Tip jar

If you like CaB and wish to support it, you can use PayPal or KoFi. Thank you, and I hope you continue to enjoy the site - Neil.

Buy Me a Coffee at ko-fi.com

Support CaB

Recent

Welcome to Cook'd and Bomb'd. Please login or sign up.

March 28, 2024, 09:44:02 PM

Login with username, password and session length

Academic Pandemic: epidemiology, engineering, maths, economics, physics etc

Started by BlodwynPig, May 12, 2020, 09:29:39 AM

Previous topic - Next topic

BlodwynPig

A thread to discuss the pandemic and related topics in an academic or pseudo-academic sense, NOT PEER REVIEWED

From another thread, I wrote:

Quote from: Fambo Number Mive on May 12, 2020, 08:58:07 AM
The trade association Water UK have highlighted that Johnson was wrote to talk about coronavirus being in the water supply: https://twitter.com/WaterUK/status/1259893298518544391

Another step closer to turning into Trump.

It is in your stools though, so can get into the wastewater... however, no evidence of persistence. I'm actually working on using data from faecal samples to predict prevalence in a population, assuming lognormal distribution in the population. If Zetetic or any stats guys want to challenge the lognormal assumption, please let me know. We can then calculate the theoretical R0 through time, but really this is down to getting more data, e.g. Bahrain seems to have a shed load as all COVID patients are quarantined within a single area with a single wastewater treatment plant as the sink of their waste.



BlodwynPig


Quote from: BlodwynPig on May 12, 2020, 09:29:39 AM
...distribution in the population...

Distribution of what?  Isn't it just a big old regression:

# diseased people = a x # diseased 💩  + b ?

Heard that they do something like this for polio, although I assume you knew that.

Edit: was a bit ambiguous - I don't mean they use regression for detecting polio, I've no idea, I mean I heard that they use waste water monitoring, and presumably some sort of statistics are involved.

bgmnts

Can we somehow harness the power of wild garlic to create a vaccine? Sprinkle some into the water supply.

BlodwynPig

Quote from: DistressedArea on May 12, 2020, 10:55:25 AM
Distribution of what?  Isn't it just a big old regression:

# diseased people = a x # diseased 💩  + b ?

Heard that they do something like this for polio, although I assume you knew that.

Edit: was a bit ambiguous - I don't mean they use regression for detecting polio, I've no idea, I mean I heard that they use waste water monitoring, and presumably some sort of statistics are involved.

in your regression, what is a and how do you measure the error/uncertainty, b? Given enough data, you could probably fit a linear regression.

The distribution for patients in a Boston hospital is shown here



where the y-axis is log10(No. viruses/gram in patient faeces), x-axis is patient number (the person who created this obviously rushed the annotation).

The red line is a log normal distribution fitted to this data set (log normal distributions are generally a good rule of thumb in biology/nature). So, the number of viruses in the population is the area under the curve (NT).

If we have a good knowledge of Nmax (the maximum number of virus in a single patient) and N0 - the modal number of viruses in the patient population, and a - the spread of that distribution, assuming a lognormal distribution we can then get a relationship for NT = f(Nmax, N0, a, ET).

Now, ET is what we want to know, i..e. the prevalence in the population. Given N0 and Nmax and a come from the data and NT is calculated, we can solve numerically the function above by : Function - NT = 0, and solving for ET.

* EDIT: we may want to trim the tails as patients with less than 1 virus/g are probably not of interest

** EDIT 2: I am new to this stuff, so please pick holes as you wish - all part of the learning

Dex Sawash

Quote from: DistressedArea on May 12, 2020, 10:55:25 AM
Distribution of what?  Isn't it just a big old regression:

# diseased people = a x # diseased 💩  + b ?



Baby Musk is the key to a cure?

BlodwynPig

This is a serious thread for serious people and all I get are Clownes.

Dex Sawash

It'll turn around eventually, we have the most free time though

BlodwynPig

So, just concluded the modelling and have calculated that the City of Boston has (had) a prevalence (COVID excreters) of 4.7% of the population in early May.

Currently Massachusetts indicates a 1.13% prevalence in total, so grossly underestimated of course

Dewt

Quote from: BlodwynPig on May 12, 2020, 10:12:58 AM
Thought you would have contributed something of note, disappointed
Keep studying, as your academic career progresses you'll start reaching the level where you can understand my post.

BlodwynPig

Quote from: Dewt on May 12, 2020, 06:11:16 PM
Keep studying, as your academic career progresses you'll start reaching the level where you can understand my post.

what texts do you suggest oh fucking master cunt?

Dewt

You really seem to be taking a joke about me suggesting that just saying the number 7 is scientific analysis very seriously.


BlodwynPig

Quote from: bgmnts on May 12, 2020, 07:26:06 PM
My thesis is much bigger than yours.

Suck it.

I read it, 5000 pages of a line drawn from page to page

BlodwynPig

Quote from: Dewt on May 12, 2020, 07:24:46 PM
You really seem to be taking a joke about me suggesting that just saying the number 7 is scientific analysis very seriously.

Nice theorem. Proof please

Quote from: BlodwynPig on May 12, 2020, 01:26:21 PM
So, just concluded the modelling and have calculated that the City of Boston has (had) a prevalence (COVID excreters) of 4.7% of the population in early May.

Currently Massachusetts indicates a 1.13% prevalence in total, so grossly underestimated of course

Quite interested in this, but also bit busy. Sorry.

BlodwynPig



BlodwynPig


EOLAN

Quote from: BlodwynPig on May 12, 2020, 01:26:21 PM
So, just concluded the modelling and have calculated that the City of Boston has (had) a prevalence (COVID excreters) of 4.7% of the population in early May.

Currently Massachusetts indicates a 1.13% prevalence in total, so grossly underestimated of course

City of Boston has about 10% of population of Massachusetts and would likely have a higher infection rate. Still likely some disparity though 

BlodwynPig



BlodwynPig


evilcommiedictator

Problem is of course, someone needs to make a political decision on X thousand deaths is worth getting Y in taxation income and economic activity, that's your actuaries calculating that, and the inbred scum terrified of doing things called the Tories who get to make the call that since they fucked up and the virus is rampant, that it being slightly more rampant but allowing their favourite brothels to reopen is not a bad idea

bgmnts

Quote from: BlodwynPig on May 13, 2020, 06:49:59 AM

I like your wild garlic musings

I wonder if I could write a thesis on wild garlic, get a degree out of it.

BlodwynPig

So I'm now part of the Belgian COVID slack group and wondering if I've got in too deep - man they are fucking professional, nothing like that seen here in UK...