OT: Corona virus

[Ian R]

ORIGINAL: BBfanboy

ORIGINAL: Ian R

ORIGINAL: BBfanboy



On the BBC news ( a very trustworthy source) a few minutes ago they said Australia had just reported a rise of over 1000 cases, its biggest one-day increase. Doesn't sound like "poof" to me. Perhaps different regions of Australia are having different experiences. Melbourne is much more temperate than Darwin, for example.

The latest numbers on the Department of health website are:

"As at 6.30am on 21 March 2020, there have been 874 confirmed cases of COVID-19 in Australia. There have been 165 new cases since 6.30am yesterday. "

It is now 1057 hrs on the 22nd here, so I'll watch for the next update and confirm (or not) that report.

Looks like BBC got the total cases instead of the daily increase. Maybe their research people are working from home ...
Thanks for the info!

Here is the update:

"As at 6.30am on 22 March 2020, there have been 1,098 confirmed cases of COVID-19 in Australia. There have been 224 new cases since 6.30am yesterday."

[alanschu]

I found a different site that also had numbers similar to the WHO, but find its daily numbers jump around a bit (WHO had an issue with that too). Though Raw numbers aren't really what I'm looking at for this particular analysis.

I still used Worldometers charts, which don't include today's values yet.

Looking at China's death rate, there is a bit more than 3 weeks of steady increases to the daily deaths. Jan 22 - Feb 12th (21 days). Then it plateaued for about another 11 days (Feb 12 to Feb 23). After that it was a quick decline, and then a steady decline for the the tail.


Italy's is about 2 weeks in on their increasing numbers. If it follows China, it could still be about another week of increasing daily totals and might continue until March 28. If it plateaus the same way, it'll remain around that daily count until 11 days or so, which would be April 8th. I didn't pick up the first few days of Italy's deaths as they didn't seem to match a linear trend line. It's possible that Italy is a couple days closer in any case (which would obviously be ideal).


I'm not sure I agree with that one medium post's that this isn't particularly contagious though. There's conflating variables for the spike in Italy's deaths, and I am more inclined to believe "this got around and our health care system couldn't keep up" as being a bigger factor than "the quarantine started on March 6th, and that's when the infection/death counts started to go up."

Based on how many more deaths Italy has than China, with Hubei and Italy both having similarish populations, I wouldn't be surprised if the infection rate of Italy is quite a bit higher and testing measures is difficult to ascertain.


I really hope no other countries have issues with health care capacity.

[alanschu]

ORIGINAL: RangerJoe

Just remember, that China is a little more densely populated than Alberta. If you need some fresh air, just go visit Lake Louise.

Of course. Was just talking about what the messaging was like. I live in Edmonton and there were a few cases (including one in my office tower). But I think most of us are "we don't want Italy" and there's a hefty dose of caution.

I'm not particularly worried and my wife and I can go for a walk from our condo. But my wife is still working and her store was super busy (she works at a Home Depot and people were stocking up on home improvement projects). Today was the first day I had heard where they were actually mitigating the amount of people allowed into stores which sounds a bit more in line with what Singapore is asking.


As always, if the numbers remain low at what point is it "this was a waste of time and money and did more harm than good" versus "it's actually good that we did this."

[RangerJoe]

Just say "I don't know if it worked or not, but we did not get hit as hard as we could have." Say that if people complain that it was not needed, just like suspenders and a belt.

[pontiouspilot]

I'm in Alberta. Ontario is one of the worst in terms of tests/100k. Here one test being done/185 Albertans, with positive ratio of 1/105 tests. (as per Calgary Herald and CBC).

Manitoba looks to be doing well so far...fingers crossed.

[Kull]

Finally getting some granularity out of Spain (see attached graphs showing current cases per province and increases over time. The number of tests reported on the wiki spiked from 30K to 355K based on this article. Probably a mixture of previous under-reporting and dubious numbers. How they went from nowhere to "more tests than any country in the world" in just two days stretches credulity (round numbers are usually a tipoff that somebody is guessing). Especially given the healthcare system meltdown currently in-process. I don't buy it.

Edit: To clarify, the cases counts, deaths, and province numbers look legit.

[obvert]

ORIGINAL: pbiggar

I have been lurking in he forum for some time, but I do feel a need to respond to this post. I read the article a couple of times to follow the data that led to conclusion that "The data is overwhelming at this point that community-based spread and airborne transmission is not a threat" and we should all keep going to work.

Included in the article on many graphics pulled from reputable sources, but when you add them up I do not follow the authors conclusion.

I have not studied virology and public health, and so I rely on those who have. This conclusion in the article is opposite from everything I have read from those who specialize in this area. How can that be?

So on the one hand I have articles like this one https://spiral.imperial.ac.uk:8443/bits ... 3-2020.pdf
that convinced your President that the risk from COVID-19 isreal and that the situation was not under control in the US.

And this article, written by a guy with BA in Economics from Texas Christian University, whose qualifications are that he is viral marketer and a member of the California Republican Executive Committee.

The graphs are interesting, but I do not see how he has connected them to reach the conclusions that are counter to what our public health experts are telling us.

When you read the article, do the conclusions follow the data in your mind? <this is a serious question>

ORIGINAL: Canoerebel

Early today, Ranger posted this link: https://medium.com/six-four-six-nine/ev ... 767def5894

I've just started reading through it. Again, the odd experience of reading thoughts previously articulated in here at length. As though this Forum really is a remarkable set.

Here's one chart that caught my eye:

Thanks for this post. We have to be careful to check sources, look at how data is presented carefully, and of course our government experts (everywhere) have not followed this advice and kept everything open.

I teach data visualisation in my Graphics Design course and this particular graph would be one I'd flag for a misleading representation of data. One, as previously mentioned, the initial high percentage points are inevitable as new cases will be doubling when it's 10 to 20 and 40 to 80, just being discovered as the epidemic begins in a new place. Even if percentage growth slows, the numbers are huge by the time you get to day 10-15. Being at 40k cases and adding 20% is huge. I would not be comforted at all by an increase of 8,000 cases in a day, from 40,000 to 48,000. Especially knowing those were only more severe cases actually being tested in hospitals, as is the case in most European countries.

[alanschu]

Here's an interesting chart of deaths on a logarithmic scale. Italy has slightly flattened, although it seems to be in a bit of a slower, linear state right now. Hopefully in the next few days.

France seems to be overlapping Italy thus far. I'm concerned about Spain and the UK though. Belgium is off to an aggressive start, although they started their lockdown before anyone had actually died so hopefully those measures see it flatten quickly.

US has sped up a bit in the past 3 days which may be a cause for concern.


I also found this link that explores the largely the period it takes to double the death count. Runaway speed tends to be 3 days. (also extra explantions about the exponential data bias

https://ourworldindata.org/coronavirus

Interesting thing they did is they actually *stopped* relying on the WHO numbers. Sounds like they changed the report cutoff time between situation reports 57 and 58, and while verifying the situation reports they found a bunch of errors.

[Ian R]

Well that graph explains why Australia has just gone into lockdown.

I guess the pollies have to work from the worst case scenario.

I don't know how it works in your country, but here they are so dumb they have no chance of understanding what that graph actually means.

[Zorch]

'COVID-19 coronavirus epidemic has a natural origin' https://www.sciencedaily.com/releases/2 ... 175442.htm

'An analysis of public genome sequence data from SARS-CoV-2 and related viruses found no evidence that the virus was made in a laboratory or otherwise engineered.'

'The novel SARS-CoV-2 coronavirus that emerged in the city of Wuhan, China, last year and has since caused a large scale COVID-19 epidemic and spread to more than 70 other countries is the product of natural evolution, according to findings published today in the journal Nature Medicine.'

[MakeeLearn]

Just checking in, haven't read any posts since yesterday.

Early yesterday morning I stopped by the grocery store to get some things for lunch. Empty shelves here and there. The man in front of me at checkout had a buggy that was overflowing, with multiples of some items.

I said to him "Late start?" He said "Yea."

In my head I'am thinking "YEARS LATE"

Heard that CV19 cases are in towns within 50miles of here.

I've been given a "Permission to Travel" paper, in case of a mandatory stay in place.

I feel like a 6 year old that for years has been asking "Are we there yet?....Are we there yet?" and now we are pulling in to the driveway.

Back to the trenches.

[Canoerebel]

Bloomberg News prepared this graph.

It wasn't offered to suggest that the pandemic is over. It was offered to show that the rate of increase is decreasing, day be day, which is true, and to show that the various countries are experiencing roughly similar "trajectories."

For what limited purpose it was used, I think that's fine.

Or am I missing something?

ORIGINAL: obvert

ORIGINAL: pbiggar

I have been lurking in he forum for some time, but I do feel a need to respond to this post. I read the article a couple of times to follow the data that led to conclusion that "The data is overwhelming at this point that community-based spread and airborne transmission is not a threat" and we should all keep going to work.

Included in the article on many graphics pulled from reputable sources, but when you add them up I do not follow the authors conclusion.

I have not studied virology and public health, and so I rely on those who have. This conclusion in the article is opposite from everything I have read from those who specialize in this area. How can that be?

So on the one hand I have articles like this one https://spiral.imperial.ac.uk:8443/bits ... 3-2020.pdf
that convinced your President that the risk from COVID-19 isreal and that the situation was not under control in the US.

And this article, written by a guy with BA in Economics from Texas Christian University, whose qualifications are that he is viral marketer and a member of the California Republican Executive Committee.

The graphs are interesting, but I do not see how he has connected them to reach the conclusions that are counter to what our public health experts are telling us.

When you read the article, do the conclusions follow the data in your mind? <this is a serious question>

ORIGINAL: Canoerebel

Early today, Ranger posted this link: https://medium.com/six-four-six-nine/ev ... 767def5894

I've just started reading through it. Again, the odd experience of reading thoughts previously articulated in here at length. As though this Forum really is a remarkable set.

Here's one chart that caught my eye:

I find it stra

Thanks for this post. We have to be careful to check sources, look at how data is presented carefully, and of course our government experts (everywhere) have not followed this advice and kept everything open.

I teach data visualisation in my Graphics Design course and this particular graph would be one I'd flag for a misleading representation of data. One, as previously mentioned, the initial high percentage points are inevitable as new cases will be doubling when it's 10 to 20 and 40 to 80, just being discovered as the epidemic begins in a new place. Even if percentage growth slows, the numbers are huge by the time you get to day 10-15. Being at 40k cases and adding 20% is huge. I would not be comforted at all by an increase of 8,000 cases in a day, from 40,000 to 48,000. Especially knowing those were only more severe cases actually being tested in hospitals, as is the case in most European countries.

[obvert]

Question: where is the daily info graph Kull and others have posted that shows daily testing trends? Thx

[Canoerebel]

Daily death graph, USA.

This would be encouraging if the trend were set. But there are always daily perturbations in the rising leg of the bell-shaped-curve, as Italy showed this week.

The graph of the number of daily cases in the US is similar.

[obvert]

The Financial Times is doing a good trending graph that has three important data points represented in relation to all countries that have had at least 100 cases of the virus.

The case trends are here shown against the days since 100 cases and the total number of cases. Keeps some in perspective and shows the relative position of each country as their version of this pandemic continues.

It's easy to see what you want to discover and doesn't need any interpretation, really.

https://www.ft.com/coronavirus-latest

[obvert]

I also found these on testing which show the US racing up the chart to the top five, which is fantastic. Glad it's coming through more now. [&o]

This site is great. Lots more there.

https://ourworldindata.org/coronavirus

[Canoerebel]

Wait a minute, Erik. The graph two posts above does the same thing that the one you found fault does - shows trends using numbers.

This one suffers from at least one flaw, if read out of context or isolated. It uses raw numbers instead of cases per capita. Raw numbers don't mean much (and are further subject to serious uncertainty due to disparities in testing percentages).

Why is it reliable and the other not???

P.S. The graph is fine but users need to understand its limitations, context and biases, just as with the other one.

[Canoerebel]

Number of serious cases reported had been oddly low and consistent day after day in the US. Today it jumped to 700+. Probably something wonky in the reporting.

Germany continues with comparatively good numbers.

[mind_messing]

ORIGINAL: Canoerebel

Number of serious cases reported had been oddly low and consistent day after day in the US. Today it jumped to 700+. Probably something wonky in the reporting.

Germany continues with comparatively good numbers.

Testing in the US ramping up. Likely a lot of serious cases where COVID was suspected but not confirmed. I think that number is going to rise a fair bit for the foreseeable.

[Canoerebel]

It can't be that. That number almost certainly applies to just hospitalized cases. It was very weirdly low and stable for a week or so. Then a sudden jump today. Rational trends can't explain it. It's human error or neglect almost certainly, either in how the USA compiled or reported or by the table creator in updating.