Self-servicing in Kahua Analytics

Self-servicing in Kahua Analytics enables builders to update report files according to their needs. Reference the sections below for a high-level overview of best practices and techniques. For more in-depth technical and instructional information, refer to the training available in our Training Portal.

Note For more general information on Kahua Analytics, refer to Kahua Analytics.

Key elements to know about this feature include the following:

  • The Kahua Analytics team hands over the desired report as a .pbix file so that the Analytics Builder can work on the most up-to-date file.

  • Analytics Builders connect reports to the Kahua Data Store through the API.

  • Updated report files are submitted to the Analytics team for review before they are released to Production.

  • The Analytics Builder 2.0 course provides more in-depth information about query folding, parameters, DAX, M code, and other technical topics.

How to . . .

The process for updating report files as a Kahua Analytics builder is as follows:

  1. File Handover: The Kahua Analytics team hands over the desired report as a .pbix file. This enables the builder to work on the most up-to-date file.

  2. Building in Power BI: The Kahua Analytics Builder updates the file to meet their Analytics needs.

  3. Email to Analytics Requests: The Kahua Analytics Builder sends a release email to AnalyticsRequests@kahua.com. In this email, the builder should attach the latest .pbix file along with any information that needs to be updated on the file, such as the addition of partition hierarchy filtering or data scope.

  4. Release: A member of the Data Science Team will review the file to ensure the self-servicing standards are met. Once these checks are validated, the file will be released to Production.

To pull in data from Kahua, all Analytics Builders must connect to the Data Store via API. Follow the steps below to do so.

  1. In Power Query Editor, navigate to the Parameters folder on the left-hand side of the screen.

  2. Click the BearerToken parameter and set it to your Data Store API Key.

  3. Ensure the SourceType parameter is set to API.

  4. Upon selecting a query, you will receive an error message with an Edit Permissions button. Click this button. The file should now populate the data preview.

  5. Once the refresh is complete, click Close and Apply in the top-left corner of the Power Query Editor. The file is now connected to the Data Store via the API Key and credentials.

To change environments, obtain the API Key for the desired environment and update the HostURL parameter.

You can check whether Partition Hierarchy Filtering is enabled for your current report in the Analytics tab of the Configuration application. If it is enabled, the URL Query String field will be filled out. If it is not enabled, follow these steps:

  1. Ensure ProjectLineageParent is loaded into Power BI and related to Project[Id].

  2. For the applicable dashboard, navigate to the Analytics tab of the Configuration application and enter the filter string in the URL Query String field.

Power BI parameters enable reports to connect to the Data Store accurately. All Analytics Builders must set up and utilize the appropriate parameters in their reports. These are outlined in detail in the Analytics Builder 2.0 course.

Follow the steps below to create a new parameter.

  1. In Power Query Editor, navigate to the Manage Parameters button in the top ribbon.

  2. Click New.

  3. Enter the parameter name, select the type Text, and enter the appropriate value in the Current Value field.

Launch Links enable the user to click on a data point in Analytics and open the corresponding Kahua record. These links can appear as hyperlinked text, using cell-specific conditional formatting, or as link icons. They are powered by parameters and can be set up as calculated columns. For the exact DAX necessary to build them out, refer to the Analytics Builder 2.0 course.

All files must be equipped with proper error handling functions. These are set up within Power Query as blank queries. The exact functions to use can be found in the Analytics Builder 2.0 course.

If there is no data in a table or column, these functions create blank tables and/or columns in its place. For example, if the RFI table has no data populated, the error handling will create a blank table called RFI. Once records are created in Kahua and pulled in accurately, the table will refresh with the appropriate data. This prevents Power BI visuals from erroring out. Instead, they will show blank.

Query folding offloads the processing burden from Power BI to the original data source. It pushes query transformations, such as filtering or merging, back to the data source rather than processing them locally. To check if your queries are folding correctly, follow the steps below:

  1. In Power Query Editor, select a query.

  2. In the Applied Steps pane, right-click on the last step.

  3. If your queries are folding, View Native Query will be clickable. If it is not, query folding has broken along the way. Refer to our Analytics Builder 2.0 course for information on how to handle this.

Question Description
I want more technical information than is included here. Refer to our Analytics Builder course for in-depth instruction on query folding, parameterizing, and more. It includes specific DAX and M code to help you build out scalable and efficient reports.
My queries aren’t folding.

There are several transformations that can break query folding. These include:

  • Adding index columns

  • Adding custom or conditional columns

  • Certain data type changes, such as changing to Text type

  • Using Group By, Sort By, or Joins for certain queries

If you are using these queries, add them as the last transformation step. This ensures query folding is active throughout previous steps.
In the same table, some of my data points are hyperlinked while others are not.

This issue could point to a problem with the relationships in the data model. However, if these are accurate, it might be that Power BI has become confused with naming/numbering conventions. If multiple records have the same values, such as Project A having Record 01 and Project B having its own version of Record 01, Power BI will not generate hyperlinks for them.

In situations like these, it is best to set up the launch link as a URL icon.
All users can see the same data in the report regardless of their permissions.

Partition Hierarchy Filtering is not enabled. Follow the steps in Set the URL for Partition Hierarchy Filtering above to enable it.
Visuals in Power BI are generating errors.

This is likely due to a lack of error handling. Check Power Query Editor to verify all tables have error handling applied.

If the visuals rely on calculated columns or measures, check to ensure these have not been deleted. If they have not, verify that the DAX syntax is correct and the necessary columns are present.

Note For extensive reporting, an in-house data warehouse is recommended.