Aggregate data by time

You can downsample and upsample data from the catalogue using the /parameters/data endpoint.

Before you begin

• If you don’t already have any Stream data in your workspace, you can use any Source from our Code Samples to set some up.

• Get a Personal Access Token to authenticate each request.

Aggregating and interpolating

The JSON payload can include a groupByTime property, an object with the following members:

• timeBucketDuration
  The duration, in nanoseconds, for one aggregated value.

• interpolationType
  Specify how additional values should be generated when interpolating.

For example, imagine you have a set of speed data, with values recorded at 1-second intervals. You can group such data into 2-second intervals, aggregated by mean average, with the following:

{
    "groupByTime": {
        "timeBucketDuration": 2000000000,
    },
    "numericParameters": [{
        "parameterName": "Speed",
        "aggregationType": "Mean"
    }]
}

You can specify an interpolationType to define how any missing values are generated. Linear will provide a value in linear proportion, whilst Previous will repeat the value before the one that was missing.

{
    "from": 1612191286000000000,
    "to":   1612191295000000000,
    "numericParameters": [{
        "parameterName": "Speed",
        "aggregationType": "First"
    }],
    "groupByTime": {
        "timeBucketDuration": 2000000000,
        "interpolationType": "None"
    },
    "streamIds": [ "302b1de3-2338-43cb-8148-3f0d6e8c0b8a" ]
}