COVID-19 California County Trends

These graphs work best on a computer. Phones, not so well. Phone users: check out the blog.

These graphs are interactive (and not very useful or intelligible if viewed without interacting). You can click-and-drag to zoom, single-click on a particular item to turn it on or off, or double-click to show just that item.

Severity

The following graph attempts to show broadly “how bad” the epidemic is in each county, based on daily cases. It does this by multiplying DCPM, by the week-over-week growth factor in new cases. Here are the top 20:

And all counties severity:

Total Cases

Daily Cases

Daily Cases Per Million

To really compare the numbers of different regions, we need to adjust for population. Hence “daily cases per million” (DCPM), which lets us see how relatively bad a region has it. I define a region as “warm” if DCMP is more than 100, and “hot” if DCPM is more than 200. You’ll definitely want to click-and-drag to zoom in to the busy part of this graph, since there are a few wild outliers that smash everything to the bottom:

The graph below is similar but shows all countries, and shows the moving average rather than the individual days with trendline:

Daily Cases Trajectories

The next graph shows the relative trajectories in daily-cases-per-million of the counties, in days since each county reached 100 DCPM (those that haven’t are not shown). So each county’s graph is shifted in time, so they align horizontally at the moment they reached 100 DCPM:

Daily Cases Growth

This next graph is the “rope chart,” which shows the daily growth of new cases each region. This allows to predict the direction a region is headed. When a line crosses zero headed downward, that region has turned the corner toward recovery. I define a region as “warm” if growth is greater than 1%, and “hot” if growth is greater than 5%. Remember that you can double-click a name at the right, to see a graph of just that region. Here again you’ll probably want to zoom in with click-and-drag to see the details better:

And the rope chart for all counties:

This next graph is similar to the rope chart, in that it shows percent growth. But rather than fitting a broad trendline, it uses a smoothed moving average, to show more of the fine details of the trajectories of each county. Here is the growth for the top 20 DCPM counties:

And here are all counties (this is mostly just a snarl unless you double-click the legend to select a particular county):

Testing