Lesson 11 — Refactor AI code for readability

Same behaviour, better shape

AI code often works but is hard to read: one-letter names, deep nesting, magic numbers. You will maintain this code, so make it readable. That is refactoring: changing the shape of code without changing what it does.

Typical moves:

• Good names: g(s) → letter_grade(score).
• Flatten nesting: else: if ... → elif ..., or an early return.
• Data instead of branches: a table of thresholds and a loop instead of five ifs.

SIZES = [(1000, "large"), (100, "medium")]
for limit, label in SIZES:
    if amount >= limit:
        return label

How do you know you did not change the behaviour? Compare old and new on many inputs. Here we check every score from 0 to 100.

Why it matters: readable code is code you (and your teammates) can still fix in six months. Tests make refactoring safe.