Skip to main content Skip to complementary content

Creating a data model

You can create relationships between the datasets that are included in a Storage or Transform data task to create a data model. The data model can be used by a data mart to automate code generation for denormalization of dimensions and fact tables.

  • Click Model to open the data model view.

You can get recommendations for relationships, or add your own relationships.

Using recommended relationships

You can get recommendation for relationships to create, based on primary keys and column names in the datasets.

  1. Click .

    You will see a list of recommended relationships. You can expand each relationship to inspect the suggested columns used to relate the datasets.

  2. Select the recommended relationships that you want to use.

    The model view is updated with dotted lines to indicate the relationships that will be created.

  3. Click Apply to create the relationships.

    The lines in the model view become solid to indicate that the relationships is created.

Adding relationships

You can also add relationships manually, if they are not included in the recommended relationships.

  1. Click Add relationship.

  2. Select a child dataset, and a parent dataset.

    You will see recommended relationships based on the primary keys of the parent data set.

  3. You can adjust the columns selections for the child dataset.

    You can also adjust the name of the relationship.

    When you are ready, click OK to create the relationship.

Editing relationships

  1. Click to go to the tile view.

  2. Click on a relationship and select Edit.

  3. You can now edit the relationship settings.

    Click OK when you are ready.

Deleting relationships

There are two ways to delete a relationship.

  • In tile view (), click on a relationship and select Delete.

  • In model view (), click on the relationship to delete.

Information noteDeleting a relationship can affect downstream data assets.


  • Circular relationships are not supported. If you get warnings about this, consider consolidating the two datasets into one.

  • The attributes used to relate two datasets must have the same data type. If they have different data types they are not ensured to be converted automatically, which may cause a runtime error when using this relationship later.

Did this page help you?

If you find any issues with this page or its content – a typo, a missing step, or a technical error – let us know how we can improve!