Acquiring Raw History

At some point, you may need to visualize your stored data as a dashboard line chart, send it as a report, or just process it and store the result in a new table. This means you need to retrieve event or variable history as raw data.

Regardless of the type of database used for keeping historical data, Iotellect Server core provides unified methods for loading history.

Get Event History is a function of the Events context, which can be called via the {events:get()} reference. It's a basic instrument that makes a request to the storage, based on several parameters:

• The mask of the source contexts and the name of the event to load

• Filter expression for the loaded events

• Date/time range

• The sorting direction of the resulting table

• The maximum number of events to load

All events are returned in a large Data Table that can be immediately used as a source for a report, machine learning algorithm, etc.

The Get Variable History function of the Utilities context loads historical values of a certain context variable, i.e., change events matching the history of this variable's updates. It can be called via the {utilities:variableHistory()} reference.

Parameters of this function include:

• Path of context and name of the variable whose history should be loaded

• Date/time range

• Sorting direction

• Maximum result count

All historical values are returned in a large Data Table:

This table is suitable for:

• Building a query over historical values

• Building a variable history report

• Exporting data to third-party systems

Here is an example of an expression that returns the history of the variable temperature defined in the device users.admin.devices.meter: {utilities:variableHistory("users.admin.devices.meter", "temperature", "2012-05-03 12:00:00.000", "2012-05-04 12:00:00.000")}

In some cases, it's necessary to aggregate loaded values by periods and calculate averages, minimums, maximums, etc.

The Summary function of the Utilities context is a very powerful way of on-retrieval data aggregation. Its input parameters allow to:

• Specify multiple data series to be loaded and processed at once

• Specify source context masks and names of events/variables whose history should be processed

• Specify date/time range

• Specify custom expression used to extract values used for further aggregation

• Specify custom expression used to extract timestamps from real-time data samples

• Define grouping periods (e.g. hours of months) and timezone used for time slicing

• Calculate various aggregates, such as averages, number of samples in a period etc.

• Intelligently handle various counter-type values

• Process out-of-range values

The summary() function also returns all resulting data as a large Data Table that can be immediately used by any other data processing tool, such as model's rule sets.

Was this page helpful?

Tell us more

Send

Thank you!