Intel® Fortran Compiler
Build applications that can scale for the future with optimized code designed for Intel® Xeon® and compatible processors.

COVID 6

JohnNichols
Valued Contributor III
385 Views

Capture.JPG

 

the NY TImes data contains negative numbers of deaths -- you need to code to pick up the errors -- only problem is working out how to fix them 

 

Capture1.JPGThe ODE plot for the SEIR model -- alpha is about constant for the last few weeks and really it is in a narrow range for the whole time.  

Stay well and stay home

 

0 Kudos
7 Replies
jimdempseyatthecove
Honored Contributor III
385 Views

John,

I think there is a strong case to be made for GI -> GO with the data.

If you were to look at the listed causes for deaths in nursing homes, hospice, assisted living, emergency room, etc... one might have to conclude that COVID-19 miraculously cured: cancer, stroke, heart attack, and other symptoms of death in those facilities as many of these reported instances have gone to near zero.

It would be interesting to subtract from the reported the statistical expected death rate for each facility extrapolated from the prior three years of data (less any non-COVID-19 reports).

Jim Dempsey

0 Kudos
JohnNichols
Valued Contributor III
385 Views

Jim:

I have to respectfully disagree, there are excess deaths -- 

This is the current Texas data -- the SEIR model is not perfect, but it will get a lot better as we study the deaths.  This is not good.

CaptureT.JPG

0 Kudos
jimdempseyatthecove
Honored Contributor III
385 Views

Can you comment on this?

https://www.youtube.com/watch?v=If4xf-g5I48

Jim Dempsey

0 Kudos
FortranFan
Honored Contributor II
385 Views

Nichols, John wrote:

.. I have to respectfully disagree, there are excess deaths -- 

This is the current Texas data -- the SEIR model is not perfect, but it will get a lot better as we study the deaths. ..

@John,

It's effectively an impossibility to make any determinations of "excess deaths" at this stage: in terms of death statistics for most states in the US, the values are just about getting validated for year 2018.  Given the pandemic, it'll take a while to "dot all the i's and cross all the t's" for year 2019 death certificates first.  And it will be anywhere from 2+ to 5 years before the data attain any level of reliability for calendar year 2020.  Only then can one begin to address @Jim Dempsey's point which, at its root, asks whether there is a "labeling" and "date" game going on with covid-19 deaths.  Clearly, there is tremendous, tremendous incentive with massive Federal aid on offer to boost the numbers.

Any computations prior to reliable data becoming available are of little to no value.

Not that anyone is being asked to stop all this most premature number-crunching: just that, many informed citizens need to be highly vigilant no one in any sphere of influence in terms of policy decisions uses all these unbaked fits and plots to make any further decisions to rain any further damage on citizens in the great states of US and elsewhere.  Humankind has already extended tremendous leeway at incalculable cost with insufferable lockdowns to one BUGGY computer code with IRREPRODUCIBLE results and OPAQUE assumptions and QUESTIONABLE parameters at Imperial College, of all places, by Ferguson.  One can only hope the same ain't repeated over and over to create an echo chamber where each useless model refers to and cites umpteen other similar ones to create a 97+% consensus among armchair-flakes masquerading as "scientists" to further any and all other COMMANDMENTS to be issued by unelected bureaucrats seeing themselves as gods and prophets leading the citizenry away from the "sins" of simply seeking independent livelihoods and lives for themselves and their loved ones.

And by the way, whatever number-crunching one may wish to do with any of the information on tap thus far like those from NYT, one hardly needs Fortran for that.  In fact, one can do better, faster, and more visual processing of all those values and perform any regression analyses on them using any of the easily amenable approaches (R, Python, MATLAB, SAS, Excel, graphical interactive environments on the cloud, etc.) besides Fortran.

0 Kudos
JohnNichols
Valued Contributor III
385 Views

Jim:

Quarantine is a brutal method to stop a virus, but it has been used for millennia , see bible on skin diseases.  

It works to stop exponential growth.  The US data clearly shows exponential growth in the early weeks.  The knowledge that CV is temperature related is a nice theory -- it might be correct, but you have to make a decision --  I do not disagree with the decisions made.  

For instance leaving schools open, is not that deadly on children - aside from the ones who will get the nasty heart stuff -- but they can infect themselves, they probably will not notice it -- but it is deadly killer for anyone over 40 at slightly different rates, so the child kills the parent -- this has happened down the centuries and we wind up with a lot of orphans -- which is ok as we can reopen the old orphanages and let nice people run them. 

 

So I appreciate the guys' point of view, we are entitled to our viewpoint, but that does not mean that you are accurate in your viewpoint.  

Interestingly you and I can watch Texas for the next few weeks, it may demonstrate the problem? 

AAUnited_States_of_AmericaLR.png

 

Of course we are lucky we have had 50 years without a lot of these diseases because our parents had seen the misery of disease and vaccinated like mad -- at least they did not have it as bad as the 16th century.  

Capturea.JPG

 

of course we could have used the old method of taking the sick people and putting them on a Hawawain Island.  

I have had bouts with communicable diseases, both times I ended up very sick on a drip -- the second one was one that required absolute isolation -- it is was not nice -- a drip for a week is not a fun way to eat. 

So I appreciate his viewpoint, but he is not someone I would listen to, he has the right to speak and I have the right to ignore. 

I am playing with some old spag bog code and It is not Fortran fun. 

 

 

0 Kudos
jimdempseyatthecove
Honored Contributor III
385 Views

>>Quarantine is a brutal method to stop a virus, but it has been used for millennia , see bible on skin diseases.  ...It works to stop exponential growth...

And it may kill the patient. In this case the country. What is the cost in lives of implementing the stay at home directives? This will be in terms of future lives lost due to reduction in healthcare visits, closing of hospitals and clinics, delay or termination of medical research, suicides, increase in stress related issues, etc... From FF's input, it may be from 2-5 years before we know the true cost (except I fully expect this will be hidden unless it is politically expeditious). 

What some (many?) of the so called experts are espousing is analogous to having a single zone heating and air-conditioning control for the entire country (one control for everybody).

Jim Dempsey

0 Kudos
JohnNichols
Valued Contributor III
385 Views

Dear Jim:

I just Fortran the models, I do not make the decisions. 

I can suggest that my mother would tell you not to roll around in the mud with a pig because the pig enjoys it. 

I have buried to many children due to medical mistakes to be a good person to argue this point with, if I could give you all I own and get back a day with those children I would, so I suggest we agree to disagree -- you will not convince me and I you, let us remain friends.

Warm regards

John

0 Kudos
Reply