Tracking COVID-19 Cases

(updated 4/10/2020)

While I’ve found lots of great information and some excellent visualizations in various online sources, I still wanted to have the full picture in my own hands. Worldometer has some excellent information and is my go-to source, as previously mentioned. I’m also very impressed with Our World in Data. And there’s a nice World Map view here. But thanks to Johns Hopkins, the raw data is actually available online in Github. It’s updated regularly, usually daily.

So I went ahead and grabbed the repository and had some fun today with RStudio. Here’s the result. I plotted daily cumulative cases (confirmed and dead) for all the countries, US states, and Europe. The “by country” graphs could not fit on one chart, so I broke them up into 3. I could also plot the raw daily numbers, but that data is much more noisy, as there are frequent “catch up” days that cause visible spikes, which are in fact meaningless. The slope of the cumulative curve is what matters.

Note that the only data available from this source is total daily numbers for confirmed new cases, new deaths, and new recoveries. I wish I had demographic data to slice this by, but I haven’t found that yet… I also can’t find the raw data for hospitalizations / critical cases, unfortunately. Don’t know where Worldometer gets theirs (which is not looking good lately…)

There also appear to be some naming inconsistencies that pop up… Unfortunately, I don’t have the time to keep chasing them all down :(

I’ll be updating this every day (hopefully) and will think about automating it. If anyone is interested in the R code, here it is. Also let me know if you think some tweaks to the visuals are needed.

First, US cases, since it’s what matters to me the most:

Next, all the countries besides US and China:

And focus on Europe