South Africa November 23: Possible error in R. about covid19-viz

[shiandy]

Hi, thank you for bringing this issue to our attention. In our code we do have outlier detection implemented and the high case rate on 2021-11-23 is indeed omitted from our Rt calculations. However, the case rate in South Africa remained high in the subsequent days.

The high value of Rt for 2021-11-23 is likely because the Rt for a particular date is based on the cases in the future due to a lag between when someone is infected to when they are reported as a case. In our analysis, we assume a 7-day lag, so the Rt for 2021-11-23 is based on the cases during 2021-11-30. During that time, cases in South Africa were much higher than in the previous weeks, hence the high Rt.

[3dgiordano]

Thanks for the explanation.
It was the coincidence of the cut gave with the day of bad report.
I did not realize that the date itself is a grouping of 7 days.
Now that I understand the data format I understand what happens.
Thanks.