D103 - Dimensional Modeling

Duration  2 days
Price  £1,000
Discount  10% for 2 or more attendees, 15% for 3 or more attendees from the same organisation.
Format  Online
Dates  Please contact us

Overview
This class provides a comprehensive coverage of dimensional modelling, including concepts such as star schema, snowflake, facts, dimensions, many to many relationships among other topics.
During this time of data analysis and discovery, it is critical to have a good data model. Whether you have an in-memory data model in Tableau, Power BI, or QlikView, or you are designing a data warehouse or a data lake, it is important to understand that only a good physical or logical dimensional model will offer usability, maintainability, growth, performance, and self-service capabilities. A good data model will be able to provide the answers to many questions without creating reports.
The examples and exercises of this class are covered in Excel, however the concepts learned could be applied to any technology. 

Objectives
  o Be able to create a dimensional model using advanced techniques.
  o Identify whether to use star schema or snowflake.
  o Learn different types of facts and measures.
  o Be able to model different types of dimensions.
  o Understand when and how to use slowly changing dimensions.
  o Identify how to incorporate concepts beyond dimensional modelling such as graphs. 

Audience
Data architects, enterprise data architects, data engineers, data analysts, data scientists, data warehouse and data lake house designers.

Prerequisites
Having previous exposure to multi-dimensional models such as Tableau, Power BI, QlikView, data warehouse or data lake house with facts, measures and dimensions will help to understand the concepts but it is not required.

Day 1:
Introduction
  o Star schema - Snowflake
  o Measures
  o Dimensions
  o Measure Types

Facts
  o Grain
  o Types of facts
  o Fact less facts

Dimensions
  o Attributes, Hierarchies
  o Keys and unknown values
  o Date and Time
  o Mini Dimensions
  o Degenerate
  o Self
  o Range
  o Junk   

Enterprise Modelling
  o Bus matrix
  o Conformed dimensions
  o Master data


Day 2: 
Slowly changing dimensions
  o Types
  o Keys
  o Unknown values

Multiple facts and dimensions
  o Many to many relationships
  o Bridge tables
  o Multi-value Dimensions
  o Hot-swappable dimensions

Analytical Tools
  o Many-to-many relationships
  o Role-Playing dimensions
  o Time Intelligence

Beyond Dimensional
  o No SQL
  o Hierarchies
  o Graphs