Does this look like we’re done with the outbreak?
One thing that has been bothering me has been the shortage of animated renderings of the pandemic. I assume because they’d be too depressing. So, I searched around and found a data-set that had a time scrubber, then screengrabbed each frame and loaded them into photoshop.
What this makes me think is “wow, they sure had a long time in which the outbreak was relatively minor!” In fact most of the animation is taken up with a few small dots in Oregon, California, Texas, and New York. That would have been the time for the CDC and government to leap into action and quash it. Not try to stop more cases from coming into the country, because it was already here, but to start putting processes in place to identify it, do contact-tracing, and isolate potentially sick people. Like they did in South Korea and not in Northern Italy or the US.
The methodology here is dodgy – I don’t have a way of knowing current infection rate, i.e.: how many sick people are in New York right now? It could be dwindling away, but the data – since this is cumulative – would not show that.
My guess is that the reason people aren’t producing animations is because they look like every scary movie pandemic-outbreak rendering, ever. If you look on google image search for “pandemic outbreak animation” you won’t find much – a reference to the game “Plague, Inc.” and a ton of 3D models of coronaviruses.
From the animation, it looks to me like we’re in the early 1/3 of an outbreak that is eventually going to cover the entire country. Thanks, Obama assholes!
This animation was produced by stepping day by day through the outbreak chart on: https://www.politico.com/interactives/2020/coronavirus-testing-by-state-chart-of-new-cases/ and saving each day as a layer in photoshop, then animating it in Imageready and turning it into an MP4 using FFMpeg. A big “thank you” to the folks at Politico, and a “hey make your dataset animated!” suggestion. I probably could have done it easier if I had spent twice as long whacking on their source code and doing my own version but it was pretty easy to do 1,000 mouse-clicks and alt-tabs in an hour, than to spend 3 hours clutching my temple at obscured javascript. (is that redundant?)
You might be interested in this interactive chart which shows new confirmed cases (averaged over the week) vs. the cumulative total. Mano linked to it recently. Note the play button and corresponding dates below the graph, as well as the handy tools at the top right.
On the logarithmic scale, you would like to see the US dropping like a rock, as some other countries have, but that isn’t happening yet. Instead, it’s just bending away a bit from the two-day doubling line, but of course you can see new cases (the Y axis) remain high and not low. As GM Ben Finegold likes to say in his chess videos, “there’s good, and there’s not good. And that’s not good.” Very instructive.
Most individual states look vaguely similar to the US as a whole: “mildly better” recently — what I mean I suppose is better than “I can hardly imagine worse.” That’s when stay-at-home orders and the like were working as intended, which is not to end the thing but to buy us more time to respond appropriately, as we certainly won’t do. Anyway, you’ll see very little rock-dropping except for a few that had a small total anyway. (Alaska, Hawaii, Montana, and everyone’s favorite colony Guam.)
Some, like Washington, Arkansas, South Dakota, and Ohio, have already experienced a noticeable rebound. Since my analogy is really great, the rock sometimes decides that it doesn’t really feel like dropping anymore and lifts itself back up by its bootstraps. Because these types of rocks have bootstraps. In this case, bootstrap-lifting is a bad thing, and presumably so is the freedom to do it. But it’s definitely a good analogy, I’m sure about that.
This is just *so* not over. When this all started the Partner started reading “The Great Influenza: Story of the Deadliest Pandemic in History” by John M. Barry and now keeps pointing out to me the similarities. Seems that, as a species, we are not real good at learning stuff. People are making the exact same mistakes now that they made back then, and they are just as bad this time as they were any other time in history. Because I’m sure it’s happened over and over and over and…
consciousness razor@#1:
You might be interested in this interactive chart which shows new confirmed cases (averaged over the week) vs. the cumulative total. Mano linked to it recently. Note the play button and corresponding dates below the graph, as well as the handy tools at the top right.
Ah, I had forgotten that; thanks for the reference. That’s a great chart and it illustrates some important things about the trajectory of the outbreak, as reported, in various places.
The US looks as though it may be about to head into a down-turn in the outbreak, but … hey let’s open some bars and maybe a few big box stores, and see if we can turn that around again.
I wish Hans Rosling was still around, helping visualize this data. He was always able to pick out revealing trends.
Some, like Washington, Arkansas, South Dakota, and Ohio, have already experienced a noticeable rebound. Since my analogy is really great, the rock sometimes decides that it doesn’t really feel like dropping anymore and lifts itself back up by its bootstraps. Because these types of rocks have bootstraps.
“Here, hold my beer.”
“The ships hung in the sky in much the same way that bricks don’t.”
@1 beat me to it. Great piece of data viz.
I could weep for you, but then I’d have to weep for us, and I need to stay angry.
But some places have got it right from the start, they have politicians who listened to the experts
from https://www.theguardian.com/world/2020/may/14/the-coronavirus-slayer-how-keralas-rock-star-health-minister-helped-save-it-from-covid-19
Well, that’s unsettling.
How does animation help anything? The data is the data, whether tabular or otherwise.
I think you’re all forgetting that when it gets warm enough Covid19 is just going to disappear – it’ll be like a miracle!
/s
I keep hearing people talking about ‘the virus’ as if it is just a US and Europe thing. Assuming there is no vaccine I picture the last cases in the US as being in early September. With some luck we get a full week, perhaps two, to relax and enjoy life … and then an asymptomatic case from Africa slips in through the not entirely foolproof screening test and we are off to the races again.
By November it goes one of two ways. Either people take heart in a new president and hold on. Otherwise, given GOP resistance to supporting the poor and working class and the prospects of four more years of the same and worse people are going to lose it. Faith, patience and both financial and mental reserves will be gone and people, lacking other options, are going to take what they need to support their families. Things are going to get ugly.
@8 John Morales
Yes, Graphs are completely useless too. I just stare at 100000 rows of database tables for 16 hours straight, when I need to extract some information. Because thats how brains work best.
For people interested in scientific animation 3blue1brown did some good videos on Covid and other stuff. They use manim, a framework they build. It’s on Github, a bit of a pain to set up because it only works with python 3.7 and you need the most recent MiKTeX up to date. Not incredibly well documented either, but it’s pretty intuitive and fun to use (because it’s python).
.
kurt1:
You thought I wrote that animations are “completely useless”?
Must be really tiny text, to fit a hundred thousand rows.
After thinking about it, I recant.