Problem:
Table number
contains many numbers in column num including duplicated ones.
Can you write a sql query to find the biggest number,which only appears once.
+---+ |num| +---+ | 8 | | 8 | | 3 | | 3 | | 1 | | 4 | | 5 | | 6 |
For the sample data above,your query should return the following result:
+---+ |num| +---+ | 6 |
Note:
If there is no such number,just output null.
Solution:
这道题本不难,但引申出了一个很重要的问题,我第一次写的答案是
SELECT num FROM number GROUP BY num HAVING COUNT(*)=1 ORDER BY num DESC LIMIT 1;
但没有通过测试用例
1 Input: {"headers": {"number": ["num"]},"rows": {"number": [[8],[1],[8],[3],[4],[5],[6],[6]]}} 2 3 Output: {"headers":["num"],"values":[]} 4 5 Expected: {"headers":["num"],"values":[[null]]}
因此,若要输出null,需要满足两个条件,第一,select中应是聚合函数,第二,聚合函数处理的行数应为0。
Code:
SELECT MAX(num) AS num FROM (SELECT num FROM number GROUP BY num HAVING COUNT(*)=1) AS T;
版权声明:本文内容由互联网用户自发贡献,该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容, 请发送邮件至 dio@foxmail.com 举报,一经查实,本站将立刻删除。