问题
In SQL Server 2012, I have a table my_table that has columns state, month, ID, and sales.
My goal is to merge different rows that have the same state, month, ID into one row while summing the sales column of these selected rows into the merged row.
For example:
state month ID sales
-------------------------------
FL June 0001 12,000
FL June 0001 6,000
FL June 0001 3,000
FL July 0001 6,000
FL July 0001 4,000
TX January 0050 1,000
MI April 0032 5,000
MI April 0032 8,000
CA April 0032 2,000
This what I am supposed to get
state month ID sales
-------------------------------
FL June 0001 21,000
FL July 0001 10,000
TX January 0050 1,000
MI April 0032 13,000
CA April 0032 2,000
I did some research, and I found that the self join is supposed to do something similar to what I am supposed to get.
回答1:
Unless I am missing something in the requirements, why not just use an aggregate function with a GROUP BY:
select state, month, id, sum(sales) Total
from yourtable
group by state, month, id
order by id
See SQL Fiddle with Demo
The result is:
| STATE | MONTH | ID | TOTAL |
--------------------------------
| FL | July | 1 | 10000 |
| FL | June | 1 | 21000 |
| CA | April | 32 | 2000 |
| MI | April | 32 | 13000 |
| TX | January | 50 | 1000 |
回答2:
Considering there should be an index on column id, this query would be a better solution:
select state, month, id, sum(sales) Total
from yourtable
group by id, state, month
order by id
来源：https://stackoverflow.com/questions/13940397/sum-values-from-multiple-rows-into-one-row