Skip to main content
Version: I2024.2.x

Working with Datasets

A dataset corresponds to a physical table or view in a data warehouse, or the results of a SELECT statement. The first step in designing an AtScale cube is to import the physical tables or views that the cube will be based on. (Later, you can create additional datasets that are based on the results of SELECT statements; these datasets are called query datasets.) To import a table or view as a dataset into an AtScale cube definition, the AtScale organization in which your project and cube are located must be connected to a data warehouse.

  • Importing Tables and Views into AtScale Projects as Datasets The first step in designing an AtScale cube is to import the physical tables and views that the cube will be based on.
  • Adding Query-Based Datasets A query-based dataset allows you to select data from one or more tables in your data warehouse and save the query in AtScale as a new dataset.
  • About Composer Composer enables you to deep-dive into a dataset, view the data, and manipulate columns.
  • Using Composer How to use Composer to transform columns into a dataset.
  • Refreshing Dataset Schemas The schema of a table or view can be modified over time. You can refresh the schema to pick up changes from your data warehouse since the last time the dataset was imported, such as new columns that were added.
  • Remapping Data Sources The datasets used in a cube definition are from a data warehouse to which AtScale is connected. After a cube is created you could configure connections to other warehouses that contain the same datasets. Then you can set the cube to use a warehouse that is different than the original one.
  • Removing Datasets from Cubes Removing a dataset from the canvas takes it out of the cube definition, but the dataset will remain in the project library. Any measures or relationships modeled on columns of the dataset will be removed from the cube definition as well.
  • Removing Datasets from Projects Removing a dataset from the project library takes it out of the project entirely. Any cube metadata that depends on the dataset will be removed as well.
  • Viewing a Sample of Dataset Rows AtScale lets you to view a sample of a dataset's rows. Viewing a sample of rows allows you to examine the column values and decide which data to base your cube attributes on.
  • Adding Calculated Columns to Datasets for Simple Data Transformations Calculated columns are a way to add simple data transformations to a dataset. After a calculated column is added to a dataset, it can be used as the basis of a cube attribute, just like any other dataset column.
  • Searching for Columns in Datasets Datasets can contain a very large number of columns. If you a looking for one or more particular columns in a dataset on the Cube Designer canvas, however, you can find them quickly by using one or more search criteria in the search field at the top of the dataset. Columns not matching the criteria are filtered out, so that the dataset lists only those that match.
  • Swapping the Physical Tables or Queries that Datasets are Based on When working in the cube design canvas, you can change the underlying table or query that a dataset is pointing to.