The Paycheck Protection Program was established to provide aid to small businesses. It’s a $669-billion loan program. The data for 4.8 million loans, amounting to $521 billion so far, is now available from the Small Business Administration.

For loans less than $150,000, you can download data for all states individually. Data for loans that were more than $150,000 can be downloaded as a single file. Look up business name, type, address, and loan amount range, among several other fields.

Seems like it’s worth a closer look.

Update: The Washington Post made a search interface for the dataset.

There have been assertions that increased case counts are all from increased testing. As you might expect, it’s not so clear cut. Andrew Witherspoon and Caitlin Owens for Axios show the changes in testing against changes in cases.

So in the wideout view of every state, the more-testing-more-cases assertion isn’t so straightforward.

ProPublica provided a similar comparison a couple of weeks ago, but I like the difference charts here for every state. They make the gaps more obvious.

Going with the shaded triangle peaks to show case counts and deaths, The Economist visualizes the current status across Europe:

To assess how European countries are coping, and to monitor the danger of “second waves”, The Economist has assembled data on covid-19 cases and deaths for 39 countries, and for 173 sub-national areas for which data are available (see map above). We present the total number of deaths per 100,000 in the population. We also break down the infection and death rates for the past seven days to give a better sense of whether the virus is most active.

For Reuters, Feilding Cage provides a series of interactive tidbits to demonstrate why time perception feels like a mess these days.

Various factors skew our perception different directions. On emotion:

A busy day usually goes by quickly, but for some the opposite is true in 2020. Frontline healthcare workers, for example, know they are at high risk of exposure to the coronavirus, and the resulting anxiety heightens their attentiveness and slows their perception of how quickly a day passes.

For others, however, time can fly during joyful moments, such as a video catch-up or dinner with friends.

There’s nothing on why it feels like time is moving backwards though.

Some attribute increased Covid-19 case counts to increased testing. While that is certainly part of the reason, it doesn’t explain it all when you compare testing rates against the increase in positives.

Charles Ornstein and Ash Ngu for ProPublica:

In other states, including Arizona, Texas and Florida, which did not see a wave of early cases and deaths, the increase in positive results has far surpassed testing growth. In Florida, testing has even decreased a bit comparing the seven days through Tuesday to the same period before Memorial Day. (Florida recorded an abnormally high number of new tests on May 20, which may have inflated the rolling average on May 25.)

But what about decreased virus-related deaths? Death doesn’t come right after a positive test, which means death rate doesn’t increase at the same time that cases increase. There’s a lag. So we’re not in the clear yet.

Sometimes a chart type seems terrible. But you’ll never know until you try.

Using a voronoi map, David Yanofsky for Quartz mapped the places in the US that are the farthest away from legal fireworks sellers in case you need to get away from the early celebrations. Or, people could just stop setting off fireworks at 1am. That would be okay too.

NASA’s Solar Dynamics Observatory has been taking a picture of the sun every 0.75 seconds for a decade. Above is a time-lapse of the resulting 20 million gigabytes of data:

This 10-year time lapse showcases photos taken at a wavelength of 17.1 nanometers, which is an extreme ultraviolet wavelength that shows the Sun’s outermost atmospheric layer — the corona. Compiling one photo every hour, the movie condenses a decade of the Sun into 61 minutes. The video shows the rise and fall in activity that occurs as part of the Sun’s 11-year solar cycle and notable events, like transiting planets and eruptions.

Lauren Tierney and William Neff for The Washington Post used a rotating globe to show how connections between countries quickly shut down as the coronavirus spread.

I’m looking forward to when we get to watch the map in reverse.

Dan Kois walked all of the blocks in his ZIP Code and collected data on whether houses used serif or sans serif fonts for their house numbers:

Between March 10 and May 25, I walked every street in Arlington, Virginia’s 22207, a total distance of about 200 miles, according to my Fitbit. The ZIP code covers a lot of territory, 6.37 fairly densely populated square miles, from the Potomac River and the Washington border at the east to East Falls Church and the McLean border at the west. It includes some of Virginia’s richest and whitest neighborhoods, but also a number of apartment buildings and townhouses along Lee Highway and the historically Black middle-class neighborhood of Hall’s Hill.

Here’s the final result in a Google map. There doesn’t seem to be much of a pattern in Kois’ ZIP code, but it’s a good walking exercise. I wonder what other types of data one might collect while walking a large area.

Axios provides a straightforward state map showing the percentage change in the 7-day average for confirmed Covid-19 cases. Numbers are up in a lot of places.

Increased testing does not explain away these numbers. Other data points make clear that we’re seeing a worsening outbreak, not simply getting better data.

So frustrating.

Cameron Bennett made a relief map of Idaho, completely out of LEGO bricks:

In March, the COVID-induced quarantine sent me home, but more importantly, to my childhood Legos. What resulted was too much time, money, and effort spent entertaining some combination of my childhood and young adult self. By building a map. Out of Legos.

What have I even been doing with my time.

Here’s the good stuff for June.

Using a wide array of sources, The New York Times shows how the virus spread at a granular level. The foundation is a map with moving dots, and the piece takes you through movements based on cell phone data and air travel in concert with known coronavirus cases.

Your jaw might drop a few times as you go through it.

Back in March, NYT published a similar work for how the spread started in China. We were talking about washing hands and restaurants closing. Thinking back to then and seeing where we are now stings.

Even though there was supposedly a person in the decision-making process and a surveillance photo wasn’t actually Robert Julian-Borchak Williams, he still ended up handcuffed in front of his own home. Kashmir Hill reporting for The New York Times:

This is what technology providers and law enforcement always emphasize when defending facial recognition: It is only supposed to be a clue in the case, not a smoking gun. Before arresting Mr. Williams, investigators might have sought other evidence that he committed the theft, such as eyewitness testimony, location data from his phone or proof that he owned the clothing that the suspect was wearing.

In this case, however, according to the Detroit police report, investigators simply included Mr. Williams’s picture in a “6-pack photo lineup” they created and showed to Ms. Johnston, Shinola’s loss-prevention contractor, and she identified him. (Ms. Johnston declined to comment.)

In crime shows, they often have this amazing tool that turns a low-resolution, pixelated image of a person’s face to a high-resolution, highly accurate picture of the perp. Face Depixelizer is a step towards that with machine learning — except it seems to assume that everyone looks the same.

There might still be some limitations.

Welcome to whose bar chart is it anyway: where the geometries are made up and the numbers don’t matter. [via @dannypage]