Hey folks,

Happy 2025! I really hope that one of your resolutions for the coming year is to grow in confidence and skill using data analysis tools like R. Regardless, I’m honored that you continue to check out the content I describe here in the newsletter and on my YouTube channel. Thank you.

Looking back on my 2024, I did a lot of thinking about how we learn, where I find inspirations, and how I teach. I have been fascinated by other YouTube channels that break down why songs are great, the techniques of different directors, recreating scenes from movies, and describe how to recreate drawings and paintings. So why not look at why data visuals are great and how we can recreate them with our tools?

This is what I’ve been doing since August. I’ve really loved describing my thought process from a 30,000-foot view and then implementing that view on YouTube. Of course, the goal isn’t to recreate other people’s visuals and end there. It’s a means to an end. The goal really is to learn techniques that we can apply to our own questions, to critique visuals from other scientists, and to develop our own methods.

Sometimes I worry that I spend too much time getting just the right font size, line thickness, symbol size, color, or margin width. But by modifying those attributes so fastidiously we are learning how those attributes work. I’m not sure that I need 100 shades of gray or that I can see the difference between a line thickness of 0.10 and 0.15. But, looking at the nuances of these values has really been educational for me and a fun exploration. I have been forced to try things that I wouldn’t ordinarily try.

As things were winding down last semester, I was listening to a Spotify playlist when a song from a family-favorite band, OK GO, came on. Seriously, if you and you kids haven’t “wasted” a night watching OK GO videos on YouTube, you’re really missing out. Anyway, this was the album art for the song that came up

I thought to myself, “Wait, OK GO knows how to program in R? OK GO is interested in genomics?” Probably not. My point in sharing this humorous episode is that by looking for visuals to share with you all I have started looking for visualizations everywhere - even when it isn’t a data visualization!

I’ll hold of on a 30,000-foot description of how I’d make this and instead opt for a 100,000-foot description (is that a thing?). I suspect that I’d start by creating a stacked bar plot and removing the space between the bars on the x-axis - perhaps we’d use geom_area(). Then I’d then use coord_radial() to circularize the area plot. To get the cool color scheme, I’d have my area plot be made up of 20 or so different categorical variables. I’d use scale_fill_manual() to then map each of those variables to a different color. Totally doable :) Go ahead.

Of course, this would take a lot of fiddling to get the right color scheme and pattern. You might be surprised to learn that there are R programmers who use R to generate art. One package that I found is {aRtsy}.

As much fun as I think it would be to generate my own version of this type of art, I’ll likely pass. Part of what I actually love about the album cover is how organic and chaotic and beautiful the patterns are. That’s pretty much a theme for OK GO’s videos in general. For this album cover, I suspect someone used a turntable and dropped different colored paints. Perhaps they had different viscosities and perhaps the artist etched in channels to create the radial streaks we see. Well, I just googled it and sure enough, that’s pretty much / what they did.

My hope for all of us in 2025 is that we tune our eyes to find more data visualizations in the world. Then, the fun begins with thinking through how we’d create the visual followed by actually making the visual.

​

​