Question

我有一个像调用temp1（Original Image）

的sql表

+ --------- + --------- + ------ + ---------- +
| ClientKey | LoginFrom | Month  | Tran_Count |
+ --------- + --------- + ------ + ---------- +
| 1         | E         | 201603 | 2          |
| 3         | A         | 201504 | 3          |
| 3         | X         | 201506 | 105        |
| 3         | A         | 201602 | 3          |
| 3         | X         | 201509 | 3          |
| 3         | A         | 201512 | 1          |
| 3         | W         | 201508 | 3          |
| 3         | W         | 201505 | 17         |
| 3         | W         | 201507 | 14         |
| 3         | W         | 201504 | 34         |
| 3         | W         | 201509 | 1          |
| 3         | X         | 201504 | 170        |
| 3         | W         | 201506 | 29         |
| 3         | W         | 201510 | 1          |
| 3         | X         | 201505 | 51         |
| 3         | X         | 201508 | 11         |
| 3         | A         | 201603 | 1          |
| 3         | X         | 201510 | 3          |
| 3         | X         | 201507 | 44         |
| 4         | W         | 201601 | 2          |
| 4         | E         | 201601 | 4          |
| 4         | A         | 201602 | 1          |
| 4         | E         | 201602 | 19         |
+ --------- + --------- + ------ + ---------- +

我想总结一下表（name_summary），如（Original Image）：

+ --------- + ---------------- + ---------------- + ---------------- + --- +
| ClientKey | A type in 201504 | A type in 201505 | W type in 201504 | ... |
+ --------- + ---------------- + ---------------- + ---------------- + --- +
| 3         | 3                |                  | 34               | ... |
+ --------- + ---------------- + ---------------- + ---------------- + --- +

我尝试过以下sql代码

 update master_summary
set
A_type_in_201504=case when temp1.LoginFrom='A' and temp1.[month]='201504' then temp1.tran_count end,
A_type_in_201505=case when temp1.LoginFrom='A' and temp1.[month]='201505' then temp1.tran_count end,
W_type_in_201504=case when temp1.LoginFrom='W' and temp1.[month]='201504' then temp1.tran_count end,
W_type_in_201505=case when temp1.LoginFrom='W' and temp1.[month]='201505' then temp1.tran_count end
from master_summary
inner join 
temp1
on
master_summary.clientkey=temp1.clientkey

我认为我错了。当我尝试使用多个条件进行单列更新时，它给出了正确的结果。但是多列更新具有不同的多重条件，它无法正常工作..

是谁能帮助我？

Answer 1

您需要使用max()进行汇总，并使用子查询加入汇总查询：

update master_summary
  set
    A_type_in_201504=s.A_type_in_201504
  , A_type_in_201505=s.A_type_in_201505
  , W_type_in_201504=s.W_type_in_201504
  , W_type_in_201505=s.W_type_in_201505
from master_summary
  inner join (
  select 
      clientkey,
    , A_type_in_201504=max(case 
        when temp1.LoginFrom='A' and temp1.[month]='201504' then temp1.tran_count end)
    , A_type_in_201505=max(case 
        when temp1.LoginFrom='A' and temp1.[month]='201505' then temp1.tran_count end)
    , W_type_in_201504=max(case 
        when temp1.LoginFrom='W' and temp1.[month]='201504' then temp1.tran_count end)
    , W_type_in_201505=max(case 
        when temp1.LoginFrom='W' and temp1.[month]='201505' then temp1.tran_count end)
  from temp1 
  group by clientkey
  ) s
    on master_summary.clientkey=s.clientkey

具有多个条件的多列更新

1 个答案: