Working with relationships

Applies to: Kyvos Enterprise  Kyvos Cloud (SaaS on AWS) Kyvos AWS Marketplace

Kyvos Azure Marketplace   Kyvos GCP Marketplace Kyvos Single Node Installation (Kyvos SNI)

Once you have datasets, you can create relationships among the different datasets. 

This process is very similar to creating table relationships. You can add dataset files, and you can even add the same dataset files multiple times. For example, you may have a table of employee data and want to create a relationship between employees and their managers. However, a manager is also an employee, so you need to add the employee data once to use it for managers and add it again to use it for employees.

Relationships are shown by a line drawn between fields in different files . The line specifies the relationship. 

Creating a relationship in 2 steps

After adding a dataset with Kyvos, you need to perform a few steps to create a relationship. These are, of course, a one-time activity; once the relationship is created, the designer users can use the relationship to process a semantic model. 

Step 1: Adding a dataset 
In this step, you add a dataset. See Creating a dataset for more details. 

Step 2: Creating a relationship 
In this step, you drag a field onto another field to connect the fields and create a relationship. 

Many-to-many relationships

While designing a relationship, you can create a many-to-many relationship between a Fact and a Dimension. See Working with many-to-many relationships.

Properties

In the Properties pane, you can do the following:

• Select a node, then view its properties. 

• Specify the node to be Fact, Dimension, or Bridge using the corresponding option from the Node Type area.

• Define the Relationship Style as a Line or a Curve.

Source

In the Source pane, you can do the following:

• View the source information for the selected node.

• Change the node source by clicking the link next to a dataset or view the node details in a different window by clicking Open Source.

• Change the node source by clicking the Actions menu (...) at the top of the page and choosing Change Node Source.

Relationships pane in the Properties panel

In the Relationships pane, you can view or change the relation type. For example, you can change from one-to-many to many-to-many by clicking the blue icon next to the Actions menu (  ⋮  ), then clicking Yes to confirm the change.

To add a relation, perform the following steps: 

1. Click the Actions menu (  ⋮  ) > Add relation and specify a source and target for an outgoing relationship, and then click Apply. 

2. To change the incoming or outgoing relationship to one-to-many, click the blue icon next to the Actions menu (  ⋮  ), and then click Change.

3. To change the operator between a source and a target, click the operator (the blue equal sign in the screenshot shown here) and choose equal to or between. If you choose between you must select a target.

4. To delete a relationship, click the Actions menu (  ⋮  ), and then click Delete to confirm the deletion.

Open

Relationships pane 

Collapse or Expand

Click either the Collapse or Expand All Tables in a Relationship icon at the top of the workspace next to Tips to expand or collapse all nodes.

Tabular view

For complex relationships, you can also view and edit the nodes in a table format.  The table format is best used when you have a larger number of nodes, for example, 25.  It helps you avoid additional scrolling and moving relationship lines. You can see all of the information more easily in table format. The table view allows you to more easily search, sort, and filter relationships. 

Once you have made changes using the tabular view, you can't switch back to the graphical view. 

To switch to the tabular view, click the Actions menu (...) and click Switch to Tabular View.

Raw data

You can create relationships using data that hasn't been processed into a semantic model. Select the file you want to use and create relationships using it. See Working with non-materialized or raw data semantic models 

Raw data is data that has no transformations and has not been processed into a semantic model. Kyvos provides tools to use raw data along with datasets that have already been processed. Calculations based on raw data are calculated at the time of the query. This means that certain types of calculations that depend on pre-aggregated data are not available. 

After you add raw data to your relationship and run a data profile job, you will see an option to right-click a raw dataset and choose Show Profile Result. See Creating a data profile. This information may help you create more useful relationships. 

Validating changes

The system automatically validates the design as you make changes. However, you can manually do it too. After you have made changes, click the Actions menu (...) and click Validate. See Validating an object.

Optimizing relationships

You can view a data profile result to see information that will help you optimize the relationships that you recreate. To learn more about data profiles, see, Creating a data profile.

Related topics