Beginning Power BI for Business Users

6Modeling Data

In earlier chapters, we've talked about tables, columns, data types, and relationships. In Chapter 5, I walked you through preparing your data to be ready for reporting. But I hope there are some lingering questions in your mind: “How do we know what data should be in which tables? Should we just bring in the data from our sources and mash it up together? Even then, how should we mash it?”

This is where data modeling enters the scene. We need to have a method to model our data that both makes sense to our report consumers and is designed to work well with Power BI. While I've used the term multiple times over the previous few chapters, it would be good to review what we mean by modeling data.

A data model is a way of organizing data as it corresponds to properties of real entities. For example, we could create a data model that represents real weather events. Those weather events have properties about them such as the type of weather event, the time the weather event occurred, the duration of the weather event, and the location of the weather event. A data model might represent all that data in one table or in multiple tables, depending on how the data model will be used. For example, if we are going to create a system to record weather events as they happen by gathering data from many sensors across a region, that system will need to perform very quickly. In this case, my data model must be able to handle data entry at the millisecond (or even nanosecond) level. ...