📣 If only your chart could talk

👋 Hi!

This week, I'm writing the third module of my Matplotlib Journey project, and it's about a crucial yet often neglected part of data visualization:

​

🌈 Annotation ☀️

​

I like to think of annotations as what your chart would say if it could talk. Essentially, explaining to the reader what they need to understand.

There are 2 main use-cases for annotations:

​

1️⃣ Annotation to Provide Context

Let’s say you need to build a bubble plot showing the relationship between GDP per capita and life expectancy.

With a few lines of R code, you get something like this:

Not too bad! 😎
​
The message is clear, and the trend is obvious.

But how frustrating is it not to know which country each dot represents? Some circles are clearly out of the trend. It drives me crazy not to know who they are!

That’s where annotations become essential. A bit of text here and there can turn a good chart into a great one:

Depending on your story, it's up to you to decide which countries to highlight to guide the reader and help them understand the message.

Another example I like:
​

This graph is a connected scatterplot. It illustrates the evolution of both troop numbers and the army budget.

Without annotations, the message would be unclear, but with them, the chart tells a complete and compelling story!

---

2️⃣ Annotation to Highlight a Result

Annotations can also be used to draw attention to the specific part of the chart that tells the main story.

They don't add extra information, but they make the key result stand out, ensuring your message comes across clearly.

Here’s an example from the Python Graph Gallery:

The full chart is quite dense and could easily overwhelm the reader.

However, the annotations ensure the key points stand out, keeping the reader focused and engaged.

---

Conclusion

The takeaway from this post is simple yet fundamental: next time you present a graph in a report, paper, or presentation, add some annotations!

Otherwise, chances are you could make the graph more effective.

For a few years now, I’ve been collecting all the data visualization projects I love on Dataviz-inspiration.com. Check it out! You’ll see that about 90% of the projects include lots of annotations.

See you next week for another dataviz tip!

​

Yan

PS: If you add an annotation to your chart this week, send me a screenshot! It’ll make my day! You can find me on LinkedIn and Bluesky.

PPS: There are now 283 students enrolled in Matplotlib Journey 😱. And we’re only halfway through writing it! Thank you so much for your trust and support! 🙏
​