TimedValueHistory vs. TimedDataPoint

#1

What’s the difference between TimedValueHistory and TimedDataPoint? They seem to be incredibly similar.

#2

TimedDataHeader should only be used if you want timeseries to be normalized. In this case you should add the annotation @ts on each field you want to normalize.

If you want to simply represent a timed relation you should mixin either TimedIntervalValueHistory or TimedValueHistory

#3

Another follow-up question: What is the difference between TimedDataPoint and TimeseriesDataPoint?

Thank you

#4

TimedDataPoint should be used to record “Point In Time” values. E.g. temperature reads happen at a particular instant in time, so there is a particular timestamp associated when the read was taken. There is not natural “end” date for it. Using “start” & “end” date values to define the period of for which the temperature reading was valid would not be natural. In these scenarios use TimedDataPoint which only has a “start” datetime but NOT an “end” date.

Similarly, when recording electricity bills, you know the exact time period for which the bill is, in this case there is a “start” date and an “end” date for the bill. In these cases uses TimeseriesDataPoint