: New:

I've been doing some small-data tinkering, putting together a dashboard to answer my recurring question: are the COVID levels in San Francisco low enough such that I can go to the supermarket in person, or are the COVID levels so high that I should get my groceries delivered instead?

Up until a few weeks ago, I would check plague.wtf for that info. But plague.wtf hasn't updated in a few weeks; it's the labor of one guy (thank you, Dan Egnor), and he should be allowed to take breaks. My local news site maintains an informative Coronavirus Dashboard, but it's not useful for answering my supermarket question. It's useful for news reporting, which is appropriate. Their charts show danger levels ranging from safe to unsafe to dangerous to OMG apocalypse. But if you're trying to answer the supermarket question, the difference between safe and unsafe is just a couple of pixels. Their chart tells you the new case numbers. But when national disease experts tell you what numbers are safe and unsafe, they talk about cases-per-100-thousand people. E.g., today Santa Clara county sees about 160 new cases per day; San Francisco sees about 60 new cases per day. Is Santa Clara more than twice as dangerous? Well, no. Santa Clara county's population is about twice San Francisco's. If you're estimating the number "What are the chances that one of the other people in this supermarket has COVID?" you need to take that into account. Alas, if I haven't yet had my morning coffee, "divide by eight" isn't as easy as it sounds. And of course I need to do this math without the benefit of caffeine before I can decide "Am I safe going out for coffee or should I stay at home and make my own?"

[screen shot of confusing dashboard]

This dashboard answered my question yesterday, and so I got my groceries delivered today, I made my coffee at home instead of going out, etc, etc. I note that my dashboard's new-cases chart has the opposite problem that my local newssite's chart has: from the chart, you can't tell the difference between somewhat-dangerous levels and OMG-apocalypse levels. That's OK; if I need to know, I can visit the newssite.

(And I'm quite thankful to my local newssite; they make their data publicly available so you can slice it and dice it into whatever charts you like without figuring out the California Health Department's humungous spreadsheet.)

Tags: programming

lahosken@gmail.com

Tags