Level Up Your Labels

Five tips from my useR 2022 poster presentation

Author

Affiliation

Cara Thompson

Building Stories with Data

It was a privilege to present “Level Up Your Labels” at the #rstats _useRconf poster session this afternoon. Thanks for the good questions!

Here’s a link to the poster with the full code I used to create the plot. www.cararthompson.com/talks/user2022/

And here are the 5 tips in detail🧵👇

Tip #1: Make it easy for your readers, by using colour to help them orient themselves as they read the text introducing the plot.

• 📦 {ggtext} + {glue}
• 🪄 <span style=colour:palette$relevant_colour>blah</span>
• 🔧 theme(plot.title = element_markdown())

Tip #2: Keep the main thing the main thing, with a primary and secondary text colour

• 📦 {monochromeR} to get the colour variants + {ggtext}
• 🪄 same <span> trick as above, overriding the default colour of the text in the box which will follow scale_colour_manual()
• 🔧 geom_textbox()

Tip #3: Use alignments to direct the reader’s gaze

Once you’ve decided where your boxes are going to sit, set the alignments programmatically with:

• 🔧 case_when() in the data, and aes(hjust, halign, vjust, valign) in the plot code.

Tip #4: Add arrows sparingly

Narrow down the points you want to highlight, and add arrows going from the box coordinates to the points.

• 🔧 case_when() again to set end points programmatically
• 🪄 loop through each unique curvature to add arrows with different curves

Tip #5: Give everything space to breathe

Get rid of unnecessary legends, and play around with lineheights and margins.

• 🔧 guides(colour = "none)
• 🪄 scales_y_continuous(expansion = expand()) for an expansion proportionate to the range in the data

{fig-align=“center” fig-alt=“Tip on design:”Give everything space to breathe” with advice on using line height, margins, and removing unnecessary lege…“}

And voilà!

All done in #rstats with {ggplot} and the tips and tricks above.

Full code: www.cararthompson.com/talks/user2022/

Happy annotating!