SQL返回低分低于平均低分的数据？答案

【问题标题】：SQL return data where low score is less than average low score?SQL返回低分低于平均低分的数据？
【发布时间】：2018-03-17 15:08:18
【问题描述】：

我有一个关于家庭作业的问题，我需要列出那些低分低于平均低分的测验的测验编号和日期。

为此，我需要使用 GROUP BY 子句和 HAVING 子句。它应该看起来像这样：

select quiznum, quizdate
from quizzes
group by quiznum, quizdate, lowscore 
having lowscore < avg(lowscore);

然而，这会返回未找到的数据。

如果我手动输入平均分（大约为 14），它会重新计算我需要的信息，但关于这个问题，我不知道平均低分。

如何列出低分低于平均低分的测验编号和日期？

【问题讨论】：

提示：HAVING 子句中需要子查询，外部查询中不需要聚合。
发布带有示例数据和所需输出的 DDL
作业认为什么是“低分”？测验表中是否有“分数”列？您是否需要首先找到低分，或者此表是否仅包含名为“lowscore”的列，而您只需要找到低于 lowscore 列中值平均值的分数？请解释一下，因为如果要帮助您解答这些难题，这些部分很重要。

标签： sql group-by having

【解决方案1】：

我想这个查询可能对你有用。

您需要创建一个子查询来获取avg 的分数。然后join self 并进行比较。

SELECT T2.* 
FROM 
(
    SELECT quiznum, 
       quizdate,
       avg(lowscore) as avgScore
    FROM quizzes
    group by quiznum, quizdate
) T1 INNER JOIN quizzes T2 
ON 
    T1.quiznum = T2.quiznum 
AND 
    T1.quizdate = T2.quizdate
Where T2.lowscore < T1.avgScore

【讨论】：

【解决方案2】：

在这种情况下，我发现使用分析窗口函数来完成所需的输出更容易。在我的查询中获得所需输出的关键是：

AVG(score) OVER (PARTITION BY quiznum, quizdate) AS avg_score

完成这项工作的查询如下。我正在使用 WITH 子句来创建虚假数据，但不要让您感到困惑。 BASE 查询是您所需要的。 WITH 子句只是在这里创建一个包含要处理的数据的假表。

WITH demo_data AS
(
  SELECT    1001     AS quiznum,    '10-JAN-18'  AS quizdate,   91   AS score    FROM dual UNION ALL
  SELECT    1001     AS quiznum,    '10-JAN-18'  AS quizdate,   75   AS score    FROM dual UNION ALL
  SELECT    1001     AS quiznum,    '10-JAN-18'  AS quizdate,   35   AS score    FROM dual UNION ALL
  SELECT    1001     AS quiznum,    '10-JAN-18'  AS quizdate,   40   AS score    FROM dual UNION ALL
  SELECT    1001     AS quiznum,    '10-JAN-18'  AS quizdate,   70   AS score    FROM dual UNION ALL
  SELECT    1001     AS quiznum,    '10-JAN-18'  AS quizdate,   85   AS score    FROM dual UNION ALL
  SELECT    1002     AS quiznum,    '17-JAN-18'  AS quizdate,   35   AS score    FROM dual UNION ALL
  SELECT    1002     AS quiznum,    '17-JAN-18'  AS quizdate,   40   AS score    FROM dual UNION ALL
  SELECT    1002     AS quiznum,    '17-JAN-18'  AS quizdate,   70   AS score    FROM dual UNION ALL
  SELECT    1002     AS quiznum,    '17-JAN-18'  AS quizdate,   85   AS score    FROM dual
)
SELECT quiznum, quizdate, score, avg_score
FROM
(
  SELECT quiznum, quizdate, score, AVG(score) OVER (PARTITION BY quiznum, quizdate) AS avg_score
  FROM demo_data
)
WHERE score < avg_score
;

下图按测验编号和日期显示了所有数据，并告诉我们平均分数。我们可以通过用外部 Select 包装它来使用它，这样我们就可以在 WHERE 子句中添加一个过滤器，以便将分数与平均分数进行比较。

下图显示了在 WHERE 子句中添加过滤器后的样子：

【讨论】：