Coronavirus Data Plotter

Coronavirus Data Plotter

Welcome!

On this data portal, explore different aspects of COVID-19 / Coronavirus time evolution. I put this together because I saw a lot of different visualizations and data views over many different sites but no way to explore them all within the same tool. I hope this will be as helpful for you as it was for me, in helping process and understand the torrent of information and numbers we see in the news and on our phones every day.

On this page you can set the x, y, color, and time axes to be different aspects of the underling data, and also customize what regions to display. When you find a set of projections that you like, copy a link to share or save the image to your computer. If you find a configuration that is particularly helpful or informative, I'd be happy to hear from you!

For some examples of configurations that might be useful, check these out: (click on the links to load them)

Bhatia-Reich plots, made famous by this MinutePhysics video. Exponential growth shows up as a nice visual "highway" that regions travel along, and is easy to observe when they exit exponential growth.
Growth curve plots, displaying curves on the news that we are supposed to be flattening. Growth scaled based on population. A logistic fit is added to visualize bell curve projections.
Percent growth plots: periods of perfectly exponential growth show up as horizontal lines. Colors reflect absolute counts, to show more info at a glance.
Mortality Rate Plots: The angle/slope of travel at each point in time approximates that moment's cumulative fatality rate. The more upwards-angled, the higher the rate; the more rightward-angled, the lower. 13-day lag applied, as per this CDC recommendation.
"Baserunner" Plots: Regions all must move through a circular track, from "home base" at the bottom, counter-clockwise; when they get back "home", they are effectively out of the epidemic. Set-backs will visually appear as backtracking or getting lost. Inspired by Danny Dorling's tornado plots

Linear Growth: Assumes that every day brings a fixed change. Can often be found "in between" exponential growth and exponential decay, when growth is controlled but not yet eliminated. Can also occur if reporting measures (like tests) has a fixed capacity per-day.

Daily Change: Increase in count per day

Exponential Growth: Assumes that the change every day is proportional to the total amount of the previous day. This is the hallmark pattern of uncontrolled epidemic growth, because each infected person becomes themselves a source of future infection.

Daily % Growth: Percent growth per day
Doubling Time: Time it takes for count to double

Logistic Growth: Assumes that exponential growth in the beginning will steadily transition into exponential decay as limiting factors overpower growth factors. In order to fit properly, this transition must have already begun. Otherwise, it is indistinguishable from perfect exponential growth.

Final Total: Maximum count after growth ends
Date of Peak: Date when growth is highest
95% Date: Date when 95% of maximum count is reached

Exponential Decay: Assumes that the decrease in growth rate is proportional to the total growth of the previous day. This will happen if suppressing factors uniformly overpower spreading factors.

Final Total: Maximum count after growth ends
95% Date: Date when 95% of maximum count is reached

Quadratic Growth: Assumes the daily increases change at a constant rate. Also known as "a linear fit on daily increases". Can occur if daily increases are stable, yet experience a steady drift in one direction or the other.

Acceleration/Deceleration: Change in daily increase per day
End Date: If decelerating, the date when growth ends
Vertex Date: If accelerating, "x-intercept" (date of 0 growth) when projecting backwards (useful for gauging steepness)

Base Projection: The underlying dataset to begin building your axis data from. This data will be run through the transformations below to generate the data you want to plot.

Transformations: With the basic building blocks below, you can generate many different unique aspects to visualize. Select different ones to see how they affect your data, and what options they can be configured with. Chain as many as you want! The result of all of your transformations on your base projection will be plotted.

Scale: How to "spread out" the values on the plot, relative to each other value.

Linear: "Normal" scale — fixed distance on plot corresponds to fixed change in value. Can be the most "truthful"; however, details for small scales can be lost if larger values are present.
Logarithmic: Fixed distance on plot corresponds to fixed proportional change in value. This is the natural scale for exponential growth and decay, because they look like straight lines. Also useful for being able to see detail at all scales.
Symmetric Log: Log scales normally cannot show negative values; symmetric log plots negative and positive numbers on flipped log scales to preserve details at all scales, and stitches them together around 0.