1. PARTITION BY creates separate "groups" within the dataset, similar to GROUP BY. 2. It does NOT collapse rows like GROUP BY, keeping all original data visible. 3. Can be used with aggregate ...

[Price Rank] = ROW_NUMBER() OVER(ORDER BY A.ListPrice DESC), [Category Price Rank] = ROW_NUMBER() OVER(PARTITION BY C.Name ORDER BY A.ListPrice DESC) [Price Rank ...