Just making models – WW-02-04

Why would you want to create a Star Schema when that means extracting the data from what would be a fact table? An eternal question and one to which there is not a simple or one size fits all answer. Ultimately as with so many things in Power BI, it will come down to how much data is involved? The benefits are that your model should be quicker. However, when you don’t have much data, you may not realise any advantages. For our example, we are pulling ~ 2000 unique records from ~5500, so the benefits will start to be seen for us (we may do some side by side comparisons)

one of the significant lessons in this series is that Power BI is adaptable and able to cope with changes and growth. As a tool, it has Continuous Improvement baked into it. Ask any Power BI Developer, and they will tell you the same story, that a set of follow-ups follows every delivery. It is easy for this to become soul-destroying, but it is ultimately the power of Power BI. The results are so engaging and accessible that they lead to people asking for the next level down. You will learn to anticipate those questions and already be planning the next phase. With experience, you will find you already plan most of the requests, but you will never get them all.

In previous videos, we have often looked at “Unpivot” in Power Query or Dataflows, and this is an excellent function to help reshape data for use by an analytics engine. Today we will do almost the reverse by using the EventParticipants to build a People table. We can quickly create a new People Dimension by using “Group By” and then some aggregations. Make notes and add another tool to your arsenal.

%d bloggers like this: