SQL MAX Function 行为无分组依据

Question

Consider the following table:

CREATE TABLE test (
  id INT
);
INSERT INTO test (id) VALUES (1);
INSERT INTO test (id) VALUES (2);

If I run the query:

SELECT MAX(id) as MaxId FROM test WHERE id > 10;

I get a single row having a value of null for MaxId

However, if I run the following:

SELECT MAX(id) as MaxId FROM test WHERE id > 10 GROUP BY id;

I get no rows returned.

Since the WHERE condition in both scenario are same, no rows are selected in either case. My understanding is, the Group By in the second case has no significance as no rows are returned.

Could someone clarify why I get a row with null value in first query while no rows returned in second query?

Answer 1

An aggregation query with no GROUP BY always returns one row (well, unless the row is filtered out by a HAVING clause).

If there are no rows in the table, or the WHERE clause filters out all rows, then the aggregation results are one row with a NULL value.

That is what happens in your first query.

Your second query has a GROUP BY . It returns one row per group. But there are no groups, so the query returns no rows.