Saturday, 28 March 2020

Covid-19: A short video to explain exponential growth using the latest UK data

I've compiled a short video to:
* explain the mathematics behind exponential growth
* compare how well the latest data (for Covid-19 deaths in the UK) fits an exponential growth curve

* describe what can be done to "flatten the curve" - and what that means mathematically.

Watch it on YouTube here:

There's also a follow up here from 31/03/20 with, by request (!) some more about logarithms, then going on to say more about using log graphs to test for exponential growth and then using that to show that growth rates are hopefully now starting to fall in Italy and Spain.

And there's a third video below, posted on 04/04/20. This one looks at:
* The equation that helps model the 'S-curve' when an exponential curve flattens out and reaches a maximum limit
* How the growth in Covid-19 in the UK compares with Italy and Spain - two countries a week or two 'ahead' of the UK in their death statistics

Here's a fourth video, posted on 10/04/20 that looks at:
* What is meant by 'reaching the peak' of the logistic growth curve - but warns that this will still be a good way from the end of the crisis;

* Uses the revised daily death data for England to estimate some key variables more precisely and uses them to model a simplified growth curve based on those numbers.

Here's a fifth video examining the difficulty in analysing data from the UK and shows:
* How the official daily announcements of deaths have substantially underestimated the actual numbers per day
* How using maths - and comparisons with real data from a country like Italy - can help unpick the confusion - although still leaves plenty of uncertainty about exactly how quickly cumulative totals will continue to grow over the days and weeks ahead

1 comment:

Trevor said...

Hi Martin,

Thanks very much for taking the time to put together these informative videos. I look forward very much to a sixth edition, if yoiu have the time and inclination, following the new stats we expect soon.

Trevor Palmer,