Question

说我有一张像这样的表

ID | FieldA | FieldB | FieldC | GroupID
---------------------------------------
 1 |     50 |    NULL|     80 |      1
 2 |   NULL |     60 |   NULL |      1
 3 |     70 |    NULL|     50 |      1

我想生成一个按降序显示非NULL值的行（按ID）所以结果表看起来像这样

ID | FieldA | FieldB | FieldC | GroupID
---------------------------------------
 3 |     70 |      60|     50 |       1

有没有办法在MySQL中执行此操作？我试过了：

SELECT * 
FROM (table)
GROUP BY GroupID
ORDER BY ID DESC

但没有运气（当然）

这是一个小提琴，所以你可以快速玩这个！

http://sqlfiddle.com/#!2/71846f/2

谢谢！

Answer 1

如果没有每列的子查询，这可能是不可能的：

SELECT 
(SELECT `ID` FROM `table1` ORDER BY `ID` DESC LIMIT 1) AS `ID`,
(SELECT `FieldA` FROM `table1` WHERE `FieldA` IS NOT NULL ORDER BY `ID` DESC LIMIT 1) AS `FieldA`,
(SELECT `FieldB` FROM `table1` WHERE `FieldB` IS NOT NULL ORDER BY `ID` DESC LIMIT 1) AS `FieldB`,
(SELECT `FieldC` FROM `table1` WHERE `FieldC` IS NOT NULL ORDER BY `ID` DESC LIMIT 1) AS `FieldC`,
(SELECT `GroupID` FROM `table1` WHERE `GroupID` IS NOT NULL ORDER BY `ID` DESC LIMIT 1) AS `GroupID`

Answer 2

您应该检查您的字段，如下所示：

SELECT * 
FROM table
WHERE FieldA is not null
AND   FieldB is not null
AND   FieldC is not null
ORDER BY ID DESC

Answer 3

select MAX(feA),
       MAX(feB),
       MAX(feC) 
from test05

Answer 4

仅按组ID分组，然后选择所有列的最大值。。哪个会有所帮助

SELECT max(id),max(FIELDA),max(FIELDB),max(FIELDC),GroupID
FROM (table)
GROUP BY GroupID

Answer 5

如果您想要FIELD的最小值，那么您只需要这个。此查询未经过测试如果此查询出现任何错误，请回复

SELECT max(id),max(FIELDA),max(FIELDB),min(FIELDC),GroupID
FROM (test)
GROUP BY GroupID

Answer 6

您可以创建列名为1-20的CTE（将原始数据复制到其中）

然后您将选择每一列（max（列））并将其更新为临时表。

（将原始表复制到temp并截断它。）

使用column_number和dynamic sql上的循环执行此操作。

我希望你明白我的意思 - ：）

如何在省略空值的同时对行进行分组？

6 个答案: