Dimensional Modeling in Python

Given a dataset in a Jupyter notebook, how would you separate it into dimension and fact tables? Implement it using pandas (melt / pivot expected).

Asked in the Jupyter live coding (Python) stage. This was Question 1 of 3 in the pandas round; candidates were advised to keep the model simple and know melt and pivot syntax.

For grading, use the supplied notebook_dataset table as the wide input representation and return its normalized customer-month fact records.

Output

1. One row per customer and month, with customer_id, customer_name, country, month_start, card_spend, and transfer_count.
2. Include all customer-month combinations, including rows with NULL measures.
3. Order by customer_id, then month_start.