Monotonic 2-Degree Polynomial Curve Fit (e2b)

Note: This website has been archived. Our data source of COVID-19 confirmed cases (John Hopkins University) stopped providing data in March 2023.

Unfortunately, without incoming per-country data on COVID-19 positive test results, we cannot continue to predict the future spread of the virus during the ongoing pandemic. As a result, this website is being archived and the historical charts shown will be provided for archival purposes only.

Disclaimer: This page shows a hypothetical model attempting to predict the future spread of coronavirus. It was automatically generated using an open-source algorithm. Such a model is inexact and may be wildly inaccurate. See below for more details.

For a list of other Extrapolation Models against the COVID-19 data, see our Models Page

About This Model

The chart above attempts to predict the future number of COVID-19 cases for the given region. The Y-axis shows the number of people who have tested positive and the X-axis is time. The vertical red line (labeled "Hypothetical Data") is the present day (on our latest graph). Everything to the right of the red line is the future, and its lines are calculated from the extrapolation model.

In this model, there's three distinct predictions:

  1. Last 30 days The purple line does a second-degree polynomial curve fit against the most-recent 30-days of data in the dataset
  2. Last 7 days The green line does a curve fit on the most-recent 7 days of data
  3. Last 3 days The red line does a curve fit on the most-recent 3 days' data

You'll find that, depending on recent events in the past 3-days, the red line will flap around more wildly day-to-day than the others. This can be a useful coorelation indicator for the consequences of recent events, such as prematurely ending lockdowns.

The "e2b" model shown here attempts to fit a curve using a monotonic second-degree polynomial using numpy's poly1d() function. This is an improvement over Coviz's first model (e2a) in that it prevents the fit line from decreasing. Still, this model is very simple and has many shortcomings.

  • Model Short Name: e2b
  • Model Full Name: Monotonic 2-Degree Polynomial Curve Fit
  • Submitted By: Michael Goldenberg

Pros

  1. Easy to write in python
  2. Easy to comprehend how it works

Cons

  1. Expects exponential growth
  2. Doesn't take into account herd immunity
  3. Doesn't take into account history of previous pandemics
  4. Assumes y is infinite, yet there's a finite max human population
  5. The monotonic implementation causes some nearly-negligable precision to be lost in the polynomial curve fit plot

If you'd like to submit your own Extrapolation Model that fixes some of the issues above, see our guide to developing and submitting models.

Country-Specific Charts

You can view graphs generated from data specific to individual countries below:

Previous Graphs

You can view graphs generated from data on previous days below: