Question

我有一个查询，它返回不止一次出现的不同device_type的行数。

SELECT COUNT(*) AS C1,device_type FROM stat 
    WHERE stat_date = '2012-02-08' 
    GROUP BY 2 HAVING C1 > 1 
    ORDER BY 1 DESC

我想将剩余的（HAVING count = 1）行总结为'others'

如何将COUNT（*）和'others'的总和添加为以下查询的第二列？

SELECT COUNT(*) AS C2,device_type FROM stat 
    WHERE stat_date = '2012-02-08' 
    GROUP BY 2 HAVING C2 = 1 
    ORDER BY 1 DESC

DB中的示例数据

device_type
dt1
dt1
dt1
dt2
dt2
dt3
dt4
dt5

预期结果

3 dt1
2 dt2
3 other

Answer 1

我会这样做。

SELECT COUNT(*) AS C1,device_type FROM stat 
    WHERE stat_date = '2012-02-08' 
    GROUP BY 2 HAVING C1 > 1 
    ORDER BY 1 DESC
Union

SELECT Sum(1),'OTHERS'FROM stat 
    WHERE stat_date = '2012-02-08' 
    GROUP BY 2 HAVING C1 =1
    ORDER BY 1 DESC

Answer 2

您也可以尝试：

SELECT SUM(C1) AS C1, CASE WHEN C1 = 1 THEN 'other' ELSE device_type END as device_type
FROM (  SELECT  COUNT(*) AS C1,
                device_type 
        FROM stat 
        WHERE stat_date = '2012-02-08' 
        GROUP BY device_type) A
GROUP BY CASE WHEN C1 = 1 THEN 'other' ELSE device_type END

如何汇总仅出现一次的行？

2 个答案: