Question

SELECT * FROM 
            (
              SELECT ID,
                     CODE,
                     CAST(SUM(QTY) AS INT) AS G_QTY,
                     CAST(SUM(Amount) AS INT) AS AmountUsed
              FROM Tbl_Used
              WHERE CODE IN ('ABC','DEF','GHI')
              GROUP BY ID,CODE
            ) AS S
PIVOT
   (
    SUM(AmountUsed) 
    FOR CODE IN(ABC,DEF)
    ) AS Pivot1

我也希望将G_QTY用于CODE IN（'GHI'）

结果应该是：

ID  ABC DEF GHI
1   10  20  30

表数据

ID  CODE    QTY AMOUNT
1122    ABC 546 789653
456     GHI 5644    65433213
1455    DEF 99978   464646
1122    ABC 6565    98464
1455    DEF 88933   7933
2454    GHI 88989   5522
456     ABC 108309  789653
2454    DEF 125414  65433213
4452    GHI 142519  464646
3120    GHI 159624  98464
456     GHI 176729  7933
3120    ABC 193834  5522
456     GHI 210938  125414
4452    DEF 228043  142519
1455    ABC 245148  159624
3120    DEF 262253  176729

数据不实，但与生产环境相似

Answer 1

可能是这样的：

SELECT ID,Max(ABC) AS ABC,Max(DEF) As DEF,Max(GHI) AS GHI
FROM 
   (
    SELECT ID,
           CODE,
           CODE As CODE1,
           CAST(SUM(QTY) AS INT) AS G_QTY,
           CAST(SUM(Amount) AS INT) AS AmountUsed
    FROM Tbl_Used
    WHERE CODE IN ('ABC','DEF','GHI')
    GROUP BY ID,CODE
    ) AS S
PIVOT
    (
    SUM(AmountUsed) 
    FOR CODE IN([ABC],[DEF])
    ) AS Pivot1
PIVOT
    (
    SUM(G_QTY) 
    FOR CODE1 IN([GHI])
    ) AS Pivot2

Group By ID

<强> FIDDLE DEMO

输出：

ID      ABC     DEF     GHI
456     789653  (null)  393311
1122    888117  (null)  (null)
1455    159624  472579  (null)
2454    (null)  65433213 88989
3120    5522    176729  159624
4452    (null)  142519  142519

Answer 2

我重写了查询。这样就不需要两个数据透视，也因为枢轴已经聚合并且不需要内部查询分组。

所以我的建议如下：

测试数据

DECLARE @tbl TABLE(ID INT,CODE VARCHAR(5),QTY INT, AMOUNT INT)
INSERT INTO @tbl
VALUES
(1122,'ABC',546,789653),
(456,'GHI',5644,65433213),
(1455,'DEF',99978,464646),
(1122,'ABC',6565,98464),
(1455,'DEF',88933,7933),
(2454,'GHI',88989,5522),
(456,'ABC',108309,789653),
(2454,'DEF',125414,65433213),
(4452,'GHI',142519,464646),
(3120,'GHI',159624,98464),
(456,'GHI',176729,7933),
(3120,'ABC',193834,5522),
(456,'GHI',210938,125414),
(4452,'DEF',228043,142519),
(1455,'ABC',245148,159624),
(3120,'DEF',262253,176729)

<强>查询

SELECT
    *
FROM
(
    SELECT 
        tbl.ID,
        tbl.CODE,
        (
            CASE 
                WHEN tbl.CODE='GHI' 
                THEN QTY 
                ELSE Amount
            END
        ) AS ValueToPivot
    FROM 
        @tbl AS tbl
    WHERE tbl.CODE IN ('ABC','DEF','GHI')
) AS SourceTable
PIVOT
(
    SUM(ValueToPivot)
    FOR CODE IN ([ABC],[DEF],[GHI])
) AS pvt

这会得到结果：

ID      ABC     DEF         CHI
------------------------------------
456     789653  NULL        393311
1122    888117  NULL        NULL
1455    159624  472579      NULL
2454    NULL    65433213    88989
3120    5522    176729      159624
4452    NULL    142519      142519

关于两个聚合体的PIVOT

2 个答案: