Define a data view

The DataView object is a declarative query and shape for CONNECT data services data. This section describes the DataView object. See the Data View API section for the corresponding API routes.

At times, this section makes reference to ways the view resolves into further resources, such as the collection of data items found by each data item query. See the Resolved Data View and Resolved Data View API sections for details.

General concepts

You define multiple aspects of a data view when you define it, including data items and their organization, grouping, and shaping.

Define a data view identifier

A data view must have a unique identifier. It may have a friendly name and description. If a friendly name is not specified, the identifier will be used as the data view's name.

Include data items

One or more queries determine the set of data items (such as streams and assets) that the data view will include. Each Query should represent a collection of like data items. Use separate queries to include data items that represent very different items, such as power inverters and weather stations.

Include data fields

Information from, or about, those data items must be included explicitly in the data view as Fields within DataFields or GroupingFields. By default, a data view includes no fields, except the default IndexField.

The fields that are available for use are exposed as a "resolved" resource, Available Field Sets. It is recommended to use or adapt those fields instead of defining new fields. Individual fields are organized into DataFieldSets and GroupingFields.

Group the data items

Data items may be grouped, which amounts to grouping or partitioning them by one or more field values. To group data items in a data view, define one or more fields as .GroupingFields of the data view. This is one way of producing a meaningful, consumable shape of data.

Organize the data fields

Field sets and fields resolve in the order they are defined. They may be re-ordered.

Within each group, a field set may be associated with multiple data items. It is often necessary to disambiguate these items. If a group contains multiple data items from the same query, you may wish to identify those data items so that they are not ambiguous and so they will be aligned across groups. Items can be disambiguated by specifying an .IdentifyingField. An identifying field is a field that differentiates the data items within a group, such as the value of Measurement metadata (i.e., the data items are identified by what they measure). Identifying data items also allows the data views engine to "align" them across groups, since it is clear, for example, that streams measuring Power Out and streams measuring Power In are alike. Ensure that each data field's label includes the {IdentifyingValue} token so the field labels are unique. The default field labels already include it.

Include the index field

The field used for indexing. If not specified, a default value is applied. If specified, a label is required. Field source and keys are not applicable for index field.

Define index type and default range

The data view is targeted to include Streams and Assets of a common index property type. Often in CONNECT data services, data is indexed by its timestamp property (of type "DateTime"). In certain cases, data may instead be indexed by a numeric value such as depth. Data views supports all index type codes allowed by the Sequential Data Store with the exception of TimeSpan. If not specified, the default .IndexTypeCode is "DateTime". Compound index types are not supported.

Default values may be defined for the start index, end index, and/or interval used when data view data is queried.

Define data view shape

Data views may be set to resolve as standard shape or narrow shape. Standard shape resolves fields similar to how they are defined. Narrow shape pivots the fields vertically, resulting in a view whose schema is independent of what data items are resolved by the data view. Narrow shape may be used when an invariant output schema is required.

Data view properties

The following table lists the properties of a DataView:

Rules for Id property

• Is not case sensitive, but case is preserved
• Can contain spaces
• Cannot contain forward slash ("/")
• Can contain a maximum of 100 characters

Related object types

The following sections describe the classes and enumerations used when defining data views.

Query

A query for CONNECT data services resources to include in the view.

DataItemResourceType enumeration

The DataItemResourceType enumeration specifies the CONNECT data services resource type included in the data view query.

FieldSet

A set of fields included in the data view, sharing a common role and query. One DataView is likely to include one FieldSet per query.

Field

An individual piece of information, such as a property of an SDS stream, an asset stream reference, or metadata of that stream or asset. All sources except FieldSource.NotApplicable can be used as data fields. Fields from sources FieldSource.Id, FieldSource.Name, FieldSource.Metadata and FieldSource.Tags can be used as grouping fields and identifying fields. Some sources are used in conjunction with the Keys property (see below).

FieldSource enumeration

For fields that derive data from a data item (e.g. a stream or asset), the FieldSource enumeration specifies the part of that data item that a Field resolves to. Some sources require one or more .Keys to be specified on the field, such as PropertyId, in which a key is the id of a desired property.

SdsTypeCode enumeration

SdsTypeCode enumeration is the name of a data type. It is used when defining a data view, where the specified .IndexTypeCode determines which data items are eligible for use in the data view.

See SdsTypeCode for details.

DataViewShape enumeration

DataViewShape enumeration describes possible output shapes for a data view.