SQL将行移动为列-PIVOT查询[重复]答案

【问题标题】：SQL Moving a Row to be a column - PIVOT query [duplicate]SQL将行移动为列-PIVOT查询[重复]
【发布时间】：2018-01-21 07:07:13
【问题描述】：

给定：

+-----------+---------------+-------------+-------+
|   Name    |   Location    | Description | Value |
+-----------+---------------+-------------+-------+
| Company A | Houston       | Sales       |   100 |
| Company A | Houston       | Profit      |    50 |
| Company B | San Francisco | Sales       |   500 |
| Company B | San Francisco | Profit      |   200 |
| Company C | New York      | Sales       |   200 |
| Company C | San Francisco | Profit      |   150 |
+-----------+---------------+-------------+-------+

如何使用 SQL 将其转换为：

+----------+---------------+--------+---------+
|   Name   |   Location    | Sales  | Profit  |
+----------+---------------+--------+---------+
| CompanyA | Houston       |    100 |      50 |
| CompanyB | San Francisco |    500 |     200 |
| CompanyC | New York      |    200 |     150 |
+----------+---------------+--------+---------+

【问题讨论】：

标签： sql sql-server

【解决方案1】：

您可以使用 PIVOT 函数将行值转换为列，但是它会提供稍微不同但更合理的输出 -

SELECT Name, Location, Sales, Profit
FROM
(
    SELECT *
    FROM Table1
) AS SourceTable
PIVOT
(
  MIN(value)
  FOR Description IN (Sales, Profit)
) AS PivotTable
ORDER BY Name, Location;

将生成以下输出 -

+-----------+---------------+-------+--------+
| Name      | Location      | Sales | Profit |
+-----------+---------------+-------+--------+
| Company A | Houston       | 100   | 50     |
+-----------+---------------+-------+--------+
| Company B | San Francisco | 500   | 200    |
+-----------+---------------+-------+--------+
| Company C | New York      | 200   | null   |
+-----------+---------------+-------+--------+
| Company C | San Francisco | null  | 150    |
+-----------+---------------+-------+--------+

您可以将其用作替代解决方案。

【讨论】：

【解决方案2】：

尝试 1 或 2 查询：http://sqlfiddle.com/#!6/bb33b/7
第二个给出了您问题中显示的确切结果，但我的猜测是第一个是您正在寻找的。p>

SELECT Name, Location,
       Sum( Case Description when 'Sales' Then Value else 0 end ) Sales,
       Sum( Case Description when 'Profit' Then Value else 0 end ) Profit
FROM table1
GROUP BY Name, Location
Order by 1,2
;

|      Name |      Location | Sales | Profit |
|-----------|---------------|-------|--------|
| Company A |       Houston |   100 |     50 |
| Company B | San Francisco |   500 |    200 |
| Company C |      New York |   200 |      0 |
| Company C | San Francisco |     0 |    150 |

SELECT Name, Min( Location ) As Location,
       Sum( Case Description when 'Sales' Then Value else 0 end ) Sales,
       Sum( Case Description when 'Profit' Then Value else 0 end ) Profit
FROM table1
GROUP BY Name
Order by 1,2
;
;

|      Name |      Location | Sales | Profit |
|-----------|---------------|-------|--------|
| Company A |       Houston |   100 |     50 |
| Company B | San Francisco |   500 |    200 |
| Company C |      New York |   200 |    150 |

【讨论】：

【解决方案3】：

要将您的数据转换为您想要的形式，您可以这样做：

 SELECT NAME
          ,LOCATION
          ,SUM (CASE WHEN DESCRIPTION = 'Sales' THEN Values ELSE 0 END) AS SALES
          ,SUM (CASE WHEN DESCRIPTION = 'Profit' THEN Values ELSE 0 END) AS PROFIT
 FROM yourtable
 GROUP BY NAME
               ,LOCATION

【讨论】：

【解决方案4】：

假设您的描述值只有Sales 和Profit，这是一个 CTE 解决方案。我调用了初始表Before。

with cteSales as (select name, value from Before where description = 'Sales')
, cteProfit as (select name, value from Before where description  = 'Profit')
select 
distinct 
    name
    , location
    , (select value from cteSales where name = B.name) as Sales
    , (select value from cteProfit where name = B.name) as Profit
from Before B

【讨论】：