This is bad. I've been graphing the exp curve for Italy, UK and USA and the USA curve is fearsomely steep. Italy has more cases now but the USA is growing fastest and looks like it will match Italy's on ~March 28th (looking at the most recent 10 points, see below)
The rate of growth in the USA, extrapolated, is just huge. Looking at the whole curve it does not match an exponential well, I think the early part is skewing it. If we take the last 10 data points, they fall very well onto the curves for all 3 countries, and that curve for the USA is horrific. I'm not an epidemiologist nor statistican so it would be irresponsible to put my predicted figures here, but christ, if I'm right the US is going to be reeling in just a few day.
Edit: @mnl below has pointed out that infection is not an exponential but on the whole a logistics curve (s-shaped curve https://en.wikipedia.org/wiki/Logistic_function). He is correct. However in early stages I believe an exponential MAY be a decent approximation (and I'm reading up on it now). However, the inflection in the 'S' (where new infections would start to slow down) would be expected to be at far higher figures than those I'm looking at - a few tens of thousands of reported infections per country currently - and it's not even close to that (populations of multiple millions in each country).
The rate of growth in the USA, extrapolated, is just huge. Looking at the whole curve it does not match an exponential well, I think the early part is skewing it. If we take the last 10 data points, they fall very well onto the curves for all 3 countries, and that curve for the USA is horrific. I'm not an epidemiologist nor statistican so it would be irresponsible to put my predicted figures here, but christ, if I'm right the US is going to be reeling in just a few day.
Edit: @mnl below has pointed out that infection is not an exponential but on the whole a logistics curve (s-shaped curve https://en.wikipedia.org/wiki/Logistic_function). He is correct. However in early stages I believe an exponential MAY be a decent approximation (and I'm reading up on it now). However, the inflection in the 'S' (where new infections would start to slow down) would be expected to be at far higher figures than those I'm looking at - a few tens of thousands of reported infections per country currently - and it's not even close to that (populations of multiple millions in each country).