The Forum for Discussion about The Third Manifesto and Related Matters

You need to log in to create posts and topics.

Johns Hopkins' Covid-19 "database"

I was looking for source data behind the Covid-19 map.

The site links to github with the following text "Downloadable database: GitHub: Here."

Curiously, the "database" is a pivot table.

A bit disappinting.

 

Quote from Riivo Kolka on March 13, 2020, 11:47 am

I was looking for source data behind the Covid-19 map.

The site links to github with the following text "Downloadable database: GitHub: Here."

Curiously, the "database" is a pivot table.

A bit disappinting.

 

Hi Riivo, heheh. You mean they should be using a tool more appropriate to the quality of the data? Like the back of an envelope. Using Excel gives a spurious sense of validity.

Both Johns Hopkins' and WHO's data are merely aggregating what they're told by individual countries. The statistics are out of date before they're even published (but it's the best we've got). There's no consistency in data collection methods nor in sampling methods nor in criteria (but it's the best we've got).

In particular, there's a couple of countries with what I'd call 'shit-hole' public health systems whose statistics are utterly incredible: the USA and Indonesia. Indonesia (Bali/Lombok) had hundreds of thousands of Chinese tourists in those islands over the Lunar New Year holiday, including tens of thousands on direct flights from Wuhan/Hubei Province. Indonesia's testing facilities (like every other semi-first-world facility in the country) are on Java; hundreds of kilometres away. What must you do with samples taken for testing Coronavirus? Keep them chilled/frozen in transit. Is that difficult in a steamy climate right on the Equator? You betcha.

If by the remotest possibility a country with the fourth largest population in the world (or Bali with ~4½ million, but much swollen by tourists) does genuinely have fewer than 100 cases, there's something miraculous going on there which the world could learn from. If OTOH there's a bunch of Australian tourists (second largest tourist country visiting Bali) who've holidayed there and brought the virus back home, that could explain how Australia (pop. 26 million) has now judged the contagion out of control (nearly 200 cases) such that they've closed sports events to the public, and indeed just cancelled both the ODI series and the T20 with NZ.

The only countries in those statistics that have decent credibility are: South Korea (testing tens of thousands per day); Japan; Taiwan; Singapore; Hong Kong (and that only after intense pressure on the CCP puppet government and protests from medical staff to close the borders with the mainland). Singapore this week got a very sober and judiciously balanced TV address from its Prime Minister. Taiwan holds a press conference every day chaired by the Minister for Health. What do these countries have in common? They were devastated by the SARS outbreak in 2003; in particular many hospital staff died or were severely ill. They put in place proper measures. They have maintained readiness ever since. Taiwan's voice gets crowded out by CCP, because Taiwan has only observer status (at best) at WHO.

Oh, did I mention USA? 3rd largest populated country. Tested ~70 people this week. Pathetic. They (Bush junior) put in place a somewhat adequate public health response organisation after SARS. The incompetent fuckwit now in the White House abolished it. USA got a TV address from the President this week. He told a pack of lies, as usual, and appears to be following none of the public health advice. Since COVID-19 particularly badly affects the elderly (over 70s), it would be entirely just if he became a statistic.

Yes there's a lot to be disappointed about with the statistics. Using a .csv/pivot table is the least to worry about.

 

Quote from Riivo Kolka on March 13, 2020, 11:47 am

I was looking for source data behind the Covid-19 map.

The site links to github with the following text "Downloadable database: GitHub: Here."

Curiously, the "database" is a pivot table.

A bit disappinting.

 

Riivo, is this you  complaining to busy statisticians with some data-purist nit-picking:

It actually stems from database theory and predicate logic.

I think we should all be grateful people are working overtime to give some transparency to a fast-moving situation. Pull your head in and stop being a distraction. You've got a more than adequate response.