Jake Kaupp

2020-02-10

 

Who are you, and what do you do?

Hi! My name is Jake Kaupp (@jakekaupp on Twitter, jkaupp on GitHub). I’m a husband and father of two busy kids. I work as the Assistant Director of Institutional Research (IR) at Queen’s University in Kingston, Ontario, Canada. That entails a lot of work with a wide array of data related to the University and its operations. It’s an analytic field with a lot of opportunities for the application of data science techniques and a lot of visualization work. There are lots of reports to be done.

I manage a team of 4, and while I still do pitch in with data work, most of my time focuses on supporting my team to develop more efficient and effective ways to do our job, deliver essential information to those that need it, and do so quickly and effectively.

While I work in IR now, this wasn’t always the case. My career path looks like a random walk visualized. I’m a Mechanical engineer with a doctorate in Tissue Engineering and did a Post-doc in Engineering Education. I left the meat grinder of an academic job market and worked in various aspects of Higher Education before eventually landing in my current role.

How do you use R?

During my Post-doc work in Engineering Education, I was also working in helping engineering programs develop sound approaches to accreditation. My boss at the time was in charge of the accreditation efforts here, and he asked me to build reporting to support teaching and learning as well as program improvement efforts, as I had a knack for graphic design and visualization. I was hitting the limits of doing things in GUI based systems, and I needed something scripted. When my boss suggested R. I had used it a few times during grad school and thought, “Why not?”.

I started and soon was using R in my day-to-day work, this was around 2011. Around this time, I found that Hadley Wickham made a lot of useful packages, RStudio was a great IDE, and Yihui Xie was about to release knitr.

Here I was learning and trying to tackle problems while the tidyverse tools were being invented and evolving around me. It was challenging, but provided me with a good understanding that doing something in R wasn’t a matter of “can you do this in R?”, it was a matter of “how do you do this in R?”.

Fast-forward to a few years later and I had built a set of packages to manage and report on learning outcomes data for accreditation; built several apps to support student services; provided visualizations and infographics for research and consulting projects and developed several reporting workflows with rmarkdown and GNU tools.

Currently, I use R in my day-to-day IR work and for the projects in my portfolio, and it’s the workhorse and saving grace in reporting tasks. It helps me automate things where I can and showcase to my team the values of the language. R and ggplot2 are my go-to tools for visualization.

What are your favorite packages?

Favourite packages are hard.

There are so many amazing ones. I have a soft spot for rmarkdown, shiny, dplyr, knitr, htmlwidgets and openxlsx as they were huge in showing others the value and power of R. My favourites for visualization are: ggplot2, paletter, ggforce, ggraph, patchwork and ggiraph (with a special mention to grid and gridExtra)

My workhorse packages are magrittr, glue, janitor, here and usethis, which are staples for everyday work.

What hardware do you run R on? Do you use any other software tools?

I run on a maxed-out 13” inch MacBook Pro as it offers many of my favourite graphics design tools in a UNIX-ish system that I don’t have to manage. In the past, I ran everything off of a server with Rstudio/Shiny Server with a simple SQLServer so I could have a central analytical hub. Recently, I’m starting to do more stuff on an Azure platform using R with DataBricks, as cloud services become more prevalent where I work.

I use Tableau (gasp) at work for interactive reporting, creating dashboards and developing data products for the use across the University. The majority of custom, one-off visualization work/reporting gets done in R.

I use a variety of other tools, Affinity Designer and Publisher, for graphic design work and publishing. Alfred for personal productivity. DataGrip for a database IDE, Visual Studio Code for other languages (SAS & Web stuff).

Do you have any projects that you’re working on right now?

Mainly just [TidyTuesday]() to keep the polish on my design, visualization and communication skills. I might look to expanding nord (shameless plug for my one package on CRAN) to include some built-in ggplot2 themes, and I always like to keep expanding my own personal utilities package to help in making graphics.

Aside from that, I am active in teaching locally with The Carpentries and help out mentoring in the R4DS Community. I’ve been taking it easy on consulting and commissioned works as I try to keep my off hours for family time while my kids are young.

My sincerest thank you goes out to the developers, package authors and maintainers as well as the R community. These two things make R what it is, and without them, I wouldn’t be where I am.

Thanks for the interview!