Question

我有一个包含许多列和行的表，并且一列下有组，另一列下有人，每组中可以有一个或多个人。我想做这样的事情（伪代码）：

if there is more than one instance of that group showing up:
    iterate through it and count how many members in that group have family members, then insert into a table
else:
    just take out that member and count how many family members he has and insert into a table

我来自python背景，SQL对我来说仍然是一个新手，但我的（非常糟糕）尝试：

if (select count(groupname) from table group by groupname) > 1 
then _________
else
insert___ into ___

表A看起来有点像这样：

groups                   people
A                        tom
A                        jerry
B                        sarah

表B如下所示：

person                   familyMembers
tom                        daughter
tom                        son
tom                        wife
sarah                      husband

汤姆有3个家庭成员，莎拉只有1个

有点失落，见解？

Answer 1

有点不清楚该组织如何发挥作用。听起来你想要计算每个人的家庭成员数量，对吗？一种简单的方法是在tableB中计算族：

SELECT
  person,
  COUNT(familyMembers) AS count_familyMembers
FROM tableB
GROUP BY person

这忽略了＆＃34; group＆＃34;列一起，但如果他们没有在tableB中有记录（例如你的例子中的Jerry），将会错过任何人。

要包括Jerry，您需要执行以下操作。这将检查familyMembers是否为NULL并返回0，否则将返回1.然后将所有这些加起来，这样就可以得到记录总数，按组和人员列分开。

SELECT
  groups,
  people,
  SUM(CASE
    WHEN familyMembers IS NULL THEN 0
    ELSE 1
  END) AS count_familyMembers
FROM tableA
LEFT JOIN tableB
  ON tableA.people = tableB.person
GROUP BY groups,
         people

如果上述方法有效，您可以添加＆＃34; INTO＆＃34;创建一个新表：

SELECT
  groups,
  people,
  SUM(CASE
    WHEN familyMembers IS NULL THEN 0
    ELSE 1
  END) AS count_familyMembers INTO new_table
FROM tableA
LEFT JOIN tableB
  ON tableA.people = tableB.person
GROUP BY groups,
         people

或＆＃34; INSERT INTO＆＃34;对于现有表格。如果您执行了＃34; INSERT INTO＆＃34;，则需要匹配数据和列：

INSERT INTO existingTable (groups, people, familyCount)
  SELECT
    groups,
    people,
    SUM(CASE
      WHEN familyMembers IS NULL THEN 0
      ELSE 1
    END) AS count_familyMembers INTO new_table
  FROM tableA
  LEFT JOIN tableB
    ON tableA.people = tableB.person
  GROUP BY groups,
           people

迭代行SQL + Count

1 个答案: