Sql 相关子查询的结果不正确

我的任务是： 查找只在同一年级有朋友的学生的姓名和年级

我的桌子：

高中生 朋友 我的问题是：

select name, grade from highschooler
where id in
(  select distinct id1 
   from friend out
   where not exists
     (select id1
      from friend
      where    
          (select grade from highschooler where id = out.id1) <>
          (select grade from highschooler where id = out.id2)
     )
)

预期查询结果：

Jordan  9
Brittany    10
Haley   10
Kris    10
Gabriel 11
John    12
Logan   12

我与Tiffany一起仔细检查我的代码：

1381    Tiffany 9   1247    Alexis  11
1381    Tiffany 9   1510    Jordan  9

我的相关子查询有问题。

---

你能给我一个提示吗？

要让学生只在自己的年级有朋友，而没有其他朋友，请尝试以下方法：

SELECT DISTINCT H.Id, H.Name, H.Grade
FROM HighSchooler H
  JOIN Friend F ON H.Id = F.ID1
  JOIN HighSchooler H2 ON F.ID2 = H2.ID AND H2.Grade = H.Grade
WHERE NOT EXISTS 
 (SELECT *
  FROM HighSchooler H3
  JOIN Friend F2 ON H3.Id = F2.Id1
  JOIN HighSchooler H4 ON F2.Id2 = H4.Id
  WHERE H4.Grade <> H.Grade AND H3.Id = H.Id)

这是我的答案

---

祝你好运。

我会使用TDQD——测试驱动的查询设计——来解决这个问题

确定某个高中生朋友的分数 这列出了每个朋友的高中生的朋友、高中生的成绩和朋友的成绩

生成一个学生列表，列出他们的朋友不在自己的年级 生成至少有一个朋友但没有不在其年级的朋友的学生列表 这将选择出现在朋友表id1列中的高中生；他们至少列出了一个朋友，而且他们的每个朋友都在同一年级

生成一个没有不在其年级的朋友的学生列表 这将选择没有出现在好友表id1列中的高中生；他们没有列出任何朋友，也没有输入任何数据，而是非常孤独，因此他们的朋友中没有一个处于不同的年级

注意，所有这些查询都假设学生B不一定因为学生A将学生B列为朋友而将学生A列为朋友。也就是说，如果条目1234、2345出现在朋友表中，说明学生1234认为2345是朋友，则它不会自动跟随2345认为1234是朋友；必须有一个明确的逆向条目23451234来表示互惠关系。如果友谊被自动回复，查询会变得更加复杂。

这是有效的：

select "name", "grade" from highschooler a
where not exists
(
  select 1 from (
(select id1, id2, count(distinct "grade") from (
SELECT *
FROM friend f, highschooler h
where f.id1 = h.id
or f.id2 = h.id
order by id1,id2)
group by id1, id2
having count(distinct "grade") > 1))
  where id1 = a.id or id2 = a.id)
order by "grade", "name"

---

这肯定有效，我已经试过运行它

Select distinct h1.name,h1.grade  
From Highschooler h1,Highschooler h2,Friend
Where h1.ID = Friend.ID1
AND h2.ID = Friend.ID2
AND h1.grade = h2.grade

EXCEPT
Select distinct h1.name,h1.grade
From Highschooler h1,Highschooler h2,Friend
Where h1.ID = Friend.ID1
AND h2.ID = Friend.ID2
AND h1.grade <> h2.grade
order by h1.grade,h1.name

---

非常感谢。但不幸的是，这不是我需要的。这将不会返回正确的结果。请查看编辑-了解不存在-将WHERE H.ID not IN更改为WHERE not EXISTS，这应该会做同样的事情。让我创建一个测试小提琴…不，你的查询返回所有高中生。不幸的是，没有工作。好的-请看编辑-这里是小提琴：工作。但我不明白我的变种有什么问题。

SELECT DISTINCT H.Id, H.Name, H.Grade
FROM HighSchooler H
  JOIN Friend F ON H.Id = F.ID1
  JOIN HighSchooler H2 ON F.ID2 = H2.ID AND H2.Grade = H.Grade
WHERE NOT EXISTS 
 (SELECT *
  FROM HighSchooler H3
  JOIN Friend F2 ON H3.Id = F2.Id1
  JOIN HighSchooler H4 ON F2.Id2 = H4.Id
  WHERE H4.Grade <> H.Grade AND H3.Id = H.Id)

SELECT f.id1 AS h1_id, h1.grade AS h1_grade, h2.grade AS h2_grade
  FROM Friends AS f
  JOIN Highschoolers AS h1 ON f.id1 = h1.id
  JOIN Highschoolers AS h2 ON f.id2 = h2.id

SELECT DISTINCT h1_id
  FROM (SELECT f.id1 AS h1_id, h1.grade AS h1_grade, h2.grade AS h2_grade
          FROM Friends AS f
          JOIN Highschoolers AS h1 ON f.id1 = h1.id
          JOIN Highschoolers AS h2 ON f.id2 = h2.id
       )
 WHERE h1_grade != h2_grade;

SELECT h.id, h.name, h.grade
  FROM Highschooolers AS h
  JOIN Friends AS f ON h.id = f.id1
 WHERE h.id NOT IN
       (SELECT DISTINCT h1_id
          FROM (SELECT f.id1 AS h1_id, h1.grade AS h1_grade, h2.grade AS h2_grade
                  FROM Friends AS f
                  JOIN Highschoolers AS h1 ON f.id1 = h1.id
                  JOIN Highschoolers AS h2 ON f.id2 = h2.id
               )
         WHERE h1_grade != h2_grade
       )

SELECT h.id, h.name, h.grade
  FROM Highschooolers AS h
 WHERE h.id NOT IN
       (SELECT DISTINCT h1_id
          FROM (SELECT f.id1 AS h1_id, h1.grade AS h1_grade, h2.grade AS h2_grade
                  FROM Friends AS f
                  JOIN Highschoolers AS h1 ON f.id1 = h1.id
                  JOIN Highschoolers AS h2 ON f.id2 = h2.id
               )
         WHERE h1_grade != h2_grade
       )

select "name", "grade" from highschooler a
where not exists
(
  select 1 from (
(select id1, id2, count(distinct "grade") from (
SELECT *
FROM friend f, highschooler h
where f.id1 = h.id
or f.id2 = h.id
order by id1,id2)
group by id1, id2
having count(distinct "grade") > 1))
  where id1 = a.id or id2 = a.id)
order by "grade", "name"

Select distinct h1.name,h1.grade  
From Highschooler h1,Highschooler h2,Friend
Where h1.ID = Friend.ID1
AND h2.ID = Friend.ID2
AND h1.grade = h2.grade

EXCEPT
Select distinct h1.name,h1.grade
From Highschooler h1,Highschooler h2,Friend
Where h1.ID = Friend.ID1
AND h2.ID = Friend.ID2
AND h1.grade <> h2.grade
order by h1.grade,h1.name