Minute data — aggregation/rounding
I am trying to reconcile my minute level data with the aggregated half-hourly, hourly, and daily data. However, there seems to be some discrepancy.
The half-hourly, hourly, and daily data reported by the API, Bright App, and my energy supplier (Octopus) are consistent (subject to rounding).
However, if I take the minute level data returned by the API and sum it, then there is a discrepancy between the summed data and the half-hourly, hourly, and daily data.
Do the aggregation functions apply some form on intermediate rounding, or is there some other explanation? I also note that the minute level data is reported to an inconsistent number of significant figures.
Any help would be much appreciated.
Comments
Sadly they will never match - the minute level data is calculated as the average of the ‘instant use’ readings taken 5/6 times in that minute. They are summed into the 30 minute slot. Overnight the real 30 minute reading comes from the DCC and corrects the 30 minute slot. It is the best we can do given there is no actual measurement of minute level data.
Ah right, I see. Thanks, Clive.