React for Data Visualization
Student Login
  • Introduction

Explore Uber's WebGL dataviz libraries

Uber has built a cool suite of data visualization tools for WebGL. Let's explore

Dataset: Download dataset šŸ—³

My solution šŸ‘‡

How it works āš™ļø

Today was not a great success so tomorrow's gonna be part two.

We explored Uber's suite of WebGL-based data visualization libraries from vis.gl. There's a couple:

  1. deck.gl is for layering data visualization layers on top of maps
  2. luma.gl is the base library that everything else uses
  3. react-map-gl is a React-based base layer for maps, you then use deck.gl to add layers
  4. react-vis is Uber's take on the "react abstraction for charts" class of libraries. Renders to SVG

Trying out luma.gl seemed like the best way to get started. It powers everything else and if we're gonna build custom stuff ... well.

Implementing this example of wandering triangles seemed like a good idea.

Copy pasting a bunch of code and trying to understand what it does even better.

So we built a React harness, spelunked through the demo code, and learned a few things about WebGL.

<LumaFragment component="canvas" />

<LumaFragment> is a d3blackbox component. It creates a canvas element and lets our render function take over. Our function is mostly a copy paste job.

In this we learned that WebGL renders onto <canvas>. Somehow I never realized that before.

We create LumaFragment like this:

const LumaFragment = d3blackbox((anchor, props) => {
// copy pasta code from official example
animationLoop.start({ canvas: anchor.current });
}

The official example creates an animationLoop based on the AnimationLoop import from luma.gl. It works like a game loop approach I believe.

There's 3 event callbacks the code hooks into:

  1. onInitialize, which initializes objects and sets up the visualization
  2. onRender, which runs on every loop of the animation. I think
  3. onFinalize, which runs when the animation runs out, but it looks like that's never

I'd explain the code inside those callbacks here, if I understood it well enough. Right now it still looks a little alien and hard to grok.

What I can tell you, however, it's that it is designed to run fast. There's a lot of low level stuff creeping in. Like using Float32Array instead of just Array because it's typed.

Yes, some JavaScript is typed didn't you know?

And it uses a lot of Buffers. Those are low level memory blocks with direct access from the GPU. Makes it faster than working with higher level JavaScript abstractions.

Another trick for more speed is that much of the hard logic happens in shaders.

We put our shaders in the shaders.js file. Shaders are low-level GPU code, originally named after shading effects, but doing all sorts of stuff these days.

Our example uses 3 shaders: EMIT_VS, DRAW_VS, DRAW_FS. They seem to control how our triangles render and behave, but I haven't been able to figure out how the JavaScript part ties together with the shaders part.

Guess that's our next step.

Also figuring out why this happens:

Join me tomorrow. We'll have fun :)

About the Author

Hi, Iā€™m Swizec Teller. I help coders become software engineers.

Story time šŸ‘‡

React+D3 started as a bet in April 2015. A friend wanted to learn React and challenged me to publish a book. A month later React+D3 launched with 79 pages of hard earned knowledge.

In April 2016 it became React+D3 ES6. 117 pages and growing beyond a single big project it was a huge success. I kept going, started live streaming, and publishing videos on YouTube.

In 2017, after 10 months of work, React + D3v4 became the best book I'd ever written. At 249 pages, many examples, and code to play with it was designed like a step-by-step course. But I felt something was missing.

So in late 2018 I rebuilt the entire thing as React for Data Visualization ā€” a proper video course. Designed for busy people with real lives like you. Over 8 hours of video material, split into chunks no longer than 5 minutes, a bunch of new chapters, and techniques I discovered along the way.

React for Data Visualization is the best way to learn how to build scalable dataviz components your whole team can understand.

Some of my work has been featured in šŸ‘‡

Created bySwizecwith ā¤ļø