coronavirus

[GUBRATS]

I'm so glad I haven't a fcuking clue what the hell you are all talking about ?

And on that note , I'm nipping in the ' pub ' before going to bed 

Nighty night

[SSoutherner]

13 minutes ago, ckn said:

I’m tired. I spent far too long thinking “what’s politically correct about that? What did I miss?”

nothing just showing what simple log graph paper looks like to better explain why an exponential growth is a flat line on log paper

[Wholly Trinity]

13 minutes ago, Bleep1673 said:

So invertabally, what is the exponential decrease of the demographically diverse spread of the demise of the European population?

Presactly. 

Which is the kind of question the press should be asking the government.

[Wholly Trinity]

47 minutes ago, SSoutherner said:

In pre PC days we used special graph paper like this for it

Brings back fond memories of being a student in the 80s 

Not only trying to complete a lab experiment correctly, but then plotting the results on log paper WITH A STONKING HANGOVER 

mistakes were made

[Kayakman]

54 minutes ago, Wholly Trinity said:

some more interesting maths to highlight when the exponential growth battle is being won. (Logarithmic scales again, sorry)

 

 

Reminds me of this one.

 

[Kayakman]

4 hours ago, Bob8 said:

Firstly, viruses and bacteria are very different. We are more closely related to E.  coli than this coronavirus is. So, it is not really relevant.

Secondly, clearly anything is possible.  As I did make perfectly clear.  It is a novel virus, but we can consider likelihood. 

But our bodies  reject both of them and that is a 'constant' which leads to relevance as the immune system is triggered.

[Bob8]

5 hours ago, Kayakman said:

But our bodies reject both of them and that is a 'constant' which leads to relevance as the immune system is triggered.

Yes, but I really would not want you to be concerned with this. The E. coli bacteria are found throughout your gut, it is a very particular strain that the immune system has to react against, and as the bacterium is extraordinarily complex compared to a virion, the specific recognition is very different.

A more likely concern is the proteins on the outside of this coronavirus are very variable, like flu. Both are RNA based and so prone to mutation. But they are not related and there is no strong reason to think that is the case beyond speculation.

Again, this is rough knowledge rather than expert insight into this virus.

[Wiltshire Rhino]

 

[Dave T]

8 hours ago, Les Tonks Sidestep said:

Dave if you look at CKN's graph you can see that Italy's and Spain's curves are flattening, or curving over - a sign that things are slowing down. Looking at the 1st graph I've put up that slow down is really difficult to see, which is why the log graphs are (or can be) more useful for following things that are growing exponentially. It's a real positive that they are - as soon as the curve started to flatten the increase on day on day deaths was beginning to fall and eventually actual numbers do, as they appear to be now. 

Ah OK, I suppose it is a certain kind of chart that I am struggling to get my head around, as it is all about the curve, whereas I am thinking about it in real numbers, scale of the death rate and how comparable the numbers are to other countries. Whereas this graph is really all about the curve. 

I think this is too specialist for the mainstream, I heard a horrendous explanation of it on one of the news channels yesterday, explaining that China had now hit the line that represented a doubling every 3 days, despite the fact their line was not following the same trajectory as the 3 day line, it had flattened and was crossing through it. 

Why isn't a traditional line chart useful for these numbers (i.e. The one that has been used to demonstrate the sombrero)?

Thanks for bearing with me by the way!!!! 

[JohnM]

The situation in care homes is awful. Not pleasant reading at all. 

BBC News - Coronavirus: The grim crisis in Europe's care homes
https://www.bbc.co.uk/news/world-europe-52094491

 

[ckn]

3 minutes ago, JohnM said:

The situation in care homes is awful. Not pleasant reading at all. 

BBC News - Coronavirus: The grim crisis in Europe's care homes
https://www.bbc.co.uk/news/world-europe-52094491

 

There’s genuine terror that that’ll happen here. Minimum wage staff, zero protection, owners who can’t afford regular deep cleans, a system that penalises them for calling ambulances and a support in GPs/community care that’s startling to buckle under other strains. 

Anyone who visits one of these places unnecessarily now really needs a good flogging. 

[Les Tonks Sidestep]

38 minutes ago, Dave T said:

Ah OK, I suppose it is a certain kind of chart that I am struggling to get my head around, as it is all about the curve, whereas I am thinking about it in real numbers, scale of the death rate and how comparable the numbers are to other countries. Whereas this graph is really all about the curve. 

I think this is too specialist for the mainstream, I heard a horrendous explanation of it on one of the news channels yesterday, explaining that China had now hit the line that represented a doubling every 3 days, despite the fact their line was not following the same trajectory as the 3 day line, it had flattened and was crossing through it. 

Why isn't a traditional line chart useful for these numbers (i.e. The one that has been used to demonstrate the sombrero)?

Thanks for bearing with me by the way!!!! 

On a straightforward (cumulative) line chart it's difficult to see when things begin to slow down, particularly around the tipping point. When zero growth is reached the curve will be 'S' shaped and the point of change is more obvious (unfortunately we're still a few weeks away from that). With a sombrero graph (daily death versus day) the peak daily increase is easy to spot (and to some extent rate of daily increase). For some reason they don't seem to be being used widely - possibly because those collecting data are academics? Here's a link to data for Italy which has a mixture of linear, log and sombreros. Other country's data can also be seen in the same formats https://www.worldometers.info/coronavirus/country/italy/ 

[Bob8]

20 minutes ago, Celt said:

19% now Smudger.

I assume you are just about to start "battening down the hatches"?

Indeed.

In a true serious pandemic though, Smudger would be vital with his HGV licence and keeping the supply chain going.

[Bob8]

43 minutes ago, Dave T said:

Ah OK, I suppose it is a certain kind of chart that I am struggling to get my head around, as it is all about the curve, whereas I am thinking about it in real numbers, scale of the death rate and how comparable the numbers are to other countries. Whereas this graph is really all about the curve. 

I think this is too specialist for the mainstream, I heard a horrendous explanation of it on one of the news channels yesterday, explaining that China had now hit the line that represented a doubling every 3 days, despite the fact their line was not following the same trajectory as the 3 day line, it had flattened and was crossing through it. 

Why isn't a traditional line chart useful for these numbers (i.e. The one that has been used to demonstrate the sombrero)?

Thanks for bearing with me by the way!!!! 

In terms of understanding, there has been an invisible plus side.

As it has been imported from ski-ing holidays to the Alps and international business people, it is something of a rich man's flu. This does seem to mean there is no stigma attached. Had it been imported by poor immigrants, I imagine there would be nastier BS going around.

[SSoutherner]

22 minutes ago, Bob8 said:

In terms of understanding, there has been an invisible plus side.

As it has been imported from ski-ing holidays to the Alps and international business people, it is something of a rich man's flu. This does seem to mean there is no stigma attached. Had it been imported by poor immigrants, I imagine there would be nastier BS going around.

This was brought home to me when i thought "i bet the channel islands are safe" checked and saw that they had cases due to skiing holiday returnees

[MattSantos]

Is there a tentative time-line for you lot?

In Singapore, the measures are in play until end of April - i don't see this being extended, but who knows..

Suspect that if other countries are in lock down etc, whilst we may loosen up the restrictions, we wouldn't be opening the flights up until everyone is out the other side.

 

 

[SSoutherner]

58 minutes ago, ckn said:

There’s genuine terror that that’ll happen here. Minimum wage staff, zero protection, owners who can’t afford regular deep cleans, a system that penalises them for calling ambulances and a support in GPs/community care that’s startling to buckle under other strains. 

Anyone who visits one of these places unnecessarily now really needs a good flogging. 

But it does re raise the comments from yesterday re which deaths are counted (at least in France)

More than a third of care homes in the Paris region alone are thought to have been affected by coronavirus. But while hospital deaths are recorded in the daily toll of coronavirus, deaths in care homes are not.

Alexandre Sanner works as a doctor for another care home, in Vosges, in the east of France, where at least 20 people have died.

[Dave T]

44 minutes ago, Les Tonks Sidestep said:

On a straightforward (cumulative) line chart it's difficult to see when things begin to slow down, particularly around the tipping point. When zero growth is reached the curve will be 'S' shaped and the point of change is more obvious (unfortunately we're still a few weeks away from that). With a sombrero graph (daily death versus day) the peak daily increase is easy to spot (and to some extent rate of daily increase). For some reason they don't seem to be being used widely - possibly because those collecting data are academics? Here's a link to data for Italy which has a mixture of linear, log and sombreros. Other country's data can also be seen in the same formats https://www.worldometers.info/coronavirus/country/italy/ 

Last one from me mate (promise ).

The issue I have with this is that a curve is being artificially manufactured isnt it? For example, using the US' numbers, the numbers are still rising, but naturally as the number gets bigger the % increase gets smaller, so this is represented as a flattening of the curve on the logarithmic scale, when the number of deaths is still large and increasing daily, as shown by the linear graph. But because the scale now being used is 100k-1m the line is shallower than the 0-100 section of the graph. 

I understand the reasons, but I'm surprised that these are the main graphs being used as they re a bit specialist imo.  Showing a 50% increase on 1000 daily deaths as shallower than a 75% increase on 30 deaths is unhelpful to the mainstream imo. 

Thanks to your posts (and others) I understand the benefit of the graph, but as the only graphs really being used on the news I find it inappropriate tbh. Particularly as the commentators are often not following them. 

[Bob8]

12 minutes ago, Dave T said:

Last one from me mate (promise ).

The issue I have with this is that a curve is being artificially manufactured isnt it? For example, using the US' numbers, the numbers are still rising, but naturally as the number gets bigger the % increase gets smaller, so this is represented as a flattening of the curve on the logarithmic scale, when the number of deaths is still large and increasing daily, as shown by the linear graph. But because the scale now being used is 100k-1m the line is shallower than the 0-100 section of the graph. 

I understand the reasons, but I'm surprised that these are the main graphs being used as they re a bit specialist imo.  Showing a 50% increase on 1000 daily deaths as shallower than a 75% increase on 30 deaths is unhelpful to the mainstream imo. 

Thanks to your posts (and others) I understand the benefit of the graph, but as the only graphs really being used on the news I find it inappropriate tbh. Particularly as the commentators are often not following them. 

If you are looking at reducing the infection rate, logarithmic makes sense. But, people die and ICUs are overwhelmed in absolute terms.  People will mistake the infection rate going down for the number of people being infected going down. 

[Dave T]

1 minute ago, Bob8 said:

If you are looking at reducing the infection rate, logarithmic makes sense. But, people die and ICUs are overwhelmed in absolute terms.  People will mistake the infection rate going down for the number of people being infected going down. 

Yes, I think this is why I'm struggling with it. As a suite of data I can see the benefit. As the only graph being used, I think it falls short. 

[gingerjon]

29 minutes ago, SSoutherner said:

But it does re raise the comments from yesterday re which deaths are counted (at least in France)

More than a third of care homes in the Paris region alone are thought to have been affected by coronavirus. But while hospital deaths are recorded in the daily toll of coronavirus, deaths in care homes are not.

Alexandre Sanner works as a doctor for another care home, in Vosges, in the east of France, where at least 20 people have died.

I don't know how they're counting in Spain but there were several reported care homes with close to 100% fatalities because staff were sick and couldn't come in, and thus residents were ill and no one was there to care for them.

A vision of horror.

[Les Tonks Sidestep]

9 minutes ago, Dave T said:

Yes, I think this is why I'm struggling with it. As a suite of data I can see the benefit. As the only graph being used, I think it falls short. 

Most of the data being published in the mainstream news comes from academic sources so probably not surprisingly it's as log plots. It takes a bit more digging around to find the other types. There's also a danger of data overload (particularly as you should be looking for different things in different types of plot) if more than one type is shown. And finally Bob's comment about avoiding confusing 'rate reduction' with '(current) number reduction' is really important at the moment.

[Saintslass]

1 hour ago, SSoutherner said:

But it does re raise the comments from yesterday re which deaths are counted (at least in France)

More than a third of care homes in the Paris region alone are thought to have been affected by coronavirus. But while hospital deaths are recorded in the daily toll of coronavirus, deaths in care homes are not.

Alexandre Sanner works as a doctor for another care home, in Vosges, in the east of France, where at least 20 people have died.

We have been doing likewise but the ONS is now producing weekly figures which includes deaths in the community where Covid features on the death certificate (not necessarily as the cause of death itself but mentioned with other disease which featured at the same time).  The first of these figures have been produced and that increases our total of deaths by 40.  However, that is a catch up figure going back to when our first death was recorded so any future increases highlighted by the ONS figures may not be as great (or they may be greater of course as the critical phase of the spread comes into play).

[SSoutherner]

17 minutes ago, Les Tonks Sidestep said:

Most of the data being published in the mainstream news comes from academic sources so probably not surprisingly it's as log plots. It takes a bit more digging around to find the other types. There's also a danger of data overload (particularly as you should be looking for different things in different types of plot) if more than one type is shown. And finally Bob's comment about avoiding confusing 'rate reduction' with '(current) number reduction' is really important at the moment.

Yes if you have 1000 cases which is 150 up but yesterday was 200 up that is better than 1000 cases which is 150 up but yesterday was 100 up and both are better than 500 cases which is 150 up with 100 up day before

[gingerjon]

I'm getting confused by you lot messing around with numbers. Definitely had enough of experts.

But, to take it back to the personal, some vaguely good news in the Ginger household as, after far, far, far too many phonecalls and dead ends, Mrs Ginger was finally added to the priority list by Sainsbury's and now we have a mahoosive delivery coming on Thursday. We could even have got one for tomorrow.