0

我有一个有 5 个列的表,看起来像这样(示例):

Conf_1   | Conf_2   | Conf_3   | Conf_4   | Conf_5
High       No Match   Perfect    No Match   Perfect
High       Perfect    Review     No Match   Perfect
No Match   No Match   Perfect    Medium     Low
Perfect    Medium     No Match   Low        Medium
Low        Low        Review     Medium     High
Medium     No Match   No Match   Perfect    High

这些值包括 Perfect、High、Medium、Low、No Match 和 Review。某些列包含所有值,而某些列仅包含一个或两个值。

我想计算每个值在每列中显示的次数 - 将表格转换为 6 个值(Perfect - Review)作为行,然后计算 5 个 Conf 字段中的每一个的计数。

所需的输出如下所示:

这里的置信度列是来自组合 conf_1,conf_2......conf_5 的不同值

Confidence | Conf_1 Count  | Conf_2 Count.....| Conf_5
Perfect      25               40               50 
High         25               Null             50
Medium       5                Null             50
Low          100              100              Null
No Match     4                5                Null
Review       Null             15               Null
4

1 回答 1

1
Select 
confidence,
count(case when Confidence = 'Perfect' then 1 else 0) end as conf_1,
count(case when Confidence = 'High' then 1 else 0) end as conf_2,
count(case when Confidence = 'Medium' then 1 else 0) end as conf_3,
count(case when Confidence = 'Low' then 1 else 0) end as conf_4,
count(case when Confidence = 'No Match' then 1 else 0) end as conf_5,
count(case when Confidence = 'Review' then 1 else 0) end as conf_6

from table
group by 1

您可以使用以下方法建立信心:

select * from (
select conf_1 as confidence from table
union 
select conf_2 as col from table
union
select conf_3 as col from table
union
select conf_4 as col from table
union 
select conf_5 as col from table
union
select conf_6 as col from table

) t where confidence is not null
order by confidence
于 2021-08-30T18:55:22.060 回答