Author Topic: COVID-19 Data Model Updated Daily  (Read 24749 times)

Beasho

  • Cortez Bank Status
  • *****
  • Posts: 2667
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #60 on: March 25, 2020, 07:07:23 PM »
^Fishman: I follow a site that graphs it. but I won't link it here due to the highly opinionated overall theme.  If the line is what he claims, (?) it's as straight as a string. 

Watched the video.  Long, complicated.  Summarized as follows:

“Reinfection is a rare thing and so in this case as in the SARS COV2 situation it seems as though if the virus follows what we would normally see that if there isn’t that much of a change in the virus hosts develop an immunity to it.”

Here --> https://youtu.be/q4P91VrfPGw?t=845



Beasho

  • Cortez Bank Status
  • *****
  • Posts: 2667
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #61 on: March 25, 2020, 07:10:13 PM »
Since you wanted Predictions: As of 3/25/20

•   Death rates have and will be much lower in the United States than in Italy.  Currently 1.5% vs 9.9%.  ITALY has been 660% more lethal than the United States (as of 3/25 WHO Data).
 
•   Death is correlated to CoMorbidity NOT TO AGE.  The Corona Virus does not kill randomly but kills those with pre-existing Cancer, Lung Disease, Heart Disease, Diabetes, Kidney problems (Google it)

•   Each Country and State will be measured differently and it will be a function of:

1)   Physiology – How old and CoMorbid are the citizens - Bloomberg News showed 99% of fatalities had pre-existing conditions (Google it)
2)   Behavior & Proximity – State or Country Behavioral response combined with density – Did the State Distance and Take Seriously?
 
3)   Health Care – How robust is the health care system?  Socialized medicine in Europe vs. Privatized in the United States will be measured and differentiated.  The better health care will limit death rates.  The more ICU beds the better the recovery.  Similar to the Ebola outbreak years ago.  Less than a handful of people died in the United States from a disease that was 50%+ lethal if left untreated.

USA Will break out with the highest number of cases in the world within the next 3 days.
« Last Edit: March 25, 2020, 07:36:09 PM by Beasho »

Beasho

  • Cortez Bank Status
  • *****
  • Posts: 2667
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #62 on: March 25, 2020, 07:38:19 PM »
The following graph uses the DAYS to DOUBLE to project the number of CASES over the next week.

The WHITE line shows the historic forecast using this approach.  The BLUE Line shows the Actuals vs that Forecast.  You can see the PREDICTION was slightly below the Actuals for the United States.

ITALY (The second set of graphs) were in-line with this doubling rate.

Number of US CASES:
                                3/25 = 55,231
                                3/24 = 46,442
« Last Edit: March 25, 2020, 07:42:50 PM by Beasho »

Beasho

  • Cortez Bank Status
  • *****
  • Posts: 2667
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #63 on: March 25, 2020, 07:44:44 PM »
For Better or Worse that Doubling Rate has proven to be an effective FORECAST mechanism.  See the estimate from 1 week ago posted on Page 1. 
Estimate for 3/24 (7 days from 3/17) Was 43,898 vs. Actual 46,442. 

The White vs. Blue lines above are very compelling.

The good news is that the Countries HIGHLIGHTED in GREEN above: ITALY, IRAN and GERMANY have Days to Double >=10.  This is an implication that the curve is starting to flatten.   Unfortunately the US is in HIGH BURN mode. 
« Last Edit: March 25, 2020, 07:47:42 PM by Beasho »

Beasho

  • Cortez Bank Status
  • *****
  • Posts: 2667
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #64 on: March 25, 2020, 07:51:11 PM »
Testing Results by STATE.  New data needs to be vetted and cleaned up.  THERE is a HUGE disparity in POSITIVE Rates for Testing by State.  Especially in the State of Washington where fewer than 10% were positive in a State that has been BURNING the longest. 

What is going on ? ? ?

I did some quick correlations between metrics and the only interesting thing was the inverse correlation even if WEAK between POSITIVE TESTING RATES and Death Rate %.
 
https://raw.githubusercontent.com/COVID19Tracking/covid-tracking-data/master/data/states_daily_4pm_et.csv



Beasho

  • Cortez Bank Status
  • *****
  • Posts: 2667
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #65 on: March 25, 2020, 07:57:26 PM »
The United States INFECTION CASE RATE is NOW growing FASTER than most other developed nations.  Faster than Everyone but Holland, Spain and the UK.  Italy has SLOWED in number of new cases but Death Rates in ITALY are the Highest in the World.
See projections for Next 7 Days.  This is a combination of 1 WEEK HISTORY with 1 WEEK FORECAST.  It looks reasonable if not terrible.

Model is updated Daily.  Data sources keep moving.  They scrambled the STATE data which I am currently working to update.  I will also add in the TESTING metrics and FORECAST Sheets. 

Link here:
https://app.powerbi.com/view?r=eyJrIjoiNzExYmM2ZWQtZWVmMS00OTI1LWE1NzYtMTU2OTk3ZGE4ZjMyIiwidCI6IjQzODE2ODE2LTY0OTYtNDU4My1iMmVmLWU5Y2U3MTg1NjI4MCIsImMiOjF9

« Last Edit: March 25, 2020, 08:00:28 PM by Beasho »

LaPerouseBay

  • Teahupoo Status
  • ******
  • Posts: 1893
  • downwind dilettante
    • View Profile
    • Email
Support your local shaper

SUP Leave

  • Sunset Status
  • ****
  • Posts: 498
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #67 on: March 26, 2020, 11:59:31 AM »
Thank you Beasho!

My biz partner and I (both numbers guys) have spent a lot of time lately discussing this and have circled April 1 as the inflection point on case growth in America (number of new cases). Your provided charts hit pretty close to that as well. The next few days are going to see huge, scary spikes in cases.

The Imperial College study that predicted 2MM American and 500k British deaths is being revised. First by Oxford, and now by the guy who wrote it at Imperial College (Ferguson). Looks like he was too high on death rate, and too low on transmission rate. Thus creating less hospitalization than they thought.

https://nymag.com/intelligencer/2020/03/oxford-study-coronavirus-may-have-infected-half-of-u-k.html

https://www.newscientist.com/article/2238578-uk-has-enough-intensive-care-units-for-coronavirus-expert-predicts/

UK has pulled CV19 off of the High Consequence Infections Disease list.

Antibody testing is going to become an even bigger deal than CV19 testing IMO. Go Science.




Make paddleboarding great again!

Dwight (DW)

  • Cortez Bank Status
  • *****
  • Posts: 4041
    • View Profile
    • supSURFmachines

OkiWild

  • Rincon Status
  • ***
  • Posts: 231
    • View Profile
    • Email
Re: COVID-19 Data Model Updated Daily
« Reply #69 on: March 26, 2020, 08:49:03 PM »
The point, of course, is to let acquired immunity grow without overwhelming medical facilities. That's the entire idea of flattening the curve. It's not to stop the infection, we don't have any tools available to do that. It's to be able to provide care to people who need it.

It blows my mind how many people still don't get this. People need to stop thinking of it as killing "old" people. A lot of young and otherwise healthy people are getting through it better, but many with medical intervention. If it comes off the rails...   

LaPerouseBay

  • Teahupoo Status
  • ******
  • Posts: 1893
  • downwind dilettante
    • View Profile
    • Email
Re: COVID-19 Data Model Updated Daily
« Reply #70 on: March 26, 2020, 10:31:34 PM »
^Indeed.  The media is a train wreck.   

https://youtu.be/6KS5Fl6CId4?t=4530

As all the experts have been saying from day 1, without testing, we are guessing.  Nothing against models, but they could be so much better... 

From this next clip: "All models are incorrect, some are useful"

In construction, we have a saying, "a picture is worth a thousand words, a model is worth a million." 

To steal a phrase from the Editor of Science magazine.  "We are asking science to build an aircraft while it's flying, and they don't have a full set of blueprints." 

https://youtu.be/oMSY2ZG0nwo

https://www.nytimes.com/interactive/2020/03/25/opinion/coronavirus-trump-reopen-america.html
« Last Edit: March 26, 2020, 11:16:13 PM by LaPerouseBay »
Support your local shaper

tarquin

  • Rincon Status
  • ***
  • Posts: 124
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #71 on: March 27, 2020, 03:53:08 AM »
Beasho amazing work. Maybe a little over complicated though. Here is one I found on another site that I like. Easy to read and the message is clear.

Admin

  • Administrator
  • Cortez Bank Status
  • *****
  • Posts: 5612
    • View Profile
    • StandUpZone
    • Email
Re: COVID-19 Data Model Updated Daily
« Reply #72 on: March 27, 2020, 04:46:05 AM »
The Imperial College study that predicted 2MM American and 500k British deaths is being revised. First by Oxford, and now by the guy who wrote it at Imperial College (Ferguson). Looks like he was too high on death rate, and too low on transmission rate. Thus creating less hospitalization than they thought.

Here are Ferguson's comments on that:

https://twitter.com/neil_ferguson?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1243294815200124928&ref_url=https%3A%2F%2Fwww.nationalreview.com%2Fcorner%2Fcoronavirus-pandemic-neil-ferguson-did-not-walk-back-covid-19-predictions%2F

I think it would be helpful if I cleared up some confusion that has emerged in recent days. Some have interpreted my evidence to a UK parliamentary committee as indicating we have substantially revised our assessments of the potential mortality impact of COVID-19.  This is not the case. Indeed, if anything, our latest estimates suggest that the virus is slightly more transmissible than we previously thought. Our lethality estimates remain unchanged.  My evidence to Parliament referred to the deaths we assess might occur in the UK in the presence of the very intensive social distancing and other public health interventions now in place.  Without those controls, our assessment remains that the UK would see the scale of deaths reported in our study (namely, up to approximately 500 thousand).

surfinJ

  • Teahupoo Status
  • ******
  • Posts: 1769
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #73 on: March 27, 2020, 06:54:55 AM »
But it is still just a guess, huh. First teen died here today. Kids as young as 10 on respirators.

Beasho amazing work. Maybe a little over complicated though. Here is one I found on another site that I like. Easy to read and the message is clear.
^Indeed.  The media is a train wreck.   

https://youtu.be/6KS5Fl6CId4?t=4530

As all the experts have been saying from day 1, without testing, we are guessing.  Nothing against models, but they could be so much better... 

From this next clip: "All models are incorrect, some are useful"

In construction, we have a saying, "a picture is worth a thousand words, a model is worth a million." 

To steal a phrase from the Editor of Science magazine.  "We are asking science to build an aircraft while it's flying, and they don't have a full set of blueprints." 

https://youtu.be/oMSY2ZG0nwo

https://www.nytimes.com/interactive/2020/03/25/opinion/coronavirus-trump-reopen-america.html

What must it be like to be my daughter’s diabetic boyfriend, my 90yo neighbor....... to hear this line of thought?

The point, of course, is to let acquired immunity grow without overwhelming medical facilities. That's the entire idea of flattening the curve. It's not to stop the infection, we don't have any tools available to do that. It's to be able to provide care to people who need it.

It blows my mind how many people still don't get this. People need to stop thinking of it as killing "old" people. A lot of young and otherwise healthy people are getting through it better, but many with medical intervention. If it comes off the rails...   

I think there are still many unknowns in play.   

It is starting to hit my home town NY in earnest, everyone fighting a terrible battle badly supported.

Here the French military with flights and field hospitals has started moving patients around to spread the impact on the healthcare systems.

SUP Leave

  • Sunset Status
  • ****
  • Posts: 498
    • View Profile
Re: COVID-19 Data Model Updated Daily
« Reply #74 on: March 27, 2020, 10:20:20 AM »
The Imperial College study that predicted 2MM American and 500k British deaths is being revised. First by Oxford, and now by the guy who wrote it at Imperial College (Ferguson). Looks like he was too high on death rate, and too low on transmission rate. Thus creating less hospitalization than they thought.

Here are Ferguson's comments on that:

https://twitter.com/neil_ferguson?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1243294815200124928&ref_url=https%3A%2F%2Fwww.nationalreview.com%2Fcorner%2Fcoronavirus-pandemic-neil-ferguson-did-not-walk-back-covid-19-predictions%2F

I think it would be helpful if I cleared up some confusion that has emerged in recent days. Some have interpreted my evidence to a UK parliamentary committee as indicating we have substantially revised our assessments of the potential mortality impact of COVID-19.  This is not the case. Indeed, if anything, our latest estimates suggest that the virus is slightly more transmissible than we previously thought. Our lethality estimates remain unchanged.  My evidence to Parliament referred to the deaths we assess might occur in the UK in the presence of the very intensive social distancing and other public health interventions now in place.  Without those controls, our assessment remains that the UK would see the scale of deaths reported in our study (namely, up to approximately 500 thousand).

Yeah, they revised their study. The point being that 500k was the number they determined if no social distancing was taken, but revised it down to 20k based on the interventions taken. Actually, with a higher transmission rate they should have given higher than 500k (with their lethality rates) at the outset.

Dr. Birx spoke of this revision on TV last night and now 90% of America assumes she is Trump and Putin's puppet. You can see that she is trying to correlate the data they are receiving from actual testing with the models and she showed some optimism. This is not allowed.

It is pretty clear now that China's numbers are false or under reported. To my untrained eye, population density is the most important factor in cases and there is 1.4 billion chinese people.
Make paddleboarding great again!

 


SimplePortal 2.3.7 © 2008-2020, SimplePortal