Which window function should fill the placeholder to return the most recent row per CustomerID?

Which window function should fill the placeholder to return the most recent row per CustomerID?

• A. ROW_NUMBER()
• B. RANK()
• C. DENSE_RANK()
• D. MAX()

Answer: (A)

The idea is to assign a unique order to rows within each CustomerID so you can pick exactly one row—the most recent one. Using ROW_NUMBER() with an order by a timestamp column in descending order creates a per-CustomerID sequence like 1, 2, 3, and you can then take the row where the sequence equals 1 to get the most recent entry. RANK() and DENSE_RANK() would assign the same rank to ties for the most recent date, which could yield multiple rows per customer. MAX() is an aggregate, not a row-identifier you can filter on to return a single per-CustomerID row without additional joins. So ROW_NUMBER() is the best fit to return the most recent row per CustomerID.

The idea is to assign a unique order to rows within each CustomerID so you can pick exactly one row—the most recent one. Using ROW_NUMBER() with an order by a timestamp column in descending order creates a per-CustomerID sequence like 1, 2, 3, and you can then take the row where the sequence equals 1 to get the most recent entry. RANK() and DENSE_RANK() would assign the same rank to ties for the most recent date, which could yield multiple rows per customer. MAX() is an aggregate, not a row-identifier you can filter on to return a single per-CustomerID row without additional joins. So ROW_NUMBER() is the best fit to return the most recent row per CustomerID.