Time series

A time series is a series of data points with corresponding timestamps. Each timestamp may have multiple versions with different values that have evolved over time. Each point therefore has the following structure:

{
  "time": "2021-11-17T00:00:00Z",
  "value": 3.1415,
  "knownTime": "2021-11-19T00:00:00Z",
}

Each time series in the Exabel data model is associated with one signal and one entity. For example, a company may have a time series for its revenue. That time series is then associated with that company and with the revenue signal.

By walking across relationship types in the graph when creating time series models, it is possible to aggregate a signal from a set of child entities to a parent entity that does not normally have that signal.

The Exabel Platform currently only supports modelling in daily or lower resolution. Timestamps must be normalised to midnight UTC.

The resource name for a a time series is on the form entityTypes/ns1.entityTypeName/entities/ns2.entityName/signals/ns3.signalName. An alternative name for the same time series is signals/ns3.signalName/entityTypes/ns1.entityTypeName/entities/ns2.entityName, but the former is the canonical version which always will be returned by the server.

Creating and updating time series points are eventually consistent.

-TimeSeriesService REST reference documentation
-TimeSeriesService gRPC reference documentation