Stream data points

A data point is an individual data point stored in a stream, representing a measurement or observation at a specific index value. Each data point contains values for all properties defined by the stream's type, forming a complete record at that particular moment or sequence position. For example, a wind turbine data point might include a timestamp of 2024-01-15T10:00:00Z, a power output value of 2450.5 kilowatts, and an operational status of Operating. Data points are the atomic units of data in stream storage—when you write data to a stream, you're creating data points, and when you read data from a stream, you're retrieving data points.

The structure of every data point is determined by its stream's type. Because types are immutable and enforce strict schemas, you can rely on every data point in a stream having the same properties with consistent data types. This predictability makes it straightforward to work with stream data programmatically and ensures data quality across your entire data collection infrastructure.

The following table shows data points from a wind turbine stream. Each row represents a single data point with its index value (the timestamp that determines data point ordering) and property values (the power output and operational status measurements at that moment):

Notice that each data point has a unique timestamp that orders it in the sequence, and all data points include values for the same properties defined by the stream's type. When you write these data points to the stream, stream storage uses the timestamp index to maintain chronological order. When you read data from the stream, you retrieve complete data points containing all property values for the index points you request.

Data point structure

Every data point consists of an index value and property values that together form a complete data record. The index value determines the data point's position in the stream's sequence—this is typically a timestamp for time-series data or a sequence number for ordered data points. The index value comes from the property designated as the key in the type definition and must be unique within the stream, ensuring that each data point occupies a distinct position in the sequence.

The property values contain the actual measurements or observations recorded at that index point. Each property value must match the data type specified in the type definition. For example, if the type defines a Temperature property as a Double, every data point must provide a numeric value for that property. Some properties may be marked as optional in the type definition, allowing data points to omit those values when they're not applicable or available. Required properties must have values in every data point written to the stream.

When you write an data point to a stream, stream storage validates that the data point conforms to the type's schema. This validation ensures that the index value is the correct data type, all required properties are present, and all property values match their expected types. If an data point doesn't conform to the schema, the write operation fails, maintaining the integrity of the stream's data.

Options for writing data points

Stream storage provides multiple ways to write data points, each designed for different integration scenarios and data collection patterns. The method you choose depends on your data source, technical infrastructure, and whether you're building custom integrations or using existing systems. For more information on these options, see Data connections (early access).

Query patterns for reading data points

Stream storage provides several query patterns for retrieving data points from streams. The following patterns are available through the portal, enabling you to access data point data without writing code.

Tip: Additional query capabilities are available when using the REST API programmatically.

• Get first value retrieves the earliest data point stored in the stream based on the index value. This is useful when you need to know the initial measurement or starting point of your data collection, such as when a sensor first came online or when monitoring began for a particular asset.

• Get last value retrieves the most recent data point stored in the stream based on the index value. This is particularly valuable for dashboards and monitoring applications that need to display current conditions or the latest measurement from equipment.

• Get range values by range retrieves all data points between a start index and end index. When you specify these boundary values, stream storage returns all data points whose index values fall within that range, maintaining their sequential order. This pattern is fundamental for trend analysis, generating reports over specific time windows, or comparing measurements across different periods.

• Get range values by count retrieves a defined number of data points starting from a specified index value. Rather than specifying an end index, you provide a start point and indicate how many data points you want. For example, you might request the 100 data points immediately following a particular timestamp. This pattern simplifies queries when you know how many data points you need but don't know the exact end index value.