September 12 - October 19
Mondays & Wednesdays
6:30pm - 9:30pm
Check out some of Kevin's favorite pieces he and his colleagues have created:
You can view more of Kevin's data visualization articles for The New York Times here.
D3 can be challenging to learn, and lots of the difficulty comes with learning syntax and understanding data joins, probably the most fundamental aspect of D3. On top of that, there are the headaches associated with what Hadley Wickham has called the "fiddly bits" of charts: scales, axes, labels, annotations and the like. These basically come for free with software like Excel, R, Chartbuilder or Data Wrapper, but with D3 you're in charge of every pixel, which makes it incredibly powerful.
Ultimately, the reason to learn D3 instead of or in addition to those other (great!) programs is because it enables you to tell stories and communicate information interactively in ways that are simply not possible outside a web browser. Making these kinds of applications is worth a little bit of extra headaches up front.
This is a course for anyone who wants to be proficient in the use of D3 and seeks expertise visualizing quantitative information. This course will not make you a data scientist. (For that, Metis offers a licensed, accredited bootcamp).
About a third of our course will be discussions and theory. We'll read from classic data visualization texts, discuss the merits and flaws of work published at some of the world's best institutions and get a strong understanding of effective visual communication.
If you don't yet have this familiarity, we suggest completing some of the following free tutorials in advance of the course start:
The 36-hour course is held on Monday and Wednesday evenings from 6:30pm - 9:30pm at Metis in both San Francisco and New York City where the Metis Data Science Bootcamps and other evening courses are also held.
San Francisco: 633 Folsom Street, 6th Floor.
New York City: 27 East 28th Street, 3rd Floor.
Learning to make charts form by form – scatter plots, then bar charts, then line charts, and so on – is not the most creative way to learn about data visualization. But because the course is so technical, this structure will help provide a foundation we can build on each week.
Each class will roughly be split into two. Half will be discussions and a hands-on, computers off activity about that week's subject, and half will be a deeply technical guided lab making things in D3.
We'll do boring things like configure our computers, make our first charts, understand why data joins are helpful and get a sense of all the things we need to learn.
If you want to be good, you really only need to be good at making a few kinds of forms: bar charts, line charts, scatterplots and maybe a histogram. We'll make demos of all of these and understand when to use which.
Since your computer is drawing the charts instead of you, making 100 charts is as easy as making 1. We'll explore the power of exploratory sketching and the data manipulation you'll need to be able to master to do so.
Mapping with D3 has exploded in the last few years. We might not explore Great Circle Arc Intersects, but we'll learn how to make bubble maps, choropleth maps and create topojson files from scratch.
Most things don't need to be interactive, but when they do, you'll be ready. We'll use D3 to make dynamic charts and applications that let us answer questions and solve problems that couldn't have happened in a printed format.
Here, we focus on honing ideas and making publication-grade data visualizations. We'll work on small touches, like custom annotations and styles, managing your data visualizations on mobile devices, incorporating feedback and pitching work for publication. We'll also do a "show and tell" of projects we've been working on throughout the 6 weeks.