w******e
发帖数: 953 | 1 My modeling work shows the total cases in New York city will be around 100
thousand and the total cases in USA will be over half million by April 12th.
My model for USA was based on data reported by March 26. My USA model
predicted that the total cases are 120,625 for March 28 and 142,094 for
March 29, respectively. The case total reported as of 6 AM today March 29
stood 124,686 in USA.
My model for New York city was based on the data reported by last Friday.
New York city had the shelter in place on March 18. But this "shelter in
place" is totally different from the shut down in Wuhan. My model showed it
took Wuhan 21 to 23 days to show the sign of slowing down after the city was
completely shut down. and it took 36 days to reach the maximum.
It is clear that it will be worsen in the coming two weeks. It
would take months to reach the maximum. | w******e
发帖数: 953 | 2 As of 3/29/2020, 66:14:06 PM, USA has 142,106 confirmed cases. | g********0
发帖数: 6201 | 3 死10w+是没跑的。（其实我看这只是计算模型给的下限，上限根本没敢提）
现在的数据已经足够流行病专家做准确计算了。 | r*********t
发帖数: 4911 | 4 when did you publish your model? cant be this week. | w******e
发帖数: 953 | 5 I worked on my model last week when I was working from home. Initially I
just want to see if the Safer at home order placed earlier last week in my
city would produce any good results. I posted my results at my social page.
Then my friend sent some USA data to me, so I extended my model to those
data. The model result indicated very bad thing is coming on its way.
The only hope is that the increasing temperature would stop the spread of
this virus.
【在 r*********t 的大作中提到】
: when did you publish your model? cant be this week.
| D********r
发帖数: 135 | 6 library(forecast)
library(ggplot2)
count<-c(109,
130,
161,
236,
335,
451,
572,
751,
1041,
1288,
1711,
2243,
2975,
3684,
4661,
6410,
9394,
14098,
19587,
26747,
35068,
44793,
55285,
69223,
86029,
104988,
124763)
fcast <- holt(count)
print(fcast)
最简单模型，今天会到142016~147060（95%置信区间）。数据来自一亩三分地，后面几
天是准的，前面几天就懒得再一个个确认了。 |
|
