Question

See the following SQL:

SELECT M.username, count(*) as Total, date(status_date) as Date FROM com_result
   LEFT JOIN members as M on M.member_id = com_result.member_id 
GROUP BY date(status_date), com_result.member_id  Order By status_date DESC

This will show total numbers of records base on DATE and member_id

Example Result:

+----------+-------+------------+
| username | Total | DATE       |
+----------+-------+------------+
| bx7      |     3 | 2012-09-10 |
| bx2      |    25 | 2012-09-04 |
| bx2      |   401 | 2012-09-03 |
| bx1      |   703 | 2012-09-02 |
| bx4      |  1075 | 2012-09-02 |
+----------+-------+------------+

It work well, now I want to merge/union to another same and its same table structure. Second SQL:

SELECT M.username, count(*) as Total, date(status_date) as Date FROM com_result_b
   LEFT JOIN members as M on M.member_id = com_result_b.member_id 
GROUP BY date(status_date), com_result_b.member_id  Order By status_date DESC

Assume the result from com_result would be:

| bx2   |    25 | 2012-09-04 |

And the result from com_result_b would be:

 | bx2   |    50 | 2012-09-04 |

So when it merged/union – I want the result be like:

 | bx2   |    75 | 2012-09-04 |

As you can see the value of Total increased from same Date from two tables. How can that be done?

Answer 1

With a SUM

select username,SUM(Total) , DATE
(
    SELECT M.username, count(*) as Total, date(status_date) as Date FROM com_result_b 
       LEFT JOIN members as M on M.member_id = com_result.member_id  
    GROUP BY date(status_date), com_result.member_id  
   UNION ALL
   SELECT M.username, count(*) as Total, date(status_date) as Date FROM com_result_b   
      LEFT JOIN members as M on M.member_id = com_result_b.member_id    
   GROUP BY date(status_date), com_result_b.member_id 
) v

group by username, date