Data Cleaning Exercises

Module 20 · Job-ready delivery · 7 min

Take-home assessments

You'll be able to: Approach a take-home cleaning task so reviewers see your judgment.

Before this: Hand off a cleaning job: rules, exceptions, reconciliation

The idea

Read the brief twice, list assumptions where it is silent, and state them in your submission.

Deliver code that reruns, a clean table, an exception report with counts and a short note on decisions and trade-offs.

Time-box polish; explicit reasoning scores higher than extra features.

Worked example

Input: brief silent on duplicates

Visual
(document) Assumptions → survivor rule stated
SQL (DuckDB)
-- assumption: latest version per (order_id, line_id) wins
pandas
# assumption documented in README

Result: Reviewer sees the rule and why

Common mistakes

  • Hidden assumptions.
  • No reconciliation.

Check yourself

What should you do when the brief is silent?

State an assumption, apply it consistently and mention alternatives.

Practice it

Clean an order-line export.

Next: Clean an order-line export