search

CSSEGISandData / COVID-19

Novel Coronavirus (COVID-19) Cases, provided by JHU CSSE
https://systems.jhu.edu/research/public-health/ncov/
29.13k stars 18.43k forks source link

Implied Recoveries for Luxembourg Has Error #2451

Open pabloaperezfernandez opened 4 years ago

pabloaperezfernandez commented 4 years ago

To whom it may concern:

I have been using your data through https://data.humdata.org/dataset/novel-coronavirus-2019-ncov-cases. The service is great, and I use it every day. This is why I noticed that your data for Luxembourg has errors in the time series. In particular, I calculate active infections = cases - deaths - recoveries. Your data shows a massive drop it the number of active cases on April 24th. It causes a very suspect, one-day drop in the number of active infections. A quick look at sites from the Luxembourgish government show that active infections by the end of April are correct in your data, but the curve getting there has issues.

Hope this helps. I appreciate the work you have done and hope this contributes to the quality of the data.

Sincerely,

Pablo

pabloaperezfernandez commented 4 years ago

Incidentally, the Netherlands data also has problems. Implied recoveries declines over time. This is not possible. Recovies should be an increasing series, just like confirmed cases.