Concept and Tools

A Time Series Query specifies a query on the time series database of the Waylay Platform.

Timeseries data can be retrieved directly via the Waylay Broker API, aggregated or unaggregated, but the Time Series Query feature adds extra functions.

A Time Series Query

  • can include multiple series (multivariate data)
  • can be aggregated with one or more aggregation methods, and optionally interpolated for missing values
  • can be saved and executed by name
  • its results can be directly exported to CSV and other formats.
  • has an independent JSON representation that is compatible with the TS Analytics Model specifications

The concept is a spin-off of the Time Series Analytics features, and this originates its main intended usage:

A Time Series Query represents multivariate data set, with rolling or specified time window, as input to analytical processing such as BYOML models, TSA models or other analytical sensors or actuators.

Tools for working with Time Series Queries

The Query Designer is a graphical user interface that lets you define, manage and test Time Series Query definitions.

The timeseriesQueryData sensor is the primary choice when using a saved Time Series Query in the Rule Engine.

Time Series Queries can be created, modified, executed, … using the Query REST API of the Waylay Platform, and is supported by the Python Waylay SDK and the Javascript Waylay SDK.