The present doubling time is 3 days according to the parameter estimated for an exponential fit to figures for Italy done a few days back. https://towardsdatascience.com/covid-19-infection-in-italy-mathematical-models-and-predictions-7784b4d7dd8d
BeechBaby wrote:
Not once have I heard Africa mentioned. Does it still exist?
I had assumed that Africa without the highly developed medical side probably isn’t doing anything like the rate of testing so we don’t really know a great deal.
News nearly always focuses closer to home if there is a story.
I do wonder if the outcome will lead to finally having some more manufacturing brought closer to home.
There is a much younger population there who will be less likely to become seriously unwell.
Qalupalik wrote:
The present doubling time is 3 days according to the parameter estimated for an exponential fit to figures for Italy done a few days back. https://towardsdatascience.com/covid-19-infection-in-italy-mathematical-models-and-predictions-7784b4d7dd8d
As he says, at this stage of the epidemic the predictions of the models will be strongly affected by small variations in the data. Given that the large uncertainties in the number of infected persons, I don’t think you can say much at all. Which indeed is the author’s own conclusion:
But there’s something that still worries me. I’ve been fitting the logistic curve every day since the beginning of the infection and every day I got different parameter values. The number of infected people at the end increases, the maximum infection day is often the current day or the next day (which is compatible with the standard error of 1 day on this parameter). That’s why I think that, although the logistic model seems to be the most reasonable one, the shape of the curve will probably change due to exogenous effects like new infection hotspots, government actions to bind the infection and so on.
That’s why I think that the predictions of this model will start to become useful only within a few weeks, reasonably after the infection peak
I don’t think these “descriptive models” do any prediction apart from fitting the data with a curve a new parameters everyday, the doubling rate each 3 may not even relate to actual infections/transmission but rather the “exponential” improvement in testing capability…obviously, by the time you start getting 50k cases on 10million tests of 10bn population, you can start running full probabilistic simulations (population Markov Models) that account for various details like travel links, population density, weather…but only at 1% infection base and 10% testing these will have reliable parameters, the rest is just speculations, trying to extrapolate 1% & 5% confidence interval stats to 0.0001% data samples on fat tail distributions is none sense, so we really don’t know yet until we fully test about 100m-1bn people
I personally don’t believe Corona virus is 3 times more infectious and 20 more deadly than a normal flu (what is looks now), this is just an artefact from unstable parameter estimation on 0.001% infection base (highly unreliable) and 0.00005% deaths (somehow reliable), but we will only know these things for sure when infection base is about 1% of total population (100m/10bn) which is only done by testing more than 10% of population
The state of Lower Saxony just shut down all schools and preschools/kindergartens, beginning next monday and until April 18. This leads to the unsolved question of who takes care of all the children of essential workers, e.g. in healthcare.
If my wife (a GP) or I have to stay home because of this, the capacity of the health care system will be weakened. Massively so, because many many colleagues have the same problem.
MedEwok wrote:
The state of Lower Saxony just shut down all schools and preschools/kindergartens, beginning next monday and until April 18. This leads to the unsolved question of who takes care of all the children of essential workers, e.g. in healthcare.If my wife (a GP) or I have to stay home because of this, the capacity of the health care system will be weakened. Massively so, because many many colleagues have the same problem.
This is why the UK argues that closing schools etc should be left as late as possible. Including that many kids would have to be looked after by elderly relatives which may not be ideal right now.
I mostly agree with Ibra above. There are way more unknowns than most realise. For example it is (to me) blindly obvious that ski tourism is responsible for a large part of the spread. And still as I write this there are thousands who are desperate to get a ski trip in, and are chasing around Alpine locations which are not yet shut. Those who can drive can do it easily, come back and infect everyone around them, not ~1:3 but ~1:30 because their whole group caught it. And skiing will come to a rapid end now.
This thing does appear to be much more infections (contageous? – what’s the difference?) but that could just be because nobody has any immunity so they “get it” from any exposure at all.
It is really bad we are not getting “community transmission” data anymore. That was the only number which really meant something.
Sitting to have a hot choc at Shoreham today, I see 100% of people touching 100% of door handles with their bare hands. They just don’t “get it”. It will take just 1 infected person and everybody who passed through that day will get it. Some might get lucky if they just happen to wash their hands before touching their face, afterwards.
Czech Republic is closing its borders on Sunday. Not surprising again, though it won’t help with the main influx of infections which according to them was from returning skiers, and most of them live in CZ.
I work with a CZ tourist guide who spendsthe winter season in Tonale , working for a big czech tour operator .
10 days ago she said to me that almost all reservations from czech republic had been cancelled .
Yes, lot of the story about Northern Italy (and some France/Switzerland) may have to do with winter ski season, probably the relevant fact is that a typical ski trip is usually 3 days to 1 week less than virus incubation period and gather more people very closely (incubation is now 5 days average but may go up 2weeks), that still this does not explain the numbers in Austria (while geographically miles apart, it takes more ski travellers influx)
If the underlying cause of Italy outbreak is ski related, then the numbers of cases reported for France/Switzerland is really miles apart from what it is really, but Italy could have its own factors…
