🛠️ Dataviz tools: the big picture

👋 Hi!
​
Before we start, a warm welcome to the 253 people who joined this newsletter last week 🙂👋 ❤️. (You can join here btw)

Today’s menu is (almost) short and to the point, just the way I like it.

I want to help answer a very common question:

​
​Yan, what tool should I use?

You have a chart to make, an app to build—and you’re lost in a galaxy of tools.
​
You’re not alone!
​
Here’s the mental model I use to navigate it all:
​

1️⃣ Excel

You know it. Some people create impressive things in Excel, and I use it for personal paperwork—but I wouldn’t recommend it for serious data analytics.

​

2️⃣ Data Visualization Tools

These aren’t my world since I prefer programming. But if you’re in a rush or don’t code, they’re lifesavers.

I’ve seen researchers waste weeks making graphs in R. Tools like Datawrapper can get you clean charts in seconds! Just load data, set a few options, and you’re done.

3️⃣ Data Science Programming languages

The two main programming languages are R and Python.
​
Both are awesome, and there’s no reason to use proprietary tools like SAS or Matlab 🙈. If your organization does, consider moving fast!

If you doubt R and Python can produce stunning visuals, check out my lists of top charts made with Matplotlib (python) or ggplot2 (R).
​

There’s one big problem, though: the graphs produced are all static—no interactivity.

​

​

4️⃣ JavaScript Dataviz Libraries

Interactive graphs are all made thanks to the programming language of the web: javascript. Javascript allows to do stuff like: "if user clicks on this circle, then make it red".

There are a myriad of libs like Chart.js that work similarly to R or Python functions but output interactive graphs:
​
​scatterplot(data_for_x, data_for_y, color, size).
​
➡️ And voilà—an interactive scatterplot!

​

5️⃣ D3.js

D3.js isn’t a charting library;
​
It does not provide a scatterplot function—just utilities that let you create elements from scratch.
​
For example, the treemap() function just computes positions for rectangles, but you draw them yourself.
​
As a result, it is very flexible, but pretty hard to use because it requires a lot of code.
​
​

Bridges

➡️ D3.js is the foundation of web graphics, powering most JavaScript dataviz libraries. So JS libs use d3 under the hood.

➡️ It is possible to make interactive graphs with R or python, thanks to tools like the HTML widgets. Those are actually wrappers of javascript libraries! You use JS without writing a line of it!
​

The Grail

D3.js is often considered the grail of dataviz: it’s limitless. You can draw anything you can imagine with it. Two months ago I built this with it and it made me win a dataviz competition.

But the learning curve? A bit brutal...

Wanna see examples? Check my graph gallery!

Visit the D3 graph gallery

​

​

Sooooooooo,

I'm very curious! Do you know what d3.js is? Have you ever used it? If not, why? And more importantly, would you like to learn it?
​
I wish you a wonderful day and would love to hear from you!
​
Yan
​
​PS: next week, I'll try to demystify a bit what d3.js is!
​
​PPS: you've been almost 100 people taking my productive-r-workflow project last week, thanks!! 🙏

​
​