Time Series Mixin Types

You create time series in the C3 Agentic AI Platform by combining a header Type and a data series Type.

• The header Type contains metadata about the time series attributes.
• The data series Type contains the individual data points of the time series data.

The fields and structure of these Types depend on the attributes of the time series data stored in the associated data series Type.

To aid in the construction of time series Types, the C3 Agentic AI Platform provides built-in Types for common time series scenarios.

Use the mixes keyword in the Type declaration for your header or data series Type to inherit fields and methods from the appropriate built-in Type.

To decide which built-in Type to mix in to create time series data, consider the following attributes of your data:

• Determine whether the incoming time series data is interval or point data.
  • Interval data are values valid over a given time range and contain start and end values.
  • Point data are measurements from specific points in time.
• Identify whether the incoming data requires normalization.
  • Normalization is a process where duplicate, overlapping, missing, or out of order entries are preprocessed to create a consistent time series.
• Is the incoming data high or low frequency?

Use the following table to select an appropriate mixin for your header and data series Types:

For examples of how to mix in header and data series Types, see the Model Data as Time Series tutorial.