Perspectives Overview

The Sisense Perspectives feature allows users to create sub-models that are fully dependent on their parent data model, while maintaining the independence of their own logical and semantic elements. By inheriting physical elements, they support a single source of truth, without requiring duplicate versions of the data model they are based on.

They are used to expose dedicated views of a data model to specific groups. This reduces the need to duplicate data and data models while enhancing the self service capabilities, which is especially valuable for complex use cases with many models, analytics assets (dashboards) or self service dashboard designers.

Data Models

Data models can be complicated to set up because they can have dependent assets and include different sources of data. The time consuming and error prone duplication of data models can be avoided in many cases by the use of Perspectives.

For more information about editing data models, see:

• Using the Model Editor

• Navigating the Model Editor - Overview

Advantages of Perspectives

Data Admins can use Perspectives to set up an agile environment without the obstacles to high performance and quality such as many-to-many table relationships, random paths and the unexpected breakage of dependent assets.

How Data Model Updates Affect Its Perspectives

The following types of changes to the root data model are immediately reflected in the data model's Perspectives:

• Physical:

  • Table and column removals

    • Adding tables and columns should be handled manually

  • Column type changes

• Semantic:

  • Table name changes

  • Column order changes

  • Removed relationships between columns

  • Hiding and showing hidden columns

  • Table description and tag edits - Once these types of changes are made in a Perspective, subsequent changes in the root data model will no longer propagate to the Perspective.

Training Course

Click here for the Perspectives training course.