MySql 未知列数的主元 headers

Question

我有一个 MySql 查询，我想动态转换它，公司名称作为列 headers，公司财务字段作为行 headers。

目前我有这个：

SELECT Company, TotalRevenue, Overhead, TotalJobCosts, GrossProfit 
FROM comp_financials

这给了我这个：

Company	TotalRevenue	Overhead	TotalJobCosts	GrossProfit
Chicago's Best Construction	2098001	363750	1424420	673581
Jones Construction	4509458	1067008	2876568	1632890

我希望得到这样的东西：

	Chicago's Best Construction	Jones Construction
TotalRevenue	2098001	4509458
Overhead	363750	1067008
TotalJobCosts	1424420	2876568
GrossProfit	673581	1632890

公司数量及其价值未知（但通常少于 20 家）。

到目前为止，我已经能够动态地将公司名称显示为列 headers，但无法了解如何显示行值（不需要行 headers）。

SET @@group_concat_max_len = 32000;
SET @sql = NULL;
SELECT  GROUP_CONCAT(DISTINCT
        CONCAT('MAX(CASE WHEN Company = ''',
               Company,
               ''' THEN Company ELSE NULL END) AS ',
               CONCAT('`', Company, '`')
               )) INTO @sql
FROM comp_financials;

SET @sql = CONCAT('SELECT ', @sql, ' 
                   FROM comp_financials');

PREPARE stmt FROM @sql;
EXECUTE stmt;
DEALLOCATE PREPARE stmt;

Answer 1

首先要做的是将列变成行。方法简单；只需要使用 UNION ALL。像这样：

SELECT 1 as r,company, "TotalRevenue" val, TotalRevenue totals FROM comp_financials UNION ALL
SELECT 2, company, "Overhead", Overhead FROM comp_financials UNION ALL
SELECT 3, company, "TotalJobCosts", TotalJobCosts FROM comp_financials UNION ALL
SELECT 4, company, "GrossProfit", GrossProfit FROM comp_financials

我添加了编号 (.. as r) 供以后在 ORDER BY 中使用。 val 值对应于我从中检索数据的列名。此查询将 return 如下数据：

r	company	val	totals
1	Jones Construction	TotalRevenue	4509458
1	Chicago's Best Construction	TotalRevenue	2098001
2	Jones Construction	Overhead	1067008
2	Chicago's Best Construction	Overhead	363750
3	Jones Construction	TotalJobCosts	2876568
3	Chicago's Best Construction	TotalJobCosts	1424420
4	Jones Construction	GrossProfit	1632890
4	Chicago's Best Construction	GrossProfit	673581

我将上面的查询作为您正在执行的原始查询的子查询。所以最终结果是这样的：

SELECT r,val,
       MAX(CASE WHEN company="Jones construction" THEN totals END) AS "Jones construction",
       MAX(CASE WHEN company="Chicago's Best Construction" THEN totals END) AS "Chicago's Best Construction"
FROM
(SELECT 1 as r,company, "TotalRevenue" val, TotalRevenue totals FROM comp_financials UNION ALL
SELECT 2, company, "Overhead", Overhead FROM comp_financials UNION ALL
SELECT 3, company, "TotalJobCosts", TotalJobCosts FROM comp_financials UNION ALL
SELECT 4, company, "GrossProfit", GrossProfit FROM comp_financials) B
GROUP BY r,val
ORDER BY r ASC

此外，我按 r 和 val 分组，然后按 r 排序。此查询 returns 数据如下：

r	val	Jones construction	Chicago's Best Construction
1	TotalRevenue	4509458	2098001
2	Overhead	1067008	363750
3	TotalJobCosts	2876568	1424420
4	GrossProfit	1632890	673581

将此添加到您准备好的声明中：

SET @@group_concat_max_len = 32000;
SET @sql = NULL;
SELECT  GROUP_CONCAT(DISTINCT
        CONCAT('MAX(CASE WHEN Company = "',
               Company,
               '" THEN totals ELSE NULL END) AS ',
               CONCAT('"', Company, '"')
               )) INTO @sql
FROM comp_financials;

SET @sql = CONCAT('SELECT r,val, ', @sql, '
   FROM
(SELECT 1 as r,company, "TotalRevenue" val, TotalRevenue totals FROM comp_financials UNION ALL
SELECT 2, company, "Overhead", Overhead FROM comp_financials UNION ALL
SELECT 3, company, "TotalJobCosts", TotalJobCosts FROM comp_financials UNION ALL
SELECT 4, company, "GrossProfit", GrossProfit FROM comp_financials) B
GROUP BY r,val
ORDER BY r ASC');
select @sql;

PREPARE stmt FROM @sql;
EXECUTE stmt;
DEALLOCATE PREPARE stmt;

这是一个演示 fiddle：https://dbfiddle.uk/?rdbms=mysql_5.7&fiddle=678de8991e21ccd90d2195d3d5eca0a7

P/S: 下次请不要post数据表或任何我们可以复制和粘贴的东西（代码）的图像。幸运的是，这只有几行数据，因此重新创建更容易，但如果他们看到照片，则没有多少人愿意提供帮助。

MySql 未知列数的主元 headers

MySql pivot for unknown number of column headers

mysql

sql

pivot

dynamic-pivot