Serving the Quantitative Finance Community

 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 18th, 2020, 6:42 pm

Telegraph 16 May 2020

Neil Ferguson's Imperial model could be the most devastating software mistake of all time

The boss of a top software firm asks why the Government failed to get a second opinion from a computer scientist 

Read on
https://analyticsindiamag.com/the-most- ... criticism/
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 18th, 2020, 6:54 pm

This revised codebase is written in C++ and is split into multiple files, but the original program was a 15,000 line single file that was written over a decade, which the ex-Google engineer considers it to be an extremely poor practice.

No kidding, Sherlock. A 6-year knows that.
BTW there's no C++, it's all K&R C 1972.
But they use a C++ compiler because C compilers don't really exist.
Last edited by Cuchulainn on June 18th, 2020, 7:19 pm, edited 1 time in total.
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 18th, 2020, 6:57 pm

This whole episode of Neil Ferguson exposes one thing for certain that the data scientists and policymakers can be fooled by randomness and their actions can lead to irreparable outcomes. This also urges for the importance of doing data science in the most robust and transparent way possible.

Woolly waffle and apple pie..

I agree, but the fooling is more mundane; the random number s**t is 50 years old and flawed. Time for a course.
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 18th, 2020, 7:30 pm

Anonymous quote (experienced developer)

"The problem is that in general people on decision-side do not know Jack shit about modelling part, nor about creating software to execute this required modelling. They actually cannot even understand that these two issues could be different. They then rely on "experts" to do it. It also seems in this case, that this expert status for software part has been based purely on academic merits, within very small group of people. This does not seem to matter. Talking heads will get their numbers and soon these numbers will be diffused to general public as "common sense". The public only see some chart presented these talking heads, tagged with some academic authors with sexy titles. In this stage, one needs truly big guns to shoot down the illusion. People have already huge amount of vested interest to give up all lies."
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 19th, 2020, 9:24 am

Anon 2

I’d definitely like to have an evening conversation on the creation process of this .. whatever you might call it, with the person who has completed this humungous task. Now, I am not as matured in C++ as you are, but even I can see that this program has been created by someone, who
 
1.       obviously knows C (dangerously a lot of C in order to make this program to be such impenetrable mess)
2.       has not been truly understanding what is C++, but still tries to use it like “C++ is just C with classes”
3.       has never heard about design patterns, nor any general software architecture issues, not even in the simplest level
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 24th, 2020, 6:52 pm

Looks like the bespoke multimorbidity connection to COVID-19

https://www.thelancet.com/action/showPd ... %2930264-3
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 27th, 2020, 5:24 pm

Interview Prof. John Ashton

www.youtube.com/watch?v=FjqfXCJtWUY
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: Models for Covid-19

June 27th, 2020, 8:03 pm

The R number and growth rate in the UK
Latest R number range for the UK 0.7-0.9
Latest growth rate range for the UK -4% to -2% per day


Yay! We are safe! No.
jensen-epidemics.jpg
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 28th, 2020, 8:33 am

According to Jacob-Rees Mogg, it's caused by the weather

https://www.independent.co.uk/news/uk/p ... 85616.html
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: Models for Covid-19

June 28th, 2020, 10:21 am

R0 can definitely depend on the weather, or anything for that matter. There's a whole body of research based on clear and simple models, showing that heterogeneities in the system can lead to epidemic outbursts or endemic states, even if the disease seemingly died off (sometimes one hotpot is enough). It's maths, so it happens regardless of the seasons/weather. The Great Britain scientists seem to be too great for such basic models, though. Not to mention the complete disregard of the economic consequences of their advice - government taking the side of big corpos again?
Last edited by katastrofa on June 28th, 2020, 12:59 pm, edited 2 times in total.
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 28th, 2020, 11:04 am

Pride goeth before destruction, and a haughty spirit before a fall.

Proverbs 16:18
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: Models for Covid-19

June 28th, 2020, 12:56 pm

British exceptionalism. I have to admit I fell for it.
I genuinely hope they know what they are doing.
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 28th, 2020, 1:00 pm

Cured?
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: Models for Covid-19

June 28th, 2020, 1:07 pm

Yes. But I know it's incurable for someone who grew up in this culture - I have the Ukrainian version of this disease, some have it Irish ;-)
 
User avatar
Cuchulainn
Topic Author
Posts: 20252
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Models for Covid-19

June 28th, 2020, 2:54 pm

Yes. But I know it's incurable for someone who grew up in this culture - I have the Ukrainian version of this disease, some have it Irish ;-)
It's at moments like this in the project that the role of the requirements analyst becomes  indispensable as the voice of reason.