我有一个简单的查询:
SELECT sds.district_id,detail.year, detail.race, SUM(count)
FROM school_data_race_ethnicity_raw as detail
INNER JOIN school_data_schools as sds USING (school_id)
GROUP BY district_id, year, race
示例结果集:
| 68080104 | 2009 | Multiracial | 0 |
| 68080104 | 2009 | White | 847 |
| 68080104 | 2010 | American Indian | 1 |
| 68080104 | 2010 | Asian | 4 |
| 68080104 | 2010 | Black | 17 |
| 68080104 | 2010 | Hispanic | 4 |
| 68080104 | 2010 | Multiracial | 2 |
| 68080104 | 2010 | White | 823 |
| 68080104 | 2011 | American Indian | 4 |
| 68080104 | 2011 | Asian | 4 |
| 68080104 | 2011 | Black | 9 |
| 68080104 | 2011 | Hispanic | 10 |
| 68080104 | 2011 | Multiracial | 24 |
| 68080104 | 2011 | White | 767 |
+-------------+------+-----------------+------------+
我想添加一个名为total的第5列,显示给定年份和地区总人口的总和。例如,如果我在2011年的区域68080104,总数将是(4 + 4 + 9 + 10 + 24 + 767)。我需要这个作为此查询中的另一列。它也需要快速。 (10秒以下)。我正在努力如何做到这一点,而不是妥协速度和数据。
答案 0 :(得分:2)
您需要为其创建单独的查询并将其与原始查询相关联。试试这个,
SELECT a.*, b.totalCount
FROM
(
SELECT sds.district_id,detail.year, detail.race, SUM(count)
FROM school_data_race_ethnicity_raw as detail
INNER JOIN school_data_schools as sds
USING (school_id)
GROUP BY district_id, year, race
) a INNER JOIN
(
SELECT sds.district_id,detail.year, SUM(count) totalCount
FROM school_data_race_ethnicity_raw as detail
INNER JOIN school_data_schools as sds
USING (school_id)
GROUP BY district_id, year
) b ON a.district_id = b.district_id AND
a.year = b.year
答案 1 :(得分:1)
使用WITH ROLLUP
SELECT sds.district_id,detail.year, detail.race, SUM(count)
FROM school_data_race_ethnicity_raw as detail
INNER JOIN school_data_schools as sds USING (school_id)
GROUP BY district_id, year, race WITH ROLLUP
答案 2 :(得分:0)
在MySQL中,要获取同一行的数据,您几乎必须将其作为连接:
select t.*, t2.cnt as TotalDistrictYear
from (SELECT sds.district_id,detail.year, detail.race, SUM(count) as cnt
FROM school_data_race_ethnicity_raw as detail INNER JOIN
school_data_schools as sds USING (school_id)
GROUP BY district_id, year, race
) t join
(SELECT sds.district_id,detail.year,SUM(count) as cnt
FROM school_data_race_ethnicity_raw as detail INNER JOIN
school_data_schools as sds USING (school_id)
GROUP BY district_id, year
) t2
on t.district_id = t2.district_id and
t.year = t2.year