data Science
• Examples of programming skills that are transferrable between generative art and pragmatic data science projects

Materials:

So the justice department’s back office monitors the smooth processing of all traffic fines in Belgium. They gather that data from all police departments and check if any anomalies occur.

Shiny recently celebrated its 10th birthday, and since its birth, has grown tremendously in many areas; however, a hello world Shiny app still looks roughly like it did 10 years ago. The bslib R package helps solve this problem making very easy to apply modern and customizable styling your Shiny apps, R Markdown / Quarto documents, and more. In addition, bslib also provides dashboard-focused UI components like expandable cards, value boxes, sidebar layouts, and more to help you create delightful Shiny dashboards.

Materials:

But what if your org creates many apps?

Scaling good design to larger groups dials the challenge up to 11.

In this talk, I will share how we approach the problem at Appsilon.

R-Ladies’ first ten years were about growing the community: from being just one chapter in 2012 to becoming a global organization in 2016, and now fostering more than 230 chapters worldwide. But how can we face the challenges of growing an organization based solely on volunteer work?

In this talk, we discuss how good programming practices –such as modularity, refactoring, and testing– inspire the way we see the sustainable management of an ever-growing community. To that end, we will present our most recent efforts at creating and documenting workflows, distributing the workload, and automating tasks that allow volunteers to focus their time where it is most needed.

After watching this talk, you will get some ideas on how to support volunteers in your own community or project, and on how to use GitHub Actions to automate workflows and tasks.

In this talk, I want to highlight some strategies for keeping your data science meetup sustainable. Specifically, I want to highlight the role of self-care in growing and sustaining your group, as well as low-key activities like a data scavenger hunt, watching videos together, styling plots together, and sharing useful tidyverse functions.

In R, the fundamental unit of shareable code is the package. As of March 2023, there were over 19,000 packages available on CRAN. Hadley Wickham and I recently updated the R Packages book for a second edition, which brought home just how much the package development landscape has changed in recent years (for the better!).

In this talk, I’ll show why that distinction is not necessarily true by discussing my personal experience from learning R in graduate school to reporting bugs on GitHub to becoming a co-author of the tidycensus package and a practicing data scientist. The positive and supportive R community on GitHub, Twitter, and elsewhere contributes to why anyone who writes R code can become a package author.

Dive into the often-overlooked world of accessibility for developers and data scientists! Uncover common accessibility barriers in open source tools, and learn simple steps to address them. Whether you’re a seasoned maintainer or a total novice, you’ll walk away with clear action items to implement right away. Join the movement of individuals championing the next frontier of disability inclusion in technology, shaping a more equitable future for all!

ggplot2 is one of the most loved visualisation libraries. It implements a Grammar of Graphics system, which requires one to think about data in terms of columns of variables and how to transform them into geometric objects. It is elegant and powerful. This is a talk about plotnine, which brings the elegance of ggplot2 to the Python programming language. It is an invitation to learn about the Grammar of Graphics system and to appreciate it. It will include some tips on how to avoid common frustrations as you learn the system.

Materials:

Shiny for Python is a great niche for R users because none of the Python web frameworks can do what Shiny can do. Additionally, almost all of your knowledge of the R package is applicable to the Python one.

“Wiki”, “User Guide”, “Handbook” – whatever you call yours, we converted ours to Quarto!

A year ago, my team’s documentation, which had been created using Microsoft Word, was large and lacked version control. Scrolling through the document was slow, and, due to confidentiality reasons, only one person could edit it at a time, which was a significant challenge for our team of multiple developers. After realizing we needed a more flexible solution, we successfully converted our documentation to Quarto.

In this talk, I’ll discuss our journey converting to Quarto, the challenges we faced along the way, and tips and tricks for anyone else who might be looking to adopt Quarto too.

This talk showcases how to use packages, such as Plotly and Crosstalk, to create interactive data visualizations and add dashboard-like elements to Quarto. Using a fun dataset available through the “Richmondway” package, we examine the number of times Roy Kent uses salty language throughout all seasons of “Ted Lasso.” We illustrate this using animated plots, interactive selection widgets such as checkboxes, and by linking two plots with brushing capabilities.

Materials:

Every year, government agencies release large amounts of data on schools and colleges, but this information is scattered across various websites and is often difficult to use. To make these data more accessible, the Urban Institute built the Education Data Portal, a freely available one-stop shop for harmonized data and metadata for nearly all major federal education datasets. In this talk, we’ll demonstrate how the portal works and share lessons we’ve learned about making data accessible to users with varying technical skills and preferred programming languages.

The Cook County Assessor’s Office (CCAO) determines the current market value of properties for the purpose of property taxation. Since 2020, the CCAO has used R, tidymodels, and LightGBM to build predictive models that value Cook County’s 1.5 million residential properties, which are collectively worth over $400B. These predictive models are open-source, easily replicable, and have significantly improved valuation accuracy and equity over time.

Join CCAO Chief Data Officer Nicole Jardine and Director of Data Science Dan Snow as they walk through the CCAO’s modeling process, shares lessons learned, and offer a sneak peek at changes planned for the 2024 reassessment of Chicago.

Now, that same capability has come to Shiny for R, thanks to the webR project.

{epoxy} is a new package that uses {glue} to give authors templating superpowers. Epoxy works in R Markdown and Quarto, in markdown, LaTeX, and HTML outputs. It also provides easy templating for Shiny apps for dynamic data-driven reporting.

Soil health data are notoriously challenging to tidy and effectively communicate to farmers. We used functional programming with the tidyverse to reproducibly streamline data cleaning and summarization. To improve project outreach, we developed a Quarto project to dynamically create interactive HTML reports and printable PDFs. Custom to every farmer, reports include project goals, measured parameter descriptions, summary statistics, maps, tables, and graphs.

Our case study presents a workflow for data preparation and parameterized reporting, with best practices for effective data visualization, interpretation, and accessibility.

Over 20 years ago Joel Spolsky famously wrote, “All non-trivial abstractions, to some degree, are leaky.” Unsurprisingly this has not changed. However, we have introduced more and more layers of abstraction into our workflows: Virtual Machines, AWS services, WASM, Docker, R, Python, data frames, and on and on. But then on top of the computational abstractions we have people abstractions: managers, colleagues, executives, stakeholders, etc.

JD’s presentation will be a wild romp through the mental models of abstractions and discuss how we, as technical analytical types, can gain skill in traversing abstractions and dealing with leaks.

The dplyr 1.1.0 release included a number of new features, such as:

• Per-operation grouping with .by
• An overhaul to joins, including new inequality and rolling joins
• New consecutive_id() and case_match() helpers
• Significant performance improvements in arrange()

When an error happens in your API, Shiny App, or quarto document, it is not always clear what line of code you need to look at, and the error messages aren’t always helpful. By walking through a simple API example, I show how you can use logging packages like log4r to provide error messages that make sense to you. I also show how you can use cloud-based data collection platforms like Azure Monitor or AWS CloudWatch to set up alerts, so you can get notified by email or text message for those severe errors that you need to be immediately aware of.

Gain more visibility into the health of your code by incorporating logging and pushing your logs to the cloud.

Slow dashboards lead to a poor user experience and cause users to lose interest, or even become frustrated. A common culprit of this situation is a slowly rendering plot.

During the talk, we will dive deeper into how plots are rendered in Shiny, identify common bottlenecks that can occur during the rendering process, and learn various techniques for improving the speed of plots in R/Shiny dashboards.

These techniques will range from more efficient data processing to library-specific optimisations at the browser level.