Writing in R
There is a lot out there written about R, R Markdown and all its family of tools. Most of what I have learned about it I got from Out There, by googling what I needed. I expect anyone reading this can do that too. These posts will help me organize and credit all those amazing sources that have made my own progress and learning possible and enjoyable. Ideally they will also reach other people that can benefit from my findings and that –even better!– may have ideas to share themselves.
I have always liked to write, and for many years did most of it in MS Word and really enjoyed it. I love writing, I love formatting, I love seeing words in a beautiful environment and feeling all cozy inside, maybe just because I like the font. I do not underestimate the effect that graphics, colors, layout, sizes, margins, etc., have in my emotions, and therefore in my mood, my enthusiasm, my performance. Language is not only about the what, but very much about the how.
About 5 years ago I was introduced to \(\LaTeX\). I was not sure what I could get from it, because most of its advantages seemed to be related to STEM-related writing, which (back then) I was not doing. However, the neat output that required so little tweaking, after a horrible-looking input that focused more on content, captivated me. I was looking forward to move from MS Word to \(\LaTeX\), and fearful of assignments that would require an MS Word output.
I started my PhD position in Leuven shortly after, and my \(\LaTeX\) experience was limited to a professional-looking CV and a chapter for what I thought was going to be my PhD thesis. I didn’t really write for a while, in this new position –mostly notes, for which I like to use Evernote. However, I did ask a colleague about what he used to write papers, given the requirements of different journals… and he said “R Markdown.”
“You can just check the cheatsheet online,” he said. That was all. I took note of it for the future, but that was it for a while. I didn’t have chance to use it immediately (not that I could see, at least). I was already using R and R Studio, getting more and more familiar, but very slowly, and I certainly didn’t see it as a word processor.
After a year or so, my R code was looking more complex and interesting, a whole story in a programming language. The comments I needed to understand what was going on, to remember the links between the pieces of code, were growing into full paragraphs. I wanted to document what I was doing properly, ideally close to the code itself. And R Markdown (Xie, Allaire, and Grolemund 2018; Xie, Dervieux, and Riederer 2020) is perfect for that.
Suddenly, I could create texts that interlaced the documentation, my thinking process, and code –maybe the code itself was not visible in the final output, but the source document remains available and is quite readable itself. And that final output…
First, you can have the same pretty PDFs you get with \(\LaTeX\), and include \(\LaTeX\) expressions (like I’m doing here with the logo). Second, you can also have MS Word output, in case you need to write a paper with it as requirement (I only did that much later). Third –and foremost, in my opinion–, you can have HTML output. Beautiful, endless scrolling HTML, with collapsable table of contents, responsive features, tabs, and so much more! This seemed even more enticing given that I had been exploring HTML and CSS for a while (CSS is what you use to style HTML), so I could personalize and enhance my HTML’s even more.
I started writing reports, merging documentation, code, visualization, and ideas.
I explored new templates, like those from
rmdformats (Barnier 2021).
I wrote so much code for them that then I moved them to specialized scripts.
I wrote so many scripts that I created a private package for most of them.
I started writing my dissertation in
bookdown (Xie 2016),
and gave a talk about it with
xaringan (Xie 2021).
I created this site with
blogdown (Xie, Hill, and Thomas 2017).
I grew so much in R, and I would not have anticipated it when I first started looking into corpus linguistics in 2013 and a professor suggested I “use R,” as if a single letter was enough to find anything about it. It certainly is –the community is solid and active, and I look forward to giving at least half what I’ve received from it ❤️.