search

daenuprobst / covid19-cases-switzerland

Data from BAG Tweets made useful.
285 stars 74 forks source link

Geneva data seems wrong #73

Open gvb1234 opened 4 years ago

gvb1234 commented 4 years ago

Geneva data at

https://www.ge.ch/document/covid-19-situation-epidemiologique-geneve

says 3993 for 07.04, up 131 from 06.04 (I guess), but yesterday it was announcing 3562 for 06.04, which doesn't add up.

Any idea what's going on?

mcamenisch commented 4 years ago

Geneva has a special systeme of counting. On day x they publish x and x-1 provisional state on 12am. x and x-1 are provisional. So the next day the add cases on x-1 and x-2 when they publish x. I never found a cleaned timeserie. Anyone knows if they publish something like that? But issu is broader. Geneva reference to the time when the test was done other cantons refer to time when they have the result.

gvb1234 commented 4 years ago

I guess every canton doing things it's own way is one of the charms of Switzerland.

Of all the timeseries available, probably the number of ICU patients is the cleanest, as it is not subject to under-testing, under/mis-reporting (like deaths), and timestamp issue like the one you mentioned.

gvb1234 commented 4 years ago

I've been tracking the GE data here:

https://www.dropbox.com/s/9navztizaw4tdze/GE_corona_data.xlsx

I manually scrape the daily number of new cases from the pdf at www.ge.ch.

It is insane how much that data is revised. Between the data published on 04/14 and 04/15, they have revised about half the dates from as early as 03/13.

madturtles commented 4 years ago

It seems that the Geneva count hasn't been updated for a while, although www.ge.ch is currently providing official information describing the situation on 06/29. Is there anything someone could do to catch up with the other cantons?