SQL Lesson 6: Normalization

SheetDrill SQL SQL Lesson 6 · Normalization

SQL · Lesson 6

Ctrl/⌘ + Enter to run

✓

SQL Lesson 6 done.

You can look at a table and see where it will go wrong: the same fact typed in many rows, several values in one cell, a column that depends on only part of the key, or on another column instead of the key. And you can fix it by splitting the table so each fact is stored once. That's normalization.

Back to SQL

What you're learning

A table built like a spreadsheet repeats things: the customer's phone on every job, a part's price on every line. Repeats cause anomalies: change a fact in one row and forget another (update), can't save a fact without some other fact (insert), delete one thing and lose another (delete).

Normalization splits the table until each fact lives in one place. 1NF: one value per cell, so lists become rows. 2NF: with a two-column key, every column has to depend on the whole pair, not one half. 3NF: no column depends on another non-key column. The short version: every column depends on the key, the whole key, and nothing but the key.

The design drills check your tables by adding rows and changing a fact once: the change has to show up everywhere.