Beautiful maps in minutes: Meet Kepler.gl

SAN FRANCISCO — Shan He may hold Silicon Valley’s most meta job.

“When I started out, I was building maps. Then I moved on to build tools to build maps and now I’m doing tools to do tools that build maps.”

He, who dumped brick-and-mortar architecture studies for computational design, joined Uber as founding member of the data visualization team in 2014. She went on to construct Kepler.gl, a tool that helps make “beautiful maps in like 10 seconds” — without any coding. Built using the deck.gl WebGL data visualization framework, the ride-sharing company recently open-sourced the geospatial toolbox that can be used with QGIS, Carto and Mapbox Studio. Given its origins, it’s easy to see why Kepler excels at large-scale visualizations centering on geolocations.

She set the stage for a recent standing-room-only meetup hosted by Mapbox for two case studies on what the tool can do and how you can get involved.

Catching air inside Uber

Chris Gervang, software engineer at Uber, started playing around with Kepler.gl without alerting his data vis team colleagues to his tinkering. Gervang works on one of the Valley’s most (ahem) pie-in-the-sky initiatives: an urban air transportation project dubbed Elevate. The second Uber Elevate summit featured a keynote by the head of aviation, Eric Allison, that would’ve fallen flat if not bolstered with stunning maps.

Titled “Scaling Uber Air,” it featured data visualizations to show everything the company knows about how people move around in cities and how to use that to model the future Uber air network.

“We were extremely careful to use realistic data for our visualizations, to show what we think Uber Air can look like using our most current research,” Gervang says.

To produce the visuals, they used images captured from a cinematography app, on top of Kepler.gl. in a four-step process: planning the story, using Kepler.gl to visualize the data in motion, then a video editor to add effects that weren’t available in a browser and, finally, put them into a slide deck to add graphics. His team then repeated the process for all 17 visualizations in the keynote.

One of those maps showed an average weekday in Los Angeles and Orange counties, 45.8 million unique trips happen over a 24-hour period aggregated, into one kilometer by one kilometer tiles — here’s that bit from the keynote, it’s even more impressive with motion. The entire keynote or slides are also available.

Lime time

If Uber is hated enough to regularly station guards outside its Market Street headquarters, Lime has the distinction of being one of three scooter sharing companies recently booted from the streets of San Francisco.

Aash Anand, head of analytics at Lime, says the company started out in 2017 in a much less contentious space, just a “normal, not electric, bike sharing startup.” One of the visuals showed a Kepler-made map of bike patterns spidering around Greensboro, North Carolina, the first Lime town.

Anand says that speed is what gives Kepler an edge — he put together his presentation for the meetup together on the 40-minute Uber ride from Lime’s San Mateo headquarters.
“That’s actually the biggest value of Kepler has added to my life as someone who works with data,” he says. “When Lime started really growing there was no time to make things look pretty and perfect, some of these were done at lightning speed, at the request of someone in a crisis.”

What’s next

The team at Uber hopes mapmakers that will adopt Kepler — and provide feedback. As they’re planning out the next set of features —including making sharing a lot easier more visualization types, especially “cool things like flight paths” Shan He says— they are asking folks to take a survey and hashtag any projects with #keplergl. If you’re ready to try it out, try the demo or check out the GitHub repository and the tutorials on Vis Academy, a hub for visualization tutorials and classes prepared by the Uber Visualization team.

Stories, Stats & Scatterplots: Inside data visualization at the Financial Times

Four years ago, John Burn-Murdoch was a journalist who had never written a line of code. These days, the senior data visualization journalist at the Financial Times says he rarely writes anything but code.

While what Burn-Murdoch dubs a “cool journey” sounds a little extreme (from zero to R, on the job, seriously?), he was never strictly a text journalist. He holds advanced degrees in data science and interactive journalism. Before landing at the FT, he worked as data journalist at The Guardian for two years.

On loan from the London headquarters to focus on bias in artificial intelligence and geographical inequality, the spiky-haired, elfin Burn-Murdoch offered a peek inside the workings of the FT data newsroom for about 50 members of the Bay Area d3 User Group.

Continue reading

Crooked! Donald Trump’s most recent insults as a word cloud

UPDATE: The Times is still tracking the list of insults — as of January 2017 it grew to 305 — and added a visualization that shows the kinds of people and things most frequently insulted. (Spoiler alert: journalists and Democrats.)

The reporters at the New York Times combed through Republican presidential nominee Donald Trump’s Twitter feed for the most recent 250 insults to nations, people and random things – including a podium.

NYtimesThis is the kind of story that cries out for a visual representation – there has to be a better way to process the information than listing names of the people he insulted in alphabetical order and the tweets as quotes underneath them. What story does that tell?

Most commonly used words in Trump insults, by frequency.

Most commonly used words in Trump insults, by frequency. By Nicole Martinelli, via Wordle.

A quick word cloud will tell you that the most common insult for the straight-talking New Yorker is “crooked” (his go-to insult for rival Hillary Clinton) followed by “dishonest,” “bad,” and “failing.”

A couple of necessary caveats: this cloud was made with a tool called Wordle and the size of the word corresponds to the number of times it appears in the text. The text in the graphic was copied and pasted from the article on the NYT site without any additional weighting or manipulation. The program automatically cuts out common words (i.e. articles) but it would be interesting to see how the cloud shifts by cutting some filler words like “new” “news” “many” “another” etc.

Digital publishing gives public figures so many ways to broadcast a message – it’s our job as journalists to make sense of it. What would you trawl through other political figures tweets to understand?

Mapping the term ‘fiscal cliff’ with the New York Times API

8262373294_99fd141ed2_bAs time goes by, it looks like we’re all going to fall off the fiscal cliff.

When thinking about ideas to test out the New York Times API, which lets you dig into everything from campaign finance to geographic tags in the old gray lady’s formidable database, I wanted to keep it simple. (The learning curve was already steep for me, to be honest.)

The chart is my first pass at using the NYT API, based in part on Jer Thorp’s chapter about it in “Beautiful Visualization,” which you can find at Amazon  or a public library.

The result is a simple cliff graphic that mimics the momentum the term gained along with the budget crisis. Wikipedia dates the term “fiscal cliff” to 2011, but Federal reserve chairman Ben Bernanke gave it heft in February 2012 with testimony before the House Committee on Financial Services, the first mention is here: nyti.ms/Tcd7Lq