Let's say we want to get the count of students in each unique combination of a grade and a course.
For example, we want to count number of students enrolled in Java in grade 1, Java in grade 2, JavaScript in grade 2, JavaScript in grade 3, and so on for all possible unique combinations of course and grade, with at least one student in it.
When there are multiple columns in a GROUP BY clause, each unique combination of values in all those columns creates a separate group, over which the aggregate function is applied.
In the above data and the query that follows it, course Java and grade 1 is a unique group, which has 2 students (Harding and Tonya) under it. Similarly, course JavaScript and grade 6 is another unique group with just 1 student (Johnson) under it. course JavaScript also combines with grade 9 to create another unique group, with 1 student (Tony) under it.