问题
I am using sqlite database. My table schema is
CREATE TABLE performance(area TEXT, name TEXT, score INTEGER, dt TEXT)
The content in the table is like this:
uk|josh|4|2013-11-04 20:00
ca|josh|2|2013-11-05 20:00
us|josh|6|2013-11-05 20:00
uk|andy|5|2013-11-04 20:00
us|andy|1|2013-11-05 20:00
uk|sara|9|2013-11-05 20:00
ca|sara|7|2013-11-06 20:00
ca|sara|2|2013-11-06 20:00
I used the following sql statement to select name and its corresponding sum of score grouping by name and dt.
select name, sum(score), dt from performance group by name, dt;
I got
josh|4|2013-11-04 20:00
josh|8|2013-11-05 20:00
andy|5|2013-11-04 20:00
andy|1|2013-11-05 20:00
sara|9|2013-11-05 20:00
sara|9|2013-11-06 20:00
Now I want to expand my query so that the sql statement can search that whose sum of score didn't change at different time(dt). In this case the output should be like:
sara|9|2013-11-05 20:00
sara|9|2013-11-06 20:00
How to compose such a sql?
回答1:
This can be achieved by a self (anti) join:
SELECT a.*
FROM (SELECT name, dt, SUM(score) as sum_score
FROM performance
GROUP BY name, dt) a
JOIN (SELECT name, dt, SUM(score) as sum_score
FROM performance
GROUP BY name, dt) b
ON a.name = b.name AND a.sum_score = b.sum_score AND a.dt < b.dt
回答2:
select dt, name, sum(score), dt
from performance
group by name, dt
having count(*) > count(distinct score);
*Note: If you are grouping by dt, name
, I think you should display them also.
来源:https://stackoverflow.com/questions/19814654/how-to-compare-the-value-of-two-rows-with-sql