hi Andrei, all
I'm not ahead of you! I had never heard of deSolve or delay differential
equations, but managed to see something wrong with my initial thought,
googled something like "differential equations with lag" and found "delay
differential equations" which led to deSolve, which turned out to be what
you - and Alison Hill - are using!
Anyway, I shall take a good look at her code (which uses ode) and then see
what happens with dede.
I'd encourage people to play around with her site, whether or not you want
to look at the code. The display is very user-friendly and the
explanations are clear.
I think that to mimic the current UK behaviour where we are in the
exponential growth phase with doubling time T2 = ln(2)/ln(1.3) = 2.64,
and log growth rate r = 0.26 (see emails / charts yesterday), her default
parameters don't work and we must have a much higher transmission rate.
The closest I can get with her sliders (though you could bypass them and
just run the R code with whatever parameters you like) is to set the
transmission rate (mild cases) to 1.06, and the total population size (log
10) to 7.8. I haven't altered any other parameters.
That will show the scale of the problem, anyway, and what sort of tsunami
to expect: with no interventions, almost 18 million cases at the peak on
day 73, and 1.25 million deaths eventually. I'm not sure if the site is
working properly just now, but we'll see when running the R code without
the display.
I don't think her model considers the age distribution of the population.
thanks
Greg
> Hi Greg and all,
>
> I also prefer the alhill page.
>
> I only got the differential code from a friend anyway, had never
actually used it before. So I think you're probably ahead of me! But it
would be fun to do more... I'd really like to extend the model to stick
more parameters in as I'm not convinced about the infectious models that
they have. But anyway...
>
> Tomas Pueyo has created a new post:
>
> *
> https://medium.com/@tomaspueyo/coronavirus-the-hammer-and-the-dance-be9337092b56
>
> It's pretty good (again) and used the same sort of models -
> specifically, the gabgoh one.
>
> Still, I reckon the models themselves could be improved upon - with
currently available data. But that's for another time ;-)
>
> Best wishes,
>
> -- Andrei
>
>
>
> On Thu, Mar 19, 2020 at 08:18:31PM -0000, Greg Dropkin wrote:
>> hi Andrei
>> the alhill page allows you to access her commented R code directly,
which
>> will help me try to follow it. It uses deSolve. As the ReadMe page
points
>> out, it is easier to alter the R code and then update the display app,
than to alter within the app directly.
>> the gabgoh page has everything buried within a script which I assume is
also used to run the display. If you right click the page and then View
Page Source, you'll find
>> <script defer src='./bundle.js'></script>
>> clicking on ./bundle.js yields an enormous code, incomprehensible to
me,
>> with no reference to deSolve and I've no idea what it does or whether
the
>> solutions are being generated in R, python, or something else, or
whether
>> this was the wrong thing to click on!
>> So alhill looks more use.
>> Greg
>> > On Thu, Mar 19, 2020 at 10:22:36AM +0000, Kornbrot, Diana wrote:
>> >> Really nice link
>> >> Plenty to amuse oneself if isolated
>> >> In fact, this evening, I came across this link:
>> >> * https://alhill.shinyapps.io/COVID19seir/
>> >
>> > Here's another one:
>> >
>> > * https://gabgoh.github.io/COVID/index.html
>> >
>> > Also, if anyone's interested in joining a group of (mostly early
>> career
>> epidemiological) researchers thnking about how we can do something
constructive in relation to this, feel free to give me a shout off list.
>> >
>> > Best wishes,
>> >
>> > -- Andrei
>> >
>> >
>> > --
>> > Andrei Morgan MRCPCH, MSc, PhD (Epidemiology / Neonatology)
>> > https://www.andreimorgan.net
>> >
>> > Honorary Clinical Lecturer,
>> > Department of Neonatology,
>> > Institute for Women's Health,
>> > University College London
>> >
>> >
>> > ******************************************************
>> > Please note that if you press the 'Reply' button your
>> > message will go only to the sender of this message.
>> > If you want to reply to the whole list, use your mailer's
>> > 'Reply-to-All' button to send your message automatically
>> > to [log in to unmask]
>> > Disclaimer: The messages sent to this list are the views of the
sender
>> and
>> > cannot be assumed to be representative of the range of views held by
>> subscribers to the Radical Statistics Group. To find out more about
Radical Statistics and its aims and activities and read current and past
>> issues of our newsletter you are invited to visit our web site
>> > www.radstats.org.uk.
>> > *******************************************************
>> >
>
> --
> Andrei Morgan MRCPCH, MSc, PhD (Epidemiology / Neonatology)
> https://www.andreimorgan.net
>
> Honorary Clinical Lecturer,
> Department of Neonatology,
> Institute for Women's Health,
> University College London
>
>
******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
Disclaimer: The messages sent to this list are the views of the sender and cannot be assumed to be representative of the range of views held by subscribers to the Radical Statistics Group. To find out more about Radical Statistics and its aims and activities and read current and past issues of our newsletter you are invited to visit our web site www.radstats.org.uk.
*******************************************************
|