Data Warehouse Express guide

Define a data warehouse as a centralized repository that stores cleaned and transformed data from one or more sources for analysis, visualization, and reporting to reveal patterns and support decisions.

Model a data warehouse in four steps: identify a single business process and its source data; set the grain; design dimensions for filtering; and build fact metrics from events.

Dimension provides context for facts in a data warehouse by answering how, where, what, who, why, and when events occur. It links stores, customers, time, and promotions to those facts.

Identify numeric metrics from business events and record them in a fact table that tracks a single event type, such as product sales in retail, to enable summarization and aggregation.

Present the star schema as a data warehouse model that contrasts with transactional entity relationship systems, addressing data duplication and performance by reducing tables to one or two.

Examine structuring a data warehouse by defining dimensions, storing solar system data for points, considering normalization, mapping entity relationships, identifying the entity key, and handling changes from the solar system.

Explore denormalization by consolidating multiple related tables linked by foreign keys into a single customer dimension table for a data warehouse.

Define dimensions with a primary key as a unique identifier, preventing duplicates by allowing the data warehouse to generate a new key, Acerbity, with all keys in third consecutive numbers.

Avoid dimension attributes; use the scripted text values for not available, known, or not applicable, and let the business or data analyst define the default identifier in this data warehouse.

Explore slowly changing dimensions in a data warehouse, comparing type 1 and type 2 behaviors, including surrogate keys, start and end dates, and how history is handled.

Explore spatial dimensions and how the warehouse discovers and handles special dimensions such as roleplayed, Ljung dimension, and snowflake dimensions.

The date dimension enables time-based analysis by filtering and grouping on the calendar date, while using a specialized primary key that encodes the date for this role.

Explore the role playing dimension in data warehouse by linking events with multiple dates to a date dimension, establishing independent relationships, and using views to assign bills different roles.

Junk dimension concepts group attributes from different entities into real combinations that can be applied in the real world to avoid dimensions with a single column or very few rows.

Explore how Snowflake dimension structures balance plane dimensions with occasional normalization, linking two tables by a foreign key. Understand why normalization may hurt performance and confuse navigation for users.

Explore data warehouse basics, designing fact and dimension structures, where events drive time-based summaries, and primary keys link tables; cover additive, semi-additive, and non-additive measures and changes.

Explore fact types in data warehousing, including transactional facts with measurements tied to dimensions, daily event sums, and factless tables for promotion events, plus handling zero sales days.

Learn how dimension tables require a primary key, how surrogate keys are generated during etl processing, and how dimensions become foreign keys in the data warehouse for source system keys.

Identify three measurement types across five tables and show how total sales can be grouped by related dimensions. Warn about measurements like balances or percentages restricted to specific dimensions.

Examine two approaches to source changes: discard later changes to keep the original, or apply the recent changes over time; for growing table, create a snapshot to preserve historical data.

Explore data warehouse concepts through a practical hotel booking example. Learn how bookings and related dimensions define the grain by customer in a single fact table.

Learn how a data warehouse models dimensions and a fact table within a schema, using surrogate keys and type three dimensions to track current and historical date attributes.

Learn the basics of data warehouse express and prepare to continue your learning journey. See you in the next course.