Clean Lantern’s order export

Ana Lim is preparing Lantern Coffee Co.’s first review of its online shop for 1 March – 30 June 2026: revenue by region, how subscribers’ orders compare with one-off orders, and the average customer rating. Her analyst has downloaded the order export, lantern-orders.csv, straight from the shop platform: 240 rows and 14 columns. You will take it from raw to analysis-ready in the Data Cleaning Lab, then write the data quality report that lets Ana, and anyone who checks her figures, trust the numbers.

Two facts from outside the file are available as evidence: the shop system counts 231 orders in the period, and it confirms that order LC-10042 was for 2 bags. The data dictionary in the dataset’s README describes what each column should hold; the lab knows its rules.

Lantern, Ana, the customers and every value in the file are fictional; the e-mail addresses use reserved example domains. The file was generated with known defects, so each stage’s questions check your results against what was planted. Stages 1 to 4 each have their own lab: stages 3 and 4 start from the table produced by the standard steps of the stages before, so that everyone’s answers can be checked.