search

micb25 / RKI_COVID19_DATA

Archiv von Datenbank-Dumps des RKI-COVID19-Dashboards mit täglicher automatisierter Aktualisierung.
https://corona.rki.de
13 stars 2 forks source link

Data of 16/Jan/2021 and 17/Jan/2021 appears the same #1

Closed rogerlord closed 3 years ago

rogerlord commented 3 years ago

Hi there - first of all, great job you are doing in storing the data from RKI!

I am using your stored data as historical data in my repo to nowcast the number of infections. Recently the script had some issues and I traced it back to the data of the 17th of Jan 2021, it appears to be the same as that of the 16th of Jan. The data of the 17th in this [repository[(https://github.com/CharlesStr/CSV-Dateien-mit-Covid-19-Infektionen-/tree/master/Dezember) however appears different.

Could you address this issue?

micb25 commented 3 years ago

Hey, thank you very much for spotting this.

According to newspapers at this specific day (2021-01-17) there was indeed a problem with the official data from the RKI. Consequently, the official RKI dashboard also didn't work. The scraper only runs once a day in the morning and the problem seemed to be fixed later that day which explains why the database dump from @CharlesStr is different and the correct one.

I have updated the specific file using the dump from CharlesStr/CSV-Dateien-mit-Covid-19-Infektionen-.

rogerlord commented 3 years ago

Many thanks for the speedy reply, results look much better now, see here. There is still something odd with the data on that day which causes a huge spike in the nowcast, but I'll investigate that separately.