In light of the Donald Sterling brouhaha, Amanda Cox for The Upshot put up some charts for why you shouldn't be surprised that people still say racist things, based on data from the General Social Survey dating back to 1972. Mmhm.
You probably remember how Target used purchase histories to predict pregnancies among their customer base (although, don't forget the false positives). Janet Vertesi, an assistant professor of sociology at Princeton University, made sure that sort of data didn't exist during her nine months.
First, Vertesi made sure there were absolutely no mentions of her pregnancy on social media, which is one of the biggest ways marketers collect information. She called and emailed family directly to tell them the good news, while also asking them not to put anything on Facebook. She even unfriended her uncle after he sent a congratulatory Facebook message.
She also made sure to only use cash when buying anything related to her pregnancy, so no information could be shared through her credit cards or store-loyalty cards. For items she did want to buy online, Vertesi created an Amazon account linked to an email address on a personal server, had all packages delivered to a local locker and made sure only to use Amazon gift cards she bought with cash.
The best part was that her modified activity—like purchasing $500 worth of Amazon gift cards in cash from the local Rite Aid—set off other (in real life) triggers.
George & Jonathan used an interactive audio visualization for their recent album George & Jonathan III. This is a fun one. You can rotate the camera as you like, as the full album plays and notes are represented with dashes and dots.
Michal Migurski thinks about finding the right job for the tool rather than the other way around:
Near the second half of most nerd debates, your likelihood of hearing the phrase "pick the right tool for the job" approaches 100% (cf. frameworks, rails, more rails, node, drupal, jquery, rails again). "Right tool for the job" is a conversation killer, because no shit. You shouldn't be using the wrong tool. And yet, working in code is working in language (naming things is the second hard problem) so it's equally in-bounds to debate the choice of job for the tool. "Right tool" assumes that the Job is a constant and the Tool is a variable, but this is an arbitrary choice and notably contradicted by our own research into the motivations of idealistic geeks.
Along the same lines, Frank Chimero on not trying any new tools for the year and how each represents someone's perspective:
Everything that's made has a bias, but simple implements—a hammer, a lever, a text editor—assume little and ask less. The tool doesn't force the hand. But digital tools for information work are spookier. The tools can force the mind, since they have an ideological perspective baked into them. To best use the tool, you must think like the people who made it. This situation, at its best, is called learning. But more often than not, with my tools, it feels like the tail wagging the dog.
These approaches apply well to analysis and visualization. In the early goings especially, there tends to be an obsession with what tools to use. Which is best? Which is fastest? Which can handle the most data? Which makes everything beautiful? And yeah, it's good to give these some thought in the beginning, but don't get stuck asking so many questions or pondering so many scenarios that you never settle down and do actual work.
There's always going to be a new application that promises to help you do something with your data. Work on this stuff long enough and you'll find that you probably won't need that new thing.
Learning regular expressions tends to involve a lot of trial and error and can be confusing for newcomers. RegExr is an online tool that lets you learn more interactively. Add a body of text in one area and type various regular expressions in another. Matches are highlighted and errors are noted on the fly, which is kind of perfect. Even if you aren't new to regular expressions, this is worth bookmarking for later.
Researchers at the University of Washington's Center for American Politics and Public Policy built the Legislative Explorer to show the lawmaking process in action. The visualization allows you to watch over 250,000 bills and resolutions introduced from 1973 to present.
The left half represents the U.S. Senate, with senators sorted by party (blue=Democrat) and a proxy for ideology (top=liberal). The House is displayed on the right. Moving in from the borders, the standing committees of the Senate and House are represented, followed by the Senate and House floors. A bill approved by both chambers then moves upward to the President’s desk and into law, while an adopted resolutions (that does not require the president's signature) moves downward.
Each dot represents a bill, so you can see them move through the levels. Use the drop-down menus at the top to focus on a Congress, a person, party, topic, and several other categorizations. Or use the search to focus on specific bills. Finally, when you do press play, be sure to keep an eye on the counters on the bottom.
The Pyro Board is a matrix of 2,500 flames that have controllable intensity, which can be used as an audio visualizer. Yeah, really. Just watch the video below.
For the past couple of sports seasons, Facebook mapped the most liked team by county. They did it for football (NFL), the NCAA basketball tournament, and baseball (MLB). Although generalized, the maps provide a view of sports fandom and people clusters across the country, and plus you know, they're fun.
The Upshot used the same like data, provided by Facebook, and mapped it at the ZIP code level. Then they took it a step further and looked closer at regional rivalries, such as Cubs and White Sox, Yankees and Red Sox, and Dodgers and Angels. Be sure to scroll down to Mets versus Phillies. They incorporated a tidbit of Josh Katz's dialect map.
The Upshot is off to an impressive start. It's almost as if The New York Times people have been doing this for a while. [via @KevinQ]
The New York Times released PourOver, a library that lets you do database-like things client-side, so that (1) you, the developer, can worry less about database optimization and server loads and (2) users get a more responsive, faster experience.
PourOver is built around the ideal of simple queries that can be arbitrarily composed with each other, without having to recalculate their results. You can union, intersect, and difference queries. PourOver will remember how your queries were constructed and can smartly update them when items are added or modified. You also get useful features like collections that buffer their information periodically, views that page and cache, fast sorting, and much, much more.
Also: How great is it that The New York Times is now getting into the habit of releasing code?
Remember those running maps I made with limited data from RunKeeper? Strava, which also provides an app to track your runs and bike rides, has a much more expansive version of popular paths around the world. Their dataset includes over 77 million rides and 19 million runs, summing to about 220 billion data points. Just pan and zoom to your area of interest, and there you go.
One of our highest priorities will be unearthing data sets — and analyzing existing ones — in ways that illuminate and explain the news. Our first day of material, both political and economic, should give you a sense of what we hope to do with data. As with our written articles, we aspire to present our data in the clearest, most engaging way possible. A graphic can often accomplish that goal better than prose. Luckily, we work alongside The Times's graphics department, some of the most talented data-visualization specialists in the country. It's no accident that the same people who created the interactive dialect quiz, the deficit puzzle and the rent-vs-buy calculator will be working on The Upshot.
Hey I'm on board with any site where Amanda Cox introduces statistical models.
FiveThirtyEight is still evolving, and I suspect The UpShot to do the same, so it should be fun to see which way each goes (and what other sites come out of it). For now though, I'm just happy that we get to see this statistics-ish thing happen.
We calculated musical taste scores using data from the National Endowment of the Arts, the Bureau of Labor Statistics, and the U.S. Bureau of Economic Analysis (via the Martin Prosperity Institute) and state level music preferences from Wikipedia. The scores include music genre preference survey data and genre performer concentrations by metro, weighted by that metro's influence on the music scene. We took the scores for each metro and used a spatial statistics method called nearest neighbors to create the heatmap.
Matthew Klein for Bloomberg View explored mortality in America through a slidedeck of charts. The animations in between each slide grows tedious, but the topics covered, going beyond just national mortality rate, are worth browsing. (Although, can someone tell me why the female mortality rate rose between the 1970s and 2000? I know there's a perfectly valid reason behind the trend, but I can't remember.)
The data itself is also worth your time, in case you're looking for a side project. It comes from the Centers for Disease Control and Prevention and spans 1968 through 2010.
I can tell you from experience the data query process isn't the smoothest experience — as much as you can expect from a government site, I guess. That said, the amount of data, with a variety of demographic breakdowns and categorizations, can make for plenty of worthwhile projects. Highly recommended.
A Block is the smallest area unit used by the U.S. Census Bureau for tabulating statistics. As of the 2010 census, the United States consists of 11,078,300 Census Blocks. Of them, 4,871,270 blocks totaling 4.61 million square kilometers were reported to have no population living inside them. Despite having a population of more than 310 million people, 47 percent of the USA remains unoccupied.
See also Stephen Von Worley's map from a couple years ago, which shows blocks in the US with only one person per square mile.
Sebastian Raschka offers a step-by-step tutorial for a principal component analysis in Python.
The main purposes of a principal component analysis are the analysis of data to identify patterns and finding patterns to reduce the dimensions of the dataset with minimal loss of information.
Here, our desired outcome of the principal component analysis is to project a feature space (our dataset consisting of n x d-dimensional samples) onto a smaller subspace that represents our data "well". A possible application would be a pattern classification task, where we want to reduce the computational costs and the error of parameter estimation by reducing the number of dimensions of our feature space by extracting a subspace that describes our data "best".
That is, imagine you have a dataset with a lot of variables, some of them important and some of them not so much. A PCA helps you identify which is which, so the source doesn't seem so unwieldy or to reduce overhead.
As a lesson on conditional probability for himself, Walt Hickey watched 403 episodes of "The Joy of Painting" with Bob Ross, tagged them with keywords on what Ross painted, and examined Ross's tendencies.
I analyzed the data to find out exactly what Ross, who died in 1995, painted for more than a decade on TV. The top-line results are to be expected — wouldn't you know, he did paint a bunch of mountains, trees and lakes! — but then I put some numbers to Ross's classic figures of speech. He didn't paint oaks or spruces, he painted "happy trees." He favored "almighty mountains" to peaks. Once he'd painted one tree, he didn't paint another — he painted a "friend."
Other findings include cumulus and cirrus cloud breakdowns, hill frequency, and Steve Ross (son of Bob Ross) patterns.
This chart-map-looking thing from Nightly News is making the rounds, and it's not good. I'm opening the comments below for critique so that you can release your angst. Signed copy of Data Points goes to a randomly selected commenter the end of this week. Have at it.
Earthquakes are in the news a lot lately. A quick search shows a 7.6 off the coast of the Solomon Islands, a 6.6 in Nicaragua, and a 7.1 off the southwest coast of Papua New Guinea, and this was just last week. Not good news at all, but just how common are these earthquakes? Can we look back farther? Yes. In addition to a real-time feed of earthquakes, the United States Geological Survey maintains an ever growing archive of earthquakes detected around the world, and they make it easy to query and download.