Fact Table
Are generally numeric in value
They can also be states of a system (Low, Medium, High)
Values for measurements are taken continuously and each new measure is a new value
There generally include textual context like units (C, F, kg, etc.)
Fact tables have many-to-one relationship with Dimension Table
Data that has direct impact on the performance of organization → Measure, Metric, Fact, KPI (Key Performance Indicator)
They are of three types: Addictive, Semi-additive, Non-addictive
Types of Fact Tables
Fact-less Fact
Does not contain any measure (All values are FKs to Dimension Tables)
Snapshot Fact
Stores data wrt. a certain point in time
Centipede Fact
A fact table that is normalized
Conformed Fact
Stored facts that are used by multiple dimension tables
Dimension Table
They are the context around the measure that is collected
They are logically grouped into dimensions
The dimension tables have a single primary key that is mapped to the foreign key of the fact table
There is one-to-many relation between Dimension and Fact tables