在使用GROUP BY的时候，一般来说GROUP BY后跟分组列，但实际上在PG中可以在GROUP BY后跟表主键。 ``` [local]:5014 pgmaster@testdb=# drop table t; DROP TABLE Time: 39.739 ms [local]:5014 pgmaster@testdb=# create table t(id int primary key,name text,title text); CREATE TABLE Time: 59.905 ms [local]:5014 pgmaster@testdb=# insert into t values(1,'test1','test1'); INSERT 0 1 Time: 5.222 ms [local]:5014 pgmaster@testdb=# insert into t values(2,'test','test'); INSERT 0 1 Time: 5.223 ms [local]:5014 pgmaster@testdb=# insert into t values(3,'test','test'); INSERT 0 1 Time: 4.569 ms [local]:5014 pgmaster@testdb=# select name,count(*) from t group by id; name | count -------+------- test | 1 test | 1 test1 | 1 (3 rows) Time: 8.932 ms [local]:5014 pgmaster@testdb=# select name,count(*) from t group by name; name | count -------+------- test | 2 test1 | 1 (2 rows) Time: 0.906 ms [local]:5014 pgmaster@testdb=# select name,count(*) from t group by title; 2022-02-11 15:13:15.392 CST [1882] ERROR: column "t.name" must appear in the GROUP BY clause or be used in an aggregate function at character 8 2022-02-11 15:13:15.392 CST [1882] STATEMENT: select name,count(*) from t group by title; ERROR: column "t.name" must appear in the GROUP BY clause or be used in an aggregate function LINE 1: select name,count(*) from t group by title; ^ Time: 0.291 ms [local]:5014 pgmaster@testdb=# ``` 值得一提的是，在MySQL中，分组列不需要在GROUP BY中，在实际执行的时候MySQL会随机选取GROUP BY分组后该列的一个随机值返回。 **参考资料** [Functional Dependencies in SQL GROUP BY](https://blog.jooq.org/functional-dependencies-in-sql-group-by/) [group-by-handling](https://docs.oracle.com/cd/E17952_01/mysql-8.0-en/group-by-handling.html)