Join and Aggregate Data | JOIN, GROUP BY, HAVING & NULLs

Description:
This is Part 2 of the SQL series. Here, we move beyond querying one table and learn how to combine data without creating incorrect counts, duplicated totals or missing records.

The central question is simple: what does one row represent before and after a join?

You’ll learn how to:
- Understand the grain of a table
- Work with primary and foreign keys
- Recognise one-to-many relationships
- Join tables using matching keys
- Understand how joins change the meaning and number of rows
- Avoid double counting after a join
- Use COUNT, COUNT DISTINCT and SUM
- Aggregate data with GROUP BY
- Understand the difference between WHERE and HAVING
- Keep unmatched records with LEFT JOIN
- Understand the important difference between filtering in ON and WHERE
- Display zero instead of NULL with COALESCE

Practice files and SQL exercises: https://github.com/david-ikenna-ezekiel/thedatasignal/tree/main/sql-masterclass

Chapters
00:00 Why joins can change your answer
00:24 Grain, keys and table relationships
02:31 Counting the inputs before joining
03:35 Preserving grain and calculating revenue
05:45 Avoiding double counting with aggregation
08:20 LEFT JOIN, ON versus WHERE and validation

If this lesson helped you understand joins more clearly, subscribe to The Data Signal for more practical SQL and data tutorials.