-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
[No Politics] What you need to know about CoVID-19 by SARS-CoV-2 [No Politics]
- Thread starter andymok
- Start date
- Status
Hey, good news: Sunny areas don't get Covid-19 - let's open up SC, GA, LA, MS, AZ and SoCal! /S
Sorry, I didn't.
This time I got it even before seeing /S.
I only have direct experience with one. And my takeaway is to be highly skeptical of the statistics. They could be low, they could be high, but I'm quite certain that they're highly contaminated.
Maybe you are the one who is willfully ignoring the accounts of NYC officials classifying deaths as COVID related regardless of underlying conditions. There are several articles on the comments of Dr. Birx on the matter. This doesn’t appear to be a conspiracy. No doubt there are people dying because critical surgeries are being postponed due to the virus. Should the deaths of these people be considered COVID as well?
https://freedomheadlines.com/freedo...icious-on-regarding-coronavirus-deaths-video/
A quick scan of that site shows it to be so politicized as to be almost useless as a source of objective information.
Can't wait to read what @Xulonn , @Floyd Toole and other venerable aged members might answer to this disgusting statement
Being a "snark fan" myself, I agree with the tactic of adding a "/s" (end snark) tag, especially when cultural and language differences are present in a conversation.
47 old drugs lab tested that may show promise for Covid-19 treatment:
https://theconversation.com/we-foun...-and-a-whole-new-way-to-fight-covid-19-136789
https://theconversation.com/we-foun...-and-a-whole-new-way-to-fight-covid-19-136789
“Simple Model” Update
I have updated my previously posted (April 15 – Post #222) two-parameter model. This model is a curve fit to the World Totals from https://www.worldometers.info/coronavirus/
The blue diamonds (on the left scale) are the cumulative Total Cases, while the red squares (right hand scale) are the cumulative Total Deaths. The red triangles are the derivative of the Total Deaths curve scaled to allow plotting on the same (right hand) axis. (The derivative is a five-point calculation to smooth the resulting curve and hence trails the actual data by two days).
As detailed in the table, and easily discerned from the derivative, the curves are flattening. The indicated Fits all have an r2 greater than 0.997 (r2 = 1.000 is a perfect fit, r2 = 0.000 is no correlation), and once the regression coefficient drops below 0.9965, I have started a new fit.
While incredibly simple, this model provides a great illustration of how initial projections (no matter how sophisticated the model) severely overestimated the viral “kill rate” and why model estimates have been decreasing over time, as more data becomes available.
I had a chemical engineering professor (D.B. Robinson who assisted in the development of the Peng Robinson Equation Of State and thus knew a thing or two about statistics and curve fitting data) who's First Law was "If it doesn't plot as a straight line it isn't worth knowing" and Second Law was "Don't over-complicate things". Things to keep in mind when reading the models, papers, and projections that are bombarding us.
I have updated my previously posted (April 15 – Post #222) two-parameter model. This model is a curve fit to the World Totals from https://www.worldometers.info/coronavirus/
The blue diamonds (on the left scale) are the cumulative Total Cases, while the red squares (right hand scale) are the cumulative Total Deaths. The red triangles are the derivative of the Total Deaths curve scaled to allow plotting on the same (right hand) axis. (The derivative is a five-point calculation to smooth the resulting curve and hence trails the actual data by two days).
As detailed in the table, and easily discerned from the derivative, the curves are flattening. The indicated Fits all have an r2 greater than 0.997 (r2 = 1.000 is a perfect fit, r2 = 0.000 is no correlation), and once the regression coefficient drops below 0.9965, I have started a new fit.
While incredibly simple, this model provides a great illustration of how initial projections (no matter how sophisticated the model) severely overestimated the viral “kill rate” and why model estimates have been decreasing over time, as more data becomes available.
I had a chemical engineering professor (D.B. Robinson who assisted in the development of the Peng Robinson Equation Of State and thus knew a thing or two about statistics and curve fitting data) who's First Law was "If it doesn't plot as a straight line it isn't worth knowing" and Second Law was "Don't over-complicate things". Things to keep in mind when reading the models, papers, and projections that are bombarding us.
Attachments
As requested by a member, a Q&D chart for only the Excited States. Matched the tail end rather than worrying about initial data matching (Been there done that). Note that the derivative is flat.
The extrapolation to 15/05/2020 is
Cases = 1,776,600 Deaths = 125,500
And just for information, a summary table for the Top 10 ranked by Deaths.
Not quite the Zombie Apocalypse some have been screeching about, but still depressing and very scary!!
Reply
Report Edit
The extrapolation to 15/05/2020 is
Cases = 1,776,600 Deaths = 125,500
And just for information, a summary table for the Top 10 ranked by Deaths.
Not quite the Zombie Apocalypse some have been screeching about, but still depressing and very scary!!
Reply
Report Edit
Attachments
Last edited:
Dave Zan
Active Member
Thank you, that looks very informative.
I have perhaps one issue.
I don't think it is quite correct to say that the initial projections overestimated the viral kill rate, or that the decrease is due to more data.
The initial projections were probably a reasonable estimate of what would happen if the situation continued.
When people looked at those projections and then took action, imposed lock-downs, wore masks and so on, then the situation altered.
So a new model is needed for the new conditions.
It's not due to lack of sophistication in the model.
So, rather than a criticism of your work, my comment is rather that it could hardly be better.
Your numbers for the US alone are indeed scary but look quite plausible.
The flat derivative for deaths is a feature I had noticed just from eyeball inspection of the data.
It's the main reason for my position that it looks ill advised to loosen US restrictions now.
If the current restrictions are in fact loosened then we will see how it compares to your model.
I don't want to set a precedent so you have to do every countries data.
But as an Australian I am obviously interested in your view of the Australian data.
If it's not too much to ask, after your already kind work on the US data.
Best wishes
David
Last edited:
Sorry but must respectfully disagree. Initial model predictions that were splashed all over by the MSM and even some medical journals were in the millions, and had included some adjustment factors for social distancing etc. I'm not going to bother looking up and quoting references as that has been belaboured both here and all over the net. Initial problems with the models (and the interpretation) were the unrealistically high "kill rate' (it is termed many different things depending on who is discussing it) until sufficient data was available to show that it was fractions of a percent, not full percents. First big call to reality (IMHO) was the data from everybody's favourite floating petri dish The Diamond Princess cruise ship
Please also note that all I am doing is presenting these curves to illustrate that even "flat curves (constant derivative)" give massive numbers when extrapolated over time. This has always been the problem with statistical models using fits to time series.
Please also note that all I am doing is presenting these curves to illustrate that even "flat curves (constant derivative)" give massive numbers when extrapolated over time. This has always been the problem with statistical models using fits to time series.
As far as the Ozzy number go, the data is very noisy (small sample size?) but here is the plot:
Extrapolations to 15/05/2020 are
Cases 7285 Deaths 131.
Extrapolations to 15/05/2020 are
Cases 7285 Deaths 131.
Models give guidance based on their accuracy and the information available and its reliability to gain understanding of cause and affect. It is an evolving process.
If a model provides a definitive answer it is no longer a model.
Models are used as a tool to move toward the truth and in the process can discard bumph.
Last edited:
The loved-ones suffer, too. Unless they hold the 'good riddance' viewpoint.
maty
Major Contributor
https://twitter.com/i/events/1256134902258585601
-> The GOVERNOR’S Report To OPEN TEXAS
[PDF] https://gov.texas.gov/uploads/files/organization/opentexas/OpenTexas-Report.pdf
-> The GOVERNOR’S Report To OPEN TEXAS
[PDF] https://gov.texas.gov/uploads/files/organization/opentexas/OpenTexas-Report.pdf
maty
Major Contributor
YouTube Dr. Melchor Álvarez de Mon: "El conocimiento que tenemos del coronavirus es absolutamente limitado"
[ Dr. Melchor Álvarez de Mon, professor of medicine and head of service and diseases of the Immune system and Oncology at the University Hospital Príncipe de Asturias ]
Masks? YES or YES.
[Spanish] http://isanidad.com/160849/dr-melch...os-del-coronavirus-es-absolutamente-limitado/
Remember, YouTube VIDEO CONFERENCIA SOBRE CORONA VIRUS
[ Dr. Melchor Álvarez de Mon, professor of medicine and head of service and diseases of the Immune system and Oncology at the University Hospital Príncipe de Asturias ]
Masks? YES or YES.
[Spanish] http://isanidad.com/160849/dr-melch...os-del-coronavirus-es-absolutamente-limitado/
Remember, YouTube VIDEO CONFERENCIA SOBRE CORONA VIRUS
Last edited:
maty
Major Contributor
La República Checa lanza un plan innovador de "cuarentena inteligente"
[Spanish] https://www.elperiodico.com/es/inte...n-innovador-de-cuarentena-inteligente-7947005
Czech Republic launches innovative "smart quarantine" plan
https://translate.google.es/translate?sl=es&tl=en&u=https://www.elperiodico.com/es/internacional/20200501/la-republica-checa-lanza-un-plan-innovador-de-cuarentena-inteligente-7947005
[ Test, locate and isolate
The Central European country of 10.6 million inhabitants has detected 7,563 infected and has so far registered 227 deaths from COVID-19, one of the lowest death rates in Europe. The new "smart quarantine" system begins by obtaining the consent of those infected.
From there, the telephone operator provides the authorities with data on the user's movements during the last five days, until creating a "memory map". In order to safeguard privacy, this map is anonymous, so the points that appear on it cannot be identified with specific people.
Other employees, with a phone call, help the interested party, with the help of the map, to remember how many people they have met, and which may have been risky contacts due to age or health status. "The map is destroyed after six hours," Prymula recalled to ensure that no personal data is disclosed.
After completing that list of suspected contagion contacts, another team of people, this time from a network of call centers managed by the company that creates the map, contact all of them, and have the power to send these people to preventive quarantine.
Quick tests
Prymula estimates that this system, which was previously tested in the South Moravian region, detects "four out of six people" at risk of infection.
Those who are affected have two days to do a quick analysis in special centers created for this purpose, or they can receive visits from ambulant doctors at their homes.
And the centers that carry out the analyzes are obliged to have the results ready in a maximum of two days, and in case they are overwhelmed, they must send the samples to other centers in the network.
Mobile app
In addition, the Ministry of Health has promoted the eRouska cell application, which has already been downloaded by more than 100,000 Czechs, and which alerts the interested party if they came into contact, even if it was in the queue of a business, with someone later diagnosed with COVID-19.
In the case of confirming a contagion, other experts ask you by phone for the contact list generated by the application, without any of its other users knowing that it is positive for COVID-19.
Unlike the geolocation system, the eRouska works by bluetooth and does not accumulate information about movements, but from users who also have the application, although there have been no official statements about its effectiveness yet.
Both systems are inspired by the use of technology by Asian countries, such as South Korea, to fight the pandemic and have proven useful... ]
[Spanish] https://www.elperiodico.com/es/inte...n-innovador-de-cuarentena-inteligente-7947005
Czech Republic launches innovative "smart quarantine" plan
https://translate.google.es/translate?sl=es&tl=en&u=https://www.elperiodico.com/es/internacional/20200501/la-republica-checa-lanza-un-plan-innovador-de-cuarentena-inteligente-7947005
[ Test, locate and isolate
The Central European country of 10.6 million inhabitants has detected 7,563 infected and has so far registered 227 deaths from COVID-19, one of the lowest death rates in Europe. The new "smart quarantine" system begins by obtaining the consent of those infected.
From there, the telephone operator provides the authorities with data on the user's movements during the last five days, until creating a "memory map". In order to safeguard privacy, this map is anonymous, so the points that appear on it cannot be identified with specific people.
Other employees, with a phone call, help the interested party, with the help of the map, to remember how many people they have met, and which may have been risky contacts due to age or health status. "The map is destroyed after six hours," Prymula recalled to ensure that no personal data is disclosed.
After completing that list of suspected contagion contacts, another team of people, this time from a network of call centers managed by the company that creates the map, contact all of them, and have the power to send these people to preventive quarantine.
Quick tests
Prymula estimates that this system, which was previously tested in the South Moravian region, detects "four out of six people" at risk of infection.
Those who are affected have two days to do a quick analysis in special centers created for this purpose, or they can receive visits from ambulant doctors at their homes.
And the centers that carry out the analyzes are obliged to have the results ready in a maximum of two days, and in case they are overwhelmed, they must send the samples to other centers in the network.
Mobile app
In addition, the Ministry of Health has promoted the eRouska cell application, which has already been downloaded by more than 100,000 Czechs, and which alerts the interested party if they came into contact, even if it was in the queue of a business, with someone later diagnosed with COVID-19.
In the case of confirming a contagion, other experts ask you by phone for the contact list generated by the application, without any of its other users knowing that it is positive for COVID-19.
Unlike the geolocation system, the eRouska works by bluetooth and does not accumulate information about movements, but from users who also have the application, although there have been no official statements about its effectiveness yet.
Both systems are inspired by the use of technology by Asian countries, such as South Korea, to fight the pandemic and have proven useful... ]
- Status
Similar threads
