Book Image

Extreme DAX

By : Michiel Rozema, Henk Vlootman
Book Image

Extreme DAX

By: Michiel Rozema, Henk Vlootman

Overview of this book

This book helps business analysts generate powerful and sophisticated analyses from their data using DAX and get the most out of Microsoft Business Intelligence tools. Extreme DAX will first teach you the principles of business intelligence, good model design, and how DAX fits into it all. Then, you’ll launch into detailed examples of DAX in real-world business scenarios such as inventory calculations, forecasting, intercompany business, and data security. At each step, senior DAX experts will walk you through the subtleties involved in working with Power BI models and common mistakes to look out for as you build advanced data aggregations. You’ll deepen your understanding of DAX functions, filters, and measures, and how and when they can be used to derive effective insights. You’ll also be provided with PBIX files for each chapter, so that you can follow along and explore in your own time.
Table of Contents (17 chapters)
Free Chapter
1
Part I: Introduction
6
Part II: Business cases
15
Other Books You May Enjoy
16
Index

Date tables

Most, if not all, Power BI models contain data that is related to dates. A date table (or calendar table, or whatever you like to call it) is therefore a common ingredient of a Power BI model. A date table has a special place in the model, because of DAX time intelligence functions (more on these in Chapter 1.4, Context and Filtering).

A date table must have one row per day for an uninterrupted period of time. Typically, this period should be large enough to cover all the rows in your fact tables. It is advisable to start and end the date table on year start and end, respectively. The date table must have a date column, being the unique key of the table (you may choose the name of this column yourself). Other columns in the table are attributes for each day, like year, month, quarter, weekday, and so on.

Power BI models have an Auto date/time feature which, when turned on, creates a hidden date table for each column in the model that has the date or date...