The fifth annual Government & Public Sector R Conference was back in Washington D.C. (& virtually online) on December 1st & 2nd, bringing together data scientists, analysts, statisticians, engineers, professors, and government officials for a fantastic two days of connecting, sharing, and learning! This event was not just about the presentations, but also about the conversations and collaborations that took place as attendees and speakers shared their work with others from a wide range of domains. It was truly an inspiring event that highlighted the power of R (and Python) and how it continues to benefit those contributing in the public sector.
Watch every talk from the 2022 Government & Public Sector R Conference here → Lander Analytics YouTube channel
After a two-year hiatus from in-person events, Georgetown University welcomed us back with open arms to host the “R Gov” Conference once again. We are so grateful for the Data Science & Analytics Program at Georgetown and their continued support in our mission to unite R users.
Thank you speakers!
This year, 25 speakers shared their work with conference attendees across the three-day event (one day of workshops; two days of conference talks). If you missed it, you can now watch every talk on the Lander Analytics YouTube Channel! Here are some highlights from a few of our amazing speakers at this year’s event…
Pictured above from left to right (Top): Tyler Morgan-Wall, Shirley Han, Dusty Turner, David Shor, Drake Gibson, Merav Yuravlivker, Alex Gold, David Meza (Middle): Molly Huie, Max Richman & Danya Murali, Tommy Jones, Asmae Toumi, Refael Lav, Danielle Larese & Selen Stromgren, William E. J. Doane (Bottom): Marck Vaisman, Thomas Bacas, Jake Dyal, Maxine Drake, Benjy Braun, Jared Lander, Kim Ky, Nick Childs
Check out the best photos → 2022 R Gov Photo Gallery
David Shor talks data science and US politics
David Shor from Rose Research discussed the role of analytics in assisting democratic campaigns and progressive organizations with election forecasting. One of the key areas of focus for Rose Research is to drive insights and recommendations that can help political campaigns allocate resources more effectively. Shor noted that analytics has come a long way in a short period of time when it comes to the acceptance by campaign staffers and political organizations alike.
One key achievement for Rose Research has been the identification of campaign ads that had adverse effects on voter sentiment. They created a model to identify ads that could make democrats vote for the republican party. The campaign would then no longer run those ads.
Shor emphasized the importance of careful data analysis in election forecasting, noting that relying on linear thinking and failing to account for correlated error can often lead to incorrect (and costly) predictions. As a result, the team at Rose Research often employs a Bayesian approach in their analysis.
Major Maxine Drake from the US Army showed us her process to build R Shiny applications
Major Drake began her presentation with a story about her 13-year journey from engineer to operational research to data scientist at the US Army. During this time, she has witnessed the evolution of technology and the increasing acceptance of data-driven tools like R to improve army operations.
Maxine's team is responsible for providing full-stack data services to support strategic decision-making in the army, including tasks such as analysis, engineering, troubleshooting, and general support. They often build Shiny apps to deliver data and visualizations through dashboards, as user-friendly frontend applications that hide the backend have proven to be an effective way to gain buy-in from users.
One interesting point to note is that US Army analysts tend to use R for data analysis in part due to RStudio's infrastructure working well within the army's strict security restrictions.
Molly Huie returns to the R conference with news from the LAB team at Bloomberg Law
There has been rapid evolution in the area of analytical buy-in at Bloomberg Law. Since Molly gave a presentation at the New York R Conference in July, her analytics team has grown from four to twelve people. Molly, who has transitioned from being a coder to a team lead, has made it a priority to gain buy-in for analytics from key stakeholders.
To achieve this, Molly's team has found that using R Shiny apps to deliver data-driven tools and analysis has been effective. Even better, the team has recognized the value of providing training in data literacy to help legal professionals skilled in law understand how to use data and technology in their field.
Tyler Morgan-Wall builds a 3D rendering of Washington D.C. in minutes (a must-watch)
Tyler gave another highly popular presentation at the R Gov conference, demonstrating the impressive capabilities of his 3D rendering packages built in R called rayshader, rayrender, and rayverse. He amazed the audience by showing how to build an entire 3D city in just a few lines of code using R.
Tyler's interest in 3D modeling began when he played SimCity as a child. Now, as the developer of open-source 3D packages in R, he demonstrated to the audience how to easily do this on their own machines without needing any prior 3D modeling skills.
I would highly recommend checking out Tyler's presentation below - you'll see him digitally recreate an entire 3D rendering of Washington D.C. in just a few minutes.
And finally, I gave a presentation on how to deploy R in a secure environment
I want to thank Dusty Turner for his incredible words when introducing me on stage. It is amazing to think that I have met so many wonderful people in my life through the R community.
This year, I gave a talk on deploying R in a secure environment. There are several decisions to make when setting up a secure environment for programming languages like R, Python and Julia. A few of the key takeaways:
Code editors such as RStudio, Jupyter, VSCode, or Posit Workbench need to be hosted, managed, and secured when shared across users.
There are also several options for hosting your environment, including using an on-premises virtual machine or a cloud-based virtual machine, such as those offered by AWS, Azure, GCP, or Digital Ocean.
Alternatively, you may choose to install your environment into a Kubernetes cluster, which allows users to automate deployment, scalement and management of containerized applications (via docker)
For the case study in my talk, I focused on Posit Workbench and the software stack that can effectively run the infrastructure of a secure R environment. Follow along to see how easy it can be to install Workbench, Connect, Package Manager, Postgres, and NGINX in a docker container to enable R in the cloud.
If this is something you are interested in learning more about, don’t hesitate to reach out!
Interactive Workshop with William E. J. Doane: An Introduction to Natural Language Processing for Public Policy Research
Along with the conference, we hosted a hybrid workshop with William E. J. Doane on November 30th. This full-day interactive Introduction to Natural Language Processing for Public Policy Research workshop gave in-person and virtual attendees the opportunity to use R to acquire data from various sources, clean and standardize the data, and explore it for insights that can inform public policy discussions. It was a truly unique and engaging experience, and we’re thankful for William for leading such an inspiring workshop!
Continuing the hybrid experience
After the success of our first hybrid R Conference in June (the New York R Conference), we wanted to keep the momentum going and offer virtual tickets alongside in-person tickets, so that people from all over the world could attend. We had lots of giveaways for the virtual attendees, from physical books to ebooks. There was an interactive virtual platform for those viewing from home, and for our D.C. attendees, there were plenty of snacks and delicious food.
Thank you to our sponsors!
We are immensely grateful to our conference sponsors for their generous support. Without them, this conference would not have been the same. Thank you Georgetown University, Deloitte, R Consortium, Posit, Saturn Cloud, PolicyViz, Pearson, Chapman & Hall/CRC, Springer, and Manning.
I am grateful for the Lander Analytics Team!
My team gave 110% to ensure that this event ran like a well-oiled machine -- or perhaps more appropriately -- like an optimized algorithm. I am forever grateful to them for their hard work and dedication. A special thank you goes out to Nicole DelGiudice, who masterfully organized and executed this event with tremendous skill and expertise.
We invite you to join us in 2023!
The past few years have presented unprecedented challenges. We are incredibly thankful to the community for their unwavering support of the R conference series in spite of these difficulties, and for those who traveled in-person to make this year so special.
We will continue to offer hybrid opportunities for people around the world to tune into our conference and benefit from the knowledge shared. Our New York R Conference will make its return to the city in the summer of 2023. Keep an eye on our @rstatsai conference Twitter account for updates and details!
We look forward to connecting with you soon!
Jared P. Lander
Lander Analytics Chief Data Scientist