Question

这应该相当简单，但我找不到我需要的确切语法。我有一个具有name1和name2字段的表。我只想要一个字段中所有唯一值的列表以及每个值的计数。例如：

name1       name2
-----------------
smith       jones
jones       williams
evans       williams
elliot      james
williams    smith

我想要：

name     count
--------------
smith    2
jones    2
Williams 3
evans    1
Elliot   1
james    1

要获取我使用联合查询的唯一名称：

select name1 as myname
  from myTable
 union
select name2 as myname
  from myTable

但我无法找到计数所在位置的正确语法。有人可以把我从痛苦中解救出来吗？

Answer 1

您应该使用UNION ALL而不是UNION，因为UNION将删除所有重复项，而UNION ALL则不会。然后，您可以将union查询用作子查询，并执行计数：

SELECT name, COUNT(*)
FROM
  (SELECT name1 name FROM tablename
   UNION ALL
   SELECT name2 name FROM tablename) s
GROUP BY
  name

请参阅小提琴here。

Answer 2

您可以使用union all：

select name, count(*) as cnt
from ((select name1 as name from t) union all
      (select name2 from t)
     ) t
group by name;

使用union all代替union非常重要。后者删除了你不想要的重复项。

对于此类较大的查询，使用unpivot或以下构造更有效：

select (case when n.n = 1 then name1 else name2 end) as name, count(*)
from t cross join
     (select 1 as n union all select 2) n
group by (case when n.n = 1 then name1 else name2 end)

union all查询将扫描表格两次。此版本通常只扫描一次表。

Answer 3

select t.name,count(t.name) count
from 
(
select name1 as name from tbl
union all 
select name2 as name from tbl
)t
group by t.name

<强> SQL Fiddle

Answer 4

解决此类问题的另一种典型方法是使用和语法：

-- Put here whatever entangled query you have and mark it "with"
with myQuery as 
  (select name1 as myName
     from myTable
    union all
    select name2 as myName
      from myTable)

-- And now do everything you want with myQuery, as if it's an ordinary table
  select myName as Name,
         Count(*)
    from myQuery
group by myName

将count与union查询一起使用

4 个答案: