MF'in Daniels @lkjsdfammmf - Tumblr Blog

Posts

So I finally launched my data/pop-culture publication, Polygraph with this long-read on the longevity of music, measured using play counts on Spotify.

Check it: http://poly-graph.co/timeless/

This is a story about proving, with data, that No Diggity by Blackstreet is timeless.

Until recently, it was impossible to measure the popularity of older music. Billboard charts and album sales only tell us about a song’s popularity at the time of its release.

But now we have Spotify, a buffet of all of music, new and old. Tracks with fewer plays are fading into obscurity. And those with more plays are remaining in the cultural ether.

July Links and Polygraph Update

/Polygraph Instead of attempting to launch a client-services business or a Paul-Graham'ian startup, I'm playing the long-con and shipping a web journal/magazine for data viz. I'm aggressively close to publishing a few projects by the end of the month, all focused on music and data, which will essentially serve as the first issue. /Links I’ve been thinking a lot about data viz content that could evolve into a (capital P) Product. This example from the NY Times is a perfect example...A 3-D View of a Chart That Predicts The Economic Future: The Yield Curve. I could totally see a financial software company buying this software, or potentially spinning out a team to focus on building features to complement it, eventually packaging the viz as a paid service/product. Quartz launched a new channel for their charts and data viz: Atlas. Sadly, it’s still images (wtf!). Yet it’s not all too different from the several sites trying to build a home for data viz, notably dadaviz and Plotly. Turns out that another awesome home for data viz, Many Eyes by IBM, is shutting down :( There’s a parameterized, explorable visualization of K-12 Department of Education statistics. Too bad it’s made with Tableau and semi-unusable from a UI, UX, and speed perspective. When is someone going to make a proper Tableau for web content? There’s a great podcast on Data Viz called Data Stories. Check out the latest one from Amanda Cox (NY Times Graphics Editor). Via Kottke.org, “From designer Karl Sluis, a list of nine great book about information visualization not written by Edward Tufte.” They’re still print books though – so nothing interactive here... Link. Many media sites have tried to pull this off: fixed graph and scroll-triggered prose/charting storytelling. This may be the best example that I’ve seen, using global warming data. The line chart’s path also cleverly animates in while the x-axis’s position is highlighted. Link. Instead of making a really long horizontal bar-chart, Bloomberg uses a clever mouse-over zoom to create a sort-of accordion effect. The data, civil rights investigations, is also pretty compelling. Link. The WSJ transposes a stock-chart’s x-axis to the window’s scroll-position. Kinda genius: “Flash Crash' a Perfect Storm for Markets.” Link. This isn’t data viz, but Bloomberg’s design department can’t stop won’t stop. Link. This might be the most impressive example of a node-link diagram on the Internet via none other than the NY Times. I suspect that Palantir’s software does something similar to help the CIA catch terrorists. Link.

smallrooms

Data Viz Linkz

I’ve immersed myself in data viz over the past few weeks, and here’s some of my best finds:

A solid extension of Bret Victor’s explorable explanations, Mike Bostock’s long-read, Visualizing Algorithms, is a masterpiece on how data viz and D3 can depict the way algorithms work. Link.

Nikki Case and Vi Hart illustrate how gentrification and segregation work with a series of parameterized simulations. The illustrations are also adorable. Link.

Here are the best-practices for transitioning data from one form to another (e.g., from a bar chart to a scatter plot). Link.

Statmuse is a natural language-based charting for NBA stats – think Siri for random basktetball facts, such as, “Who has the most turnovers in the 3rd quarter of the NBA Finals?” Link.

Trifacta is another startup out of the University of Washington/Stanford Interactive Lab. It attempts to solve one of the main challenges in data science: data munging and cleaning. It’s a little rough around the edges, but overall allows you to make sense of what’s in a large dataset. Link.

Bloomberg uses a novel approach to the histogram to illustrate the rate at which states adopt new social-driven legislation (women’s rights, gay marriage, drug legalization). Link.

This block is the first time that I’ve seen a filter combined with an axis. It doesn’t have a name (perhaps brushing?!), because it’s so weird and new. Link.

Plottable is a D3 framework from Palantir that makes rapid creation of charts much easier. Link.

Stereotropes attempts to visualize all of the data in the massive movie/TV trope database. It’s a product out of the data viz department of Bocoup. Link.

Wow – this launch page is fucking gorgeous.

Holy shit this be amazing.

I’ve been designing more d3 charts with NBA data. My second post on waiver-wire pick-ups (via poly-graph.co) http://poly-graph.co/basketball.html

Some experiments in data viz for fantasy basketball (via poly-graph.co): http://poly-graph.co/basketball.html

The tool-tip on Medium’s charted.co is extremely satisfying. Here’s link to the data viz using a csv I uploaded (via polygraph)

Linkz

I’ve been thinking a lot about data viz recently. Here’s what I’ve been reading:

UW Interactive Data Lab: This department seems to be on the edge of data viz research. Turns out they’re the peeps behind the ubiquitous data-viz platform Tableau, the semi-popular startup Trifacta, and the Bret Victor-esque data viz design tool Lyra.

Lyra - Lyra is about the closest thing to a Photoshop/Illustrator for data viz. The idea is that you can quickly explore data visualizations without getting into code, as outlined by Bret Victor in his presentations on dynamic data visualizations.

How Your Hometown Affects Your Chances of Marriage - Again, the Upshot kills it on data viz, specifically the middle bubble chart.

BlockExplorer - this is a f’ing amazing tool that should probably displace random Stack Overflow searches for D3 code.

I’m also plugging away at a blog for Dataviz here.

Sign-up for email updates here.