search

CSSEGISandData / COVID-19

Novel Coronavirus (COVID-19) Cases, provided by JHU CSSE
https://systems.jhu.edu/research/public-health/ncov/
29.12k stars 18.39k forks source link

Ongoing Problems with current US Time Series #2318

Open DinsdalePiranha85 opened 4 years ago

DinsdalePiranha85 commented 4 years ago

A few examples:

_deaths_US.csv: Looking at USA / NY / Suffolk as example 693 Deaths logged for 4/16, 4/17, 4/18 and 4/19? No deaths in 3 days in the 3rd worst hit county in the worst hit state? Not Likely.

_confirmed_US.csv: Again, looking at USA / NY / Suffolk as example One Case between 4/10 and 4/11? Logged 20320 and 20321 respectively. Again, NOT likely.

paulavery1951 commented 4 years ago

Hi, the NYT series has very similar numbers for those dates. Below are the cumulative and daily deaths from that time series. Note that 4/19 seems to be playing catch-up for the previous 2 days with no deaths. The JHU series does catch-up for 3 days on 4/20.

New York_Suffolk,2020/04/15,663,46 New York_Suffolk,2020/04/16,706,43 New York_Suffolk,2020/04/17,706,0 New York_Suffolk,2020/04/18,706,0 New York_Suffolk,2020/04/19,845,139 New York_Suffolk,2020/04/20,887,42 New York_Suffolk,2020/04/21,918,31