透视带有文本且没有聚合的表格 [英] Pivoting a Table with Text and no Aggregation

问题描述

我一直在查看 StackOverflow 问题，但还没有真正找到与此问题相关的问题.

I've been looking through StackOverflow questions and haven't really found one for this issue.

我有一张桌子，基本上是这样排列的:

I have a table that's arranged basically like:

Table.LineID
Table.QuestionGroup
Table.Question
Table.Answer

我想旋转"表格，但问题和答案没有任何要汇总的内容.例如，它们都只是文本字段，可能会这样写:

And I'd like to "pivot" the table, but the questions and answers don't have anything to aggregate. They're all just text fields for example, one might read:

Table.LineID = 00001
Table.QuestionGroup = Color
Table.Question = Outside Color
Table.Answer = Dark Green

Table.LineID = 00001
Table.QuestionGroup = Size
Table.Question = Height
Table.Answer = 180 3/4

我正在尝试旋转表格，以便我可以获得相关联的行 ID 并使传播函数成为问题所以它看起来像

I'm trying to pivot the table so that I can get the associated ID of line and make that the spreading function the Questions So it would look like

线路ID |问题 1 |问题 2 |等等……

LineID | Question 1 | Question 2 | Etc...

我看过这个 https://stackoverflow.com/a/7744347/839330，除了这似乎更多地参考一些VB代码(也许我错了?).有谁知道我应该如何处理这个问题?

I've looked at this https://stackoverflow.com/a/7744347/839330 except this seems to be more in reference to some VB code (maybe I'm wrong?). Does anyone have an idea on how I should approach this?

推荐答案

仅仅因为您的表中有文本数据并不意味着您不能对其使用聚合函数.

Just because you have text data in your table does not mean that you cannot use an aggregate function on it.

您没有指定您使用的是什么 RDBMS，但这种类型的数据转换是 pivot.这会将行数据转换为列.一些数据库有一个数据透视函数，但是如果你使用一个没有这个函数的数据库，那么你需要使用一个带有 CASE 表达式的聚合函数:

You did not specify what RDBMS you are using but this type of data transformation is a pivot. This converts row data into columns. Some databases has a pivot function but if you are using one without that function, then you will need to use an aggregate function with a CASE expression:

select lineid,
  max(case when question ='Height' then answer else '' end) Height,
  max(case when question ='Outside Color' then answer else '' end) [Outside Color]
from yourtable 
group by lineid

参见 SQL Fiddle with Demo

如果您有一个具有数据透视功能的数据库(SQL Server 2005+/Oracle 11g+)，那么您的代码将类似于:

If you have a database that has the pivot function (SQL Server 2005+/Oracle 11g+), then your code will be similar to this:

select *
from 
(
  select lineid,
    Question,
    Answer
  from yourtable
) src
pivot
(
  max(answer)
  for question in ([Height], [Outside Color])
) piv;

参见 SQL Fiddle with Demo

现在，如果您使用的是 SQL Server 2005+，并且您想将未知数量的问题转换为列，那么您可以使用类似于这样的动态 sql:

Now if you are using SQL Server 2005+ and you have an unknown number of questions that you want to turn into columns, then you can use dynamic sql similar to this:

DECLARE @cols AS NVARCHAR(MAX),
    @query  AS NVARCHAR(MAX)

select @cols = STUFF((SELECT distinct ',' + QUOTENAME(Question) 
                    from yourtable
            FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)') 
        ,1,1,'')

set @query = 'SELECT lineid,' + @cols + ' from 
             (
                select lineid,
                  Question,
                  Answer
                from yourtable
            ) x
            pivot 
            (
                max(Answer)
                for Question in (' + @cols + ')
            ) p '

execute(@query)

参见 SQL Fiddle 演示

根据您的示例数据，结果是:

Based on your sample data the result is:

| LINEID |  HEIGHT | OUTSIDE COLOR |
------------------------------------
|      1 | 180 3/4 |    Dark Green |

这篇关于透视带有文本且没有聚合的表格的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！