Data Modeling with Microsoft Power BI

Chapter 15. Real-World Examples Using Power Query and M

In addition to the standard tasks to transform a given data model into a star schema, there is a lot we can do in Power Query to prepare data for advanced challenges. In many cases, Power Query is better than DAX when it comes to shaping data models. Keep in mind, though, that some solutions require DAX measures to be written, which can’t be replaced by even sophisticated Power Query or M. I also use scripts in M to make solutions more dynamic, lessening the effort needed to maintain solutions when there are changes in the data source.

The use cases in this chapter are listed here:

• How to group values into bins or buckets to show the name of the bucket instead of the actual value.

• How to support multi-fact data models by bridging the many-to-many relationship, which will appear between some of the relationships between a fact and a dimension table. I will demonstrate the solution on the example of a budget.

• In my multi-language solution, Power Query synthesizes the translations. I will show you how you can use Azure Cognitive Services to get the texts translated.

• Key-value pair tables must be pivoted to be able to satisfy common reporting requirements. You will learn which buttons to click in the UI to pivot the table so that every key becomes a column of its own. I will also show you how you can implement a dynamic solution that will take automatically care of new keys.

Let’s start with binning.

Binning

Let’s ...