Serving the Quantitative Finance Community

 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

May 18th, 2020, 8:23 pm

What the scientists should do is port the Imperial code to Python (I'm serious), after having done a good online course. Avoid Ball of Mud 2.
 
User avatar
bearish
Posts: 5186
Joined: February 3rd, 2011, 2:19 pm

Re: Models for Covid-19

May 18th, 2020, 8:48 pm

Not unrelated:

I've started to read (sorry haven't saved any refs) that R_0 isn't necessarily the greatest thing since sliced bread (no kidding!), and even that R_0 going above one might not matter, something about variability in R_0 makes things safer.

I am confused.

If you have a simple exponential growth (early stage of epidemic) with a growth rate which varies across populations, say, then the average number of infected is not given by the number of infected using an average growth rate. In fact it's worse. It's simple Jensen's Inequality.

There's probably a Certainty Equivalent Rate of Infection you could calculate.
You heard it first here: a risk neutral rate of infection!
 
User avatar
Paul
Posts: 6604
Joined: July 20th, 2001, 3:28 pm

Re: Models for Covid-19

May 18th, 2020, 9:51 pm

I like it! I think I’ll diversify, catch a little bit of this, a little bit of that.

We have risk-neutral rate of infection, certainty equivalent, implied rates, and I’m sure we can back out the hospital price of risk.
 
User avatar
Alan
Posts: 2958
Joined: December 19th, 2001, 4:01 am
Location: California
Contact:

Re: Models for Covid-19

May 18th, 2020, 10:14 pm

Hehe -- then there's the sharp ratio when you get vaccinated.

And the martingale clinical trial strategy: keep doubling the dosage until you get a winner!
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

May 19th, 2020, 8:23 am

Not unrelated:

I've started to read (sorry haven't saved any refs) that R_0 isn't necessarily the greatest thing since sliced bread (no kidding!), and even that R_0 going above one might not matter, something about variability in R_0 makes things safer.

I am confused.

If you have a simple exponential growth (early stage of epidemic) with a growth rate which varies across populations, say, then the average number of infected is not given by the number of infected using an average growth rate. In fact it's worse. It's simple Jensen's Inequality.

There's probably a Certainty Equivalent Rate of Infection you could calculate.
You heard it first here: a risk neutral rate of infection!
yes, it's the silly season.
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

May 19th, 2020, 2:35 pm

The Italian Connection

http://maddmaths.simai.eu/divulgazione/ ... oni-covid/

What Quateroni is essentially saying is the need for more inclusive models.

// where's the 'physics' in SIR?
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

May 21st, 2020, 7:31 pm

Small update regarding the SIR code

A number of Fortran routines for RNG from ACM where ported 1:1 (mindlessly it seems) to C. This destroys the integrity and reliability of the code. Back to square 1.

Even Python leaves Fortran as it is (you know, wrappers).
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

May 23rd, 2020, 1:47 pm

Exemption of travellers from island of Ireland from UK 14-day quarantine welcomed


https://www.irishtimes.com/news/ireland ... -1.4260426

Arlene Foster describes exemption as a ‘generous move by the UK government’
What a load of baloney; DUP playing politics again.


I think they have lost the plot.
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

May 26th, 2020, 8:04 pm

According to the Sunday Times report, the key moment came on 12 March when a group of government experts gathered to examine modelling of the spread of the virus carried out by academics at Imperial College London and elsewhere.

This predicted that if no action was taken more than half a million people would die, and that even some limited mitigation efforts would only halve this. The Sunday Times report said this changed the mind of Cummings, who before had been an adherent of the herd immunity idea.
After the 12 March meeting, Cummings changed his view and became one of the strongest advocates in government for tough restrictions to curb the spread of the virus, the Sunday Times said. It quoted one anonymous senior Conservative as saying: “He’s gone from ‘herd immunity and let the old people die’ to ‘let’s shut down the country and the economy’.”
The Downing Street spokesman said: “This is a highly defamatory fabrication which was not put to No 10 by the Sunday Times before publication. The article also includes a series of apparent quotes from meetings which are invented.”

Question: how can someone with a degree in History know this?
Another 'armchair epidemiologist"?

And 17 March BBC showed ODEs on prime time.
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

May 30th, 2020, 10:18 am

It is only now that we are all learning where the first case of Covid-19 was discovered, seemingly from a man from Codogno who was taken up in Lodi hospital (Lodi is a small village in  Lombardy; coincidentally I was twice in Lodi few years ago for business and a wedding). The video below shows how the team got into action. And the second case came ... then the doctors realised they had a pandemic on their hands..
One of the messages coming through is that the authorities ignored warnings from doctors. And most of Europe,USA, Brazil etc. for that matter. Asleep at the wheel.

www.youtube.com/watch?v=qv5GkJfhMko

Here's the story (in Italian) of the journey of "patient #1" from Lodi to Pavia and back again.

https://milano.corriere.it/notizie/cron ... resh_ce-cp
// Lodi has a famous building Banca Popolaro di Lodi designed by Renzo Piano if that means anything to you.
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 3rd, 2020, 12:22 pm

Sweden should have done more to combat coronavirus, health chief says

https://www.rte.ie/news/coronavirus/2020/0603/1145149-coronavirus-sweden/

"At the same time, we have to admit that when it comes to elderly care and the spread of infection, that has not worked. That is obvious," he said. "Too many old people have died here."

He didn't have a parent in a care home?

All models are wrong, some are useful. And Anders Tegnell was soooh sure of his models, He even went to far as to say every other country had got it wrong. As a follow-on, he is not the only scientist who messed up.

Each discipline has its own peculiar ways to brainwash people to think in a certain way. The result is often tunnel vision.
They should all take a course on Requirements Analysis and Elicitation. It is not Rocket Science.
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 4th, 2020, 5:32 am

https://www.spectator.co.uk/article/why ... eplicated-

How is it possible to replicate undocumented code?
Seems Uppsala model was not peer-reviewed..

No one, however, would have predicted this news item from last week: “Covid-19 deaths in Sweden were the highest in Europe per capita in a rolling seven-day average between 12 and 19 May.” It confirmed that Sweden’s state epidemiologist Anders Tegnell’s “mitigation” strategy of allowing shops, restaurants, gyms, schools and workplaces to remain open was a deadly folly. It does not even seem to have produced herd immunity. Just 7.3% of Stockholm’s inhabitants had developed Covid-19 antibodies by the end of April.

Britain saw a wave of support for a government that frittered it away with its repeated displays of tardiness and ineptitude. If Britain shows the dangers of a weak state and incapable politicians, Sweden shows what happens when you place too much trust in a handful of administrators, without first protecting yourself with a robustly argumentative culture that allows you to question whether they are right.

Pride comes before a fall.
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 5th, 2020, 8:35 am

ems Canada caught the bug as well

The Flawed COVID-19 Model That Locked Down Canada

https://www.iedm.org/the-flawed-covid-1 ... wn-canada/

What happened? On March 16, Professor Neil Ferguson of Imperial College London released an epidemiological model that took the world by storm.(2) The report warned that tens of millions would die in a pandemic that was compared to the Spanish flu, the deadliest epidemic in modern times.

And on 17th March the BBC posted the ODEs (not all of us were celebrating St. Patricks' Day)

Image
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 5th, 2020, 7:31 pm

I have been informed that (mostly academics and hobbyists I suspect) are working on a new version. Some  gems
1. 
My limited knowledge of stochastic processes is that it still should abide by some given laws/rules and using thread races to add "noise" is not ideal.

2. Background is that I was looking at setting up some training for our University HPC service on Intel Inspector and spotted OpenMP was used in this code and like to learn approaches researchers take so thought it would be a relevant example for users.

3. Multi-threaded testing is a harder problem, and Intel Inspector - when properly configured - does the job very well, so I can happily recommended it for the training you have in mind. We do a fresh exhaustive test on covid-sim whenever changes are pushed that concern threaded sections. Proper configuration, unfortunately, requires maximum resources and overheads, and anything less than that tends to return a lot of false positives - eg, reported memory leaks that are demonstrably not there in hello-world type examples. Usually though, true errors will be detected as well as the false reports, so if you know what to look for, the cheaper analyses of Inspector are still useful.

4.     We don't really mind which thread gets there first; the differences caused by this will sometimes cause no difference, or sometimes cause a very slightly different arrangement of people to work-places when the population is initialised. The simulation is stochastic anyway (ie, we generate many different realisations using different random seeds), and the difference caused here by the race merely causes a different (equally valid) realisation.

Looks as if the problem is almost solved.

Caveat: The OpenMP C and C++ application program interface lets you write applications that effectively use multiple processors. Visual C++ supports the OpenMP 2.0 standard.

BTW current version > 4.5.
 
User avatar
Cuchulainn
Topic Author
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 8th, 2020, 7:19 am

It seems that the Ferguson is spawning interest in the github repositories among 'external consultants and academics' . First impressions: it doesn't look good, not at all. It's like fixing holes in dykes while what is needed is a delta plan. If it were my money (bit maybe it is  indirectly via EU funding), I would dump the codebase and start again.

The management question, therefore, is not whether to build a pilot system and throw it away. You will do that. The only question is whether to plan in advance to build a throwaway, or to promise to deliver the throwaway to customers.”

― Frederick P. Brooks Jr., The Mythical Man-Month: Essays on Software Engineering