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

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

This section details how to add dimensions in a Kyvos semantic model to organize data into categories and hierarchies, enabling users to analyze measures across different perspectives.

A dimension is a structure that categorizes facts and measures, allowing users to answer business questions. Commonly used dimensions are people, products, places, and time. Dimensions typically provide filtering, grouping, and labeling.

The elements of a dimension can be organized as a hierarchy - a set of parent-child relationships, where a parent member summarizes its children.

For example, cities are members of regions; products typically are part of product groups, and individual expense items are grouped into types of expenditure.

Levels of data range from the most summarized (up) to the most detailed (down).

You can add, name, and rename dimensions, define a hierarchy level, add levels, and set up attributes.

Note that some properties include an information icon you can click to learn more about that item.

To learn about attributes, see Attributes.

To add a dimension, perform the following steps.

1. From the Toolbox, click Semantic Models.

2. Select a semantic model from the list.

3. To add a dimension, do one of the following:

   1. Click the Add Dimension button, which is the plus sign next to the header of the Dimension column.
      A new dimension is created in the list. By default, it is named Dimension and includes a hierarchy and level, and attributes.

   2. Right-click in the dimension column and choose Add Dimension to add the first dimension group.

   3. Drag a candidate dimension from the list on the left into the dimension column.
      You will see an indicator, for example, Drop as dimension or Drop as level to guide you.

4. Do one of the following:

   1. Right-click the default name, choose Rename, then enter a name.

   2. Edit the properties in the Properties column and change the name.

5. The other property choices vary depending on the item you are editing.

6. Choose the type: regular, time, or location. You may also see bridge which is only available when a bridge node is available in the relationship.
   The available choices are based on the data type you are using.

   • Regular dimensions are typically people, products, or places.

   • Time hierarchies use time and date-related data. If you have a Custom Calendar, you can click the checkbox to use a custom calendar hierarchy in your time dimension.

   • Location hierarchies use location-related data (such as country, state, county, city)

   • To use any of the facts or dimensions from a many-to-many dataset, you must create a bridge dimension in the semantic model Physical view.

7. Choose the hierarchy and field based on the type by clicking the dimension and making a selection in the properties.

   1. If this is a time hierarchy, choose Year-Month-Day, Year-Quarter-Month, Year-Quarter-Week-Day, or Year-Quarter-Month-Week-Day. See the time hierarchies section for more details. 
      If there is not a pre-defined time hierarchy shown, set the field data type to a Date type such as year-quarter-month. 

   2. If this is a Location hierarchy, and location data is available, choose from the country choices, such as city, state, country, and postal code. Select a field from the last level of the hierarchy and the others are automatically fetched.

   3. Choose a field from the list. 

8. Choose whether to materialize the data. Yes means this field will be part of the processed semantic model and No means that it won't be included in the semantic model.

9. If appropriate for the data you are using, click Is a Slowly Changing Type 1 Dimension. See Slowly changing dimensions.

10. Optionally, add a description of up to 200 characters. The optional description is used when you hover the cursor over the dimension in a worksheet.
    
    

    Open

    

Time hierarchies

Kyvos allows you to use time hierarchies in two ways:

• Predefined: These are pre-created hierarchies in the system. You can select and add them to the semantic model from the hierarchies list from Year-Month-Day, Year-Quarter-Month, Year-Quarter-Week-Day, or Year-Quarter-Month-Week-Day.

• Custom: In this type, you can define different date values for different columns. For example, year=2021 and month=09. In this case, you can drag individual columns and set up the field data type to each field such as year-quarter-month-day.

You can use the following to indicate time values:

• Year: YYYY -2020

• Quarter: Q1, Quarter1, 1 , 01

• Month: Jan, January 01, 1

• Week:  w1, 01, 1, Week1

• Day: 01, 1

If your data contains values like " 2020-Sept " for a month, it is not a true month value and you will need to convert the data to a supported format. 

Related topics

• Defining hierarchies and levels