We have generated a number of estimates for the number of cases.
Before generating estimates for future cases and the total number of cases we first have to find the midpoint for the actual number of cases. Otherwise we will not know when to “flip” our (cumulative) estimates generated up to the midpoint to create an S-Curve. See:
Is your brain full yet?
In our last post we had the following graph:
Below is the bar chart for the actual daily number of new cases up to 16 April (Day 49):
The midpoint appears to be around 31 March or 1 April. This would have been hard to see near this date since it is between two peaks (in a trough).
Below our estimates are included (e.g. if we had not gone into Lockdown Level 4 when we did):
The midpoints for these bar graphs appear to be around 30 March for Re** (single spike) and 31 March or 1 April for the rest.
Even when we have estimated the dates to get an estimate for the midpoint we then have to decide whether to double the estimate for a single date or sum two consecutive dates.
A single spike may indicate which choice to make. If two days have roughly the same number of new cases, we may choose to sum these for our estimate.
Our decision depends on whether we think the midpoint is in the middle of a day (say around midday) or between two dates (around midnight).
Fundamentally we will need to take into consideration what situation best fits the data following the midpoint.
We also need to consider whether we are using the local time or GMT (Greenwich Mean Time). New Zealand is twelve hours ahead of GMT.
My other COVID-19 posts can be found here:
Data for my posts can be found at:
I share my posts at: