Elements of a Transformation

Transformation Canvas

The Transformation Canvas is where the transformation workflow is built. To add data inputs or transformation steps, drag them onto the canvas. Inputs and steps must be connected to create the order of operations for the transformation.

Each transformation input and transformation step is configured on the right-side configuration panel once it’s added to the canvas.

Inputs

Inputs supply the data used in subsequent transformation steps. They can be added throughout the transformation sequence.

Table

Tables are either ingested data source Tables or Data Views. Tables can have multiple transformation steps as outputs.

Child Dataset

Child Datasets are tables related to the input table by a relationship established on the model canvas. The Child Dataset must use a table as its input and can only be used if it is directly connected to the input table in the model. The result of this transformation step is the equivalent of a join between the two tables.

Transformation Type

Transformation Types are available on transformations with a host table in order to convert a Calculated Column to an Aggregated Measure or vice versa.

Aggregated Measures can only have one input Table. To change the type from a Calculated Column to an Aggregate Measure you must remove secondary inputs from the Canvas.

Transformation Steps

Transformation steps are used to manipulate and transform the input data. Every valid step must have an input and an output. Some steps may require more than one input and all steps can have multiple outputs.

Step Inputs are represented by triangles at the top of the Transformation Step Tiles

Step Outputs are represented by squares at the bottom of each Transformation Step Tile

For configuration instructions see the Transformation Step Section.

Output

To complete a transformation workflow, the final step must be connected to the output. On Calculated Columns and Aggregated Measures, the output step will define the column to append to the host table. On Data Views, the output step defines the last step of the transformation workflow where all columns will be stored.

The necessary output step is automatically added to the Transformation Canvas however, you must connect the last transformation step to the output to complete the transformation.

An automatically generated name will be provided upon connecting to the Transformation Output. The name can be edited once the upstream inputs to the Output Tile are valid.