Corona Virus and Exponential Growth

@Metal-Brain
Wow, I had figured that out all by myself! But you will never stop projecting conspiracy theories. What will you latch onto after we get through THIS crisis?
Climate change hoax? Moon landing hoax? Einstein was a fraud?

@sonhouse said
@Eladar
As of today the infections in US are 12,000 and 274 deaths.

I have been keeping track of the deaths. I will check again in about 20 minutes.

@eladar said
@joe-shmo

I do not like the total cases because the US is doing a terrible job testing.

Deaths are more robust to testing because we do test sick people, especially those very sick.

So I would suggest what you are saying but with total deaths. It is a bit more morbid, but a bit more accurate.

The Ti 84 graphing calculator has a logistics regression option in the stat section. You go to the math column then down to option B.

I see what you mean about death being the better measure. I agree.

Thanks for the pointing out the Logistics Regression. I have an 89, didn't know it had the function. I used the Data Matrix App with the following data. This was the result: Starting at March 4 as Day 10 for the regression.

https://www.worldometers.info/coronavirus/country/us/

(Day, Deaths)
10 11
12 15
14 22
16 30
18 41
19 49
20 57
21 68
22 86
23 109
24 150
25 207
26 256


L(x) = a/(1+b*e^(c*x) ) + d

where
a = 2347.947362
b = 30782.831736
c = -0.314662
d = 14.58998

If it continues accordingly the inflection point is around day 33 with 1900 Deaths, which means next week by this time we should have a very good idea if that's the case. The model is predicting less than 2500 US Deaths. I don't know what stock I'll hold in the model ( its a nice fit for the data we currently have), but perhaps the measures we are taking are working?

Today's total is 324 which means 66 new deaths. I think today would be day 27 for your logistic curve.

Your equation predicts a max number of deaths 'a'. In this case 2348.

There is a warning from Stats concerning models. Models are only accurate for observations within the domain of the original data.

@Eladar

Yeah, that's correct day 27. I'll wait till the worldometer puts a few new data points the graph before recalculating. One thing I've noticed is if they loosen the quarantining, the experts ( according to your cited paper ) expect the curve to kind of reset. Which means the gift that keeps on giving...death. They are talking months of restricted behavior. I think the "two weak" first step restrictions proposed by our governor ( PA ) is just the lubrication for a very painful freedom sucking experience yet to come ... wtf.

@joe-shmo

The shutting down of society is simply slowing down the initial growth. The longer they can put it off the fewer deaths.

If the death rate maxes out in June, then that would be about 13 weeks. If they slow the spread for 2 weeks we will only be looking at 11 weeks of exponential growth.

Seems to me this is their plan, slow it down during the flu season. They may extend this shut down, we will see.

Given how the AP tests have been made a shortened take home test, it seems we will be shut down for a while.

@sonhouse said
@Metal-Brain
Wow, I had figured that out all by myself! But you will never stop projecting conspiracy theories. What will you latch onto after we get through THIS crisis?
Climate change hoax? Moon landing hoax? Einstein was a fraud?

https://www.sciencemag.org/news/2020/01/wuhan-seafood-market-may-not-be-source-novel-virus-spreading-globally

From the science article below:

"Using these differences and presumed mutation rates, several groups have calculated that the virus began to spread around mid-November 2019—which supports the thesis that spread may have occurred before any of the cases linked to the market. One group put the origin of the outbreak as early as 18 September 2019."

https://science.sciencemag.org/content/367/6477/492.full

What conspiracy theory? It is your projection.

@joe-shmo said
I see what you mean about death being the better measure. I agree.

Thanks for the pointing out the Logistics Regression. I have an 89, didn't know it had the function. I used the Data Matrix App with the following data. This was the result: Starting at March 4 as Day 10 for the regression.

https://www.worldometers.info/coronavirus/country/us/

(Day, Deaths)
10 ...[text shortened]... its a nice fit for the data we currently have), but perhaps the measures we are taking are working?

Can your tool give error bars?

@metal-brain said
https://www.sciencemag.org/news/2020/01/wuhan-seafood-market-may-not-be-source-novel-virus-spreading-globally

From the science article below:

"Using these differences and presumed mutation rates, several groups have calculated that the virus began to spread around mid-November 2019—which supports the thesis that spread may have occurred before any of the cases linke ...[text shortened]... //science.sciencemag.org/content/367/6477/492.full

What conspiracy theory? It is your projection.

The evidence is that the seafood market was the most likely point at which the first humans were infected. I haven't looked at the links you've posted, but speculation based around mutation rates isn't evidence. It's possible these groups are right and the actual source was elsewhere - but none of this constitutes evidence that the virus is artificial.

@DeepThought

No, the simpler regressions give R^2, the logistics regression does not. Neither have error bars, and I see no options for them...

@deepthought said
The evidence is that the seafood market was the most likely point at which the first humans were infected. I haven't looked at the links you've posted, but speculation based around mutation rates isn't evidence. It's possible these groups are right and the actual source was elsewhere - but none of this constitutes evidence that the virus is artificial.

There is no evidence that it started in Wuhan.

https://www.sciencemag.org/news/2020/01/wuhan-seafood-market-may-not-be-source-novel-virus-spreading-globally

I never claimed I had proof that C19 is artificial, but Russian Biologist and Former U.N. Expert Igor Nikulin is not someone I would ignore. I know you will just dismiss it as Russian propaganda, but China's spokesman for the foreign ministry also made the allegation in a tweet and of course Ayatollah Ali Khamenei is repeating the same allegation.

https://www.memri.org/tv/russian-biologist-igor-nikulin-coronavirus-conspiracy-global-government-decimate-population

If an artificial virus escaped from a bioweapons lab would my government admit it? The US has over 400 of them around the world. Here is another Russian allegation.

https://www.rt.com/news/440309-us-georgia-toxic-bioweapon-test/

I guess my numbers and that site's numbers are different.

41,57,49 and 46 are the official last 4 day death counts. They have 20 fewer deaths yesterday.

@joe-shmo said
@DeepThought

No, the simpler regressions give R^2, the logistics regression does not. Neither have error bars, and I see no options for them...

As a hack approach what you could do is divide your data into two subsets, the subset of points above the curve for the parameters you've found and one set for below the curve. Then work out the parameters for each of those two subsets. This will give you a handle on the size of the error. If it were a 1 parameter fit then this would work fine, but you have 4 parameters so it's not great. I'll have a think about it and see if I can work out a better method.

You can always calculate r^2, Khan academy has a video on it.