From @haru_cchii on the Twitter:

Local German Gets Bored And Tries To Name All American States
i think i did pretty well

Seems right to me.

It wasn’t just issues with an app. There appears to be many more problems with the Iowa caucus results. The Upshot broke it down with a closer look at the data:

Some of these inconsistencies may prove to be innocuous, and they do not indicate an intentional effort to compromise or rig the result. There is no apparent bias in favor of the leaders Pete Buttigieg or Bernie Sanders, meaning the overall effect on the winner’s margin may be small.

But not all of the errors are minor, and they raise questions about whether the public will ever get a completely precise account of the Iowa results. With Mr. Sanders closing to within 0.1 percentage points with 97 percent of 1,765 precincts reporting, the race could easily grow close enough for even the most minor errors to delay a final projection or raise doubts about a declared winner.

When did voting get so complicated?

To increase anonymity in the Census records, the bureau is testing an algorithm that removes real people and inserts imaginary people in various locations. As you can imagine, this carries a set of challenges. Gus Wezerek and David Van Riper for New York Times Opinion ask what effects this could have on small towns. [Thanks, Gus]

Welcome to another episode of Misleading or Not? where we look at a chart and decide if it’s misleading or not.

For the Financial Times, Alan Smith and Steven Bernard traced the history of railroad construction in America and mapped it over time. Literally. Bernard used digitized versions of old maps and traced each new segment by hand. Tedious, but the result is impressive.

From Lusion, CineShader is a fun editor for those who are familiar with Shadertoy:

CineShader is a real-time 3D shader visualiser. It leverages the Shadertoy.com API to bring thousands of existing shader artworks into a cinematic 3D environment.

The whole project was started as an idea of using a web demo to explain what procedural noise is to our clients at Lusion. After sending out the demo to some of our friends, we were encouraged to add the live editor support and we decided to release it to the public.

Google Maps incorporates data from smartphones to estimate traffic in any given location. Artist Simon Weckert used this tidbit to throw the statistical models off the scent. With a wagon of 99 smartphones, he turned roads red on Google Maps just by walking around.

Nice.

Michael Keller released a new version of Layer Cake:

Layer Cake is a graphics framework built on top of Svelte. It measures your target div and your data and creates scales that stay synced on layout changes. Use these scales to organize multiple, mostly-reusable Svelte components, whether they be SVG, HTML, Canvas or WebGL. Since they all share the same coordinate space, you can build your graphic one layer at a time.

I’m intrigued. (And I feel like I need to learn more about this Svelte.)

I know you don’t get enough election coverage these days and are probably like, “I wish there was a way I could be reminded of who’s running with bouncing heads across my screen whenever I come back to my computer. That would be super.” Well, look no more. The Presidential Hopefuls screensaver by Robert Tolar Haining is what you’re looking for.

Every month I collect resources and tools to help you better visualize data. Here’s the good stuff for January 2020.

Flow Fields, a generative art tool by Michael Freeman, lets you adjust various parameters, such as color, smoothness, and fluctuations, and the flows just keep coming. Pretty.

The code is up on GitHub and is based on Daniel Shiffman’s Coding Train tutorials.

For ProPublica, Ellis Simani and Ken Schwencke compiled an interactive database that you can search:

ProPublica reporters spent months collecting the lists as they were originally released by each diocese. They then made them searchable via a public database in order to provide victims of clerical abuse and members of the public a way to search across all of the released lists.

More than 6,700 names are included in the database, and over 5,800 of them are unique. A little more than half of the people named were listed as being deceased. ProPublica did not have the data necessary to merge records with the same name across dioceses, though our reporting on specific clergy indicates that some have surfaced on as many as eight lists.

Unsettling.

The data is also available for download.

Noah Veltman just posted a dataset of 23,463 personalized license plate applications that were flagged for additional review by the state of California from 2015 to 2016. Casually scrolling through, for the plates people request and why they are flagged, this is a goldmine of amusement.

Veltman writes:

This data was parsed from a set of 458 Excel workbooks that the DMV prepared for someone else’s public records request. I received the files as a consolation prize in response to my own related records request, which I was told would cost $2,000 to fulfill otherwise.

Just on this information alone, I think we are obligated to do something with this dataset.

Sometimes illustrating scientific findings is a challenge. Sometimes the illustrations are published anyways, because there are no more options. Sometimes those illustrations end up on a Twitter feed called Science Diagrams that Look Like Shitposts.

Over a year ago, Google released Dataset Search in public beta. The goal was to index datasets across the internets to make them easier to find. It came out of beta:

Based on what we’ve learned from the early adopters of Dataset Search, we’ve added new features. You can now filter the results based on the types of dataset that you want (e.g., tables, images, text), or whether the dataset is available for free from the provider. If a dataset is about a geographic area, you can see the map. Plus, the product is now available on mobile and we’ve significantly improved the quality of dataset descriptions. One thing hasn’t changed however: anybody who publishes data can make their datasets discoverable in Dataset Search by using an open standard (schema.org) to describe the properties of their dataset on their own web page.

I haven’t tried it in a while, but the last time I did, there weren’t that many sources yet, because the indexing partially relies on others to use a standard to provide metadata. Kicking the tires on it now, it still kind of feels like an index of other dataset aggregators, but I’m interested.

Using one of my recent projects as an example, I describe my non-elegant process of making a quick chart.

So get this. There are these things called radio stations that broadcast music using frequency modulation. They call it “FM radio.” You don’t download or stream the music, and you don’t get to choose what songs you want to hear right away, but sometimes you can call locally and request a song you like. It’s also free to listen to if you have this thing called a “radio.” In exchange, you have to listen to “commercials” sometimes where someone tries to sell you stuff. Seems like a fair exchange.

Anyways, Erin Davis mapped these radio stations and their coverage, based on FCC data. She joined the data with radio-locator.com data, which provides music genre. This allowed for the splits above.

Technology is amazing.

File another one under the sounds-good-on-paper-but-really-challenging-in-practice. Kashmir Hill, for The New York Times, describes the challenges of new laws that allow users to request the data that companies collect on them:

Since then, two groups of researchers have demonstrated that it’s possible to fool the systems created to comply with G.D.P.R. to get someone else’s personal information.

One of the researchers, James Pavur, 24, a doctoral student at Oxford University, filed data requests on behalf of his research partner and wife, Casey Knerr, at 150 companies using information that was easily found for her online, such as her mailing address, email address and phone number. To make the requests, he created an email address that was a variation on Ms. Knerr’s name. A quarter of the companies sent him her file.

“I got her Social Security number, high school grades, a good chunk of information about her credit card,” Mr. Pavur said. “A threat intelligence company sent me all her user names and passwords that had been leaked.”

Yay.

I’m not saying these new laws are bad, but maybe get yourself a good password manager and change all those duplicate passwords.

For The New York Times, Jennifer Valentino-DeVries looked at the current state of facial recognition in law enforcement:

Officials in Florida say that they query the system 4,600 times a month. But the technology is no magic bullet: Only a small percentage of the queries break open investigations of unknown suspects, the documents indicate. The tool has been effective with clear images — identifying recalcitrant detainees, people using fake IDs and photos from anonymous social media accounts — but when investigators have tried to put a name to a suspect glimpsed in grainy surveillance footage, it has produced significantly fewer results.

Not quite CSI levels yet, huh.