假设我有一个像这样的表:
+---+---+---+---+---+ |G1 |G2 |G3 |G4 | V | +---+---+---+---+---+ | A | B | C | D | 2 | | E | F | G | H | 4 | | C | D | A | B | 2 | | E | F | G | H | 3 | | E | I | G | L | 7 | +---+---+---+---+---+
其中G1、G3是VARCHAR类型,G2、G4是INT类型
如果我在G1..G4上进行简单的GROUP BY,我会得到:
SELECT G1,G2,G3,G4,SUM(V) as V FROM Table GROUP BY G1, G2, G3, G4; +---+---+---+---+---+ |G1 |G2 |G3 |G4 | V | +---+---+---+---+---+ | A | B | C | D | 2 | | E | F | G | H | 7 | | C | D | A | B | 2 | | E | I | G | L | 7 | +---+---+---+---+---+
我想知道是否可能在反转的值G1、G2 <-> G3、G4上进行聚合。 我想要的结果是这样的:
+---+---+---+---+---+ +---+---+---+---+---+ |G1 |G2 |G3 |G4 | V | |G1 |G2 |G3 |G4 | V | +---+---+---+---+---+ +---+---+---+---+---+ | A | B | C | D | 4 | 或 | C | D | A | B | 4 | | E | F | G | H | 7 | | E | F | G | H | 7 | | E | I | G | L | 7 | | E | I | G | L | 7 | +---+---+---+---+---+ +---+---+---+---+---+
我尝试获取反转的行,但由于重复,我仍然无法在它们上进行聚合。我的测试查询是:
SELECT DISTINCT * FROM Table t1 JOIN Table t2 on t1.G1 = t2.G3 and t1.G2=t2.G4 and t1.G3=t2.G1 and t1.G4=t2.G2 +---+---+---+---+---+---+---+---+---+---+ |G1 |G2 |G3 |G4 | V |G1 |G2 |G3 |G4 | V | +---+---+---+---+---+---+---+---+---+---+ | A | B | C | D | 2 | C | D | A | B | 2 | | C | D | A | B | 2 | A | B | C | D | 2 | +---+---+---+---+---+---+---+---+---+---+
如果我理解正确,将列的顺序反转,使得 A B C D 对应 C D A B。
db<>fiddle