Question

我有一个查询已经转动与交易数据相关的数据。我有一个包含客户编号，交易类型和交易金额的表。

我的原始查询返回了每个客户按交易类型细分的交易总数。它在下面：

;WITH query AS
(
SELECT customernum,[Cash],[Card]
FROM (
transactions
)
AS source
PIVOT
(
COUNT(TransType) FOR TransType IN ([Cash],[Card])
)
AS pvt
) SELECT customernum AS 'Customer Number', SUM([Cash]) AS 'Cash Transactions', 
SUM([Card]) AS 'Card Transactions' FROM query
GROUP BY customernum
ORDER BY customernum

我现在需要将总值添加到查询中，但仍按事务类型细分。

例如，来自表格＆＃39;交易的样本数据＆＃39;是：

╔═════════════╦═══════════╦════════╗
║ CustomerNum ║ TransType ║ Amount ║
╠═════════════╬═══════════╬════════╣
║         111 ║ Cash      ║      5 ║
║         111 ║ Card      ║      5 ║
║         111 ║ Cash      ║      5 ║
║         111 ║ Cash      ║      5 ║
║         111 ║ Card      ║      5 ║
║         222 ║ Card      ║      5 ║
║         222 ║ Cash      ║      5 ║
║         222 ║ Card      ║      5 ║
║         333 ║ Cash      ║      5 ║
║         333 ║ Cash      ║      5 ║
║         333 ║ Cash      ║      5 ║
╚═════════════╩═══════════╩════════╝

我的原始查询产生了这个结果：

╔═════════════╦══════╦══════╗
║ CustomerNum ║ Cash ║ Card ║
╠═════════════╬══════╬══════╣
║         111 ║    3 ║    2 ║
║         222 ║    1 ║    2 ║
║         333 ║    3 ║    0 ║
╚═════════════╩══════╩══════╝

我现在需要它来回复：

╔═════════════╦══════╦═════════════╦══════╦═════════════╗
║ CustomerNum ║ Cash ║ Cash Amount ║ Card ║ Card Amount ║
╠═════════════╬══════╬═════════════╬══════╬═════════════╣
║         111 ║    3 ║          15 ║    2 ║          10 ║
║         222 ║    1 ║           5 ║    2 ║          10 ║
║         333 ║    3 ║          15 ║    0 ║           0 ║
╚═════════════╩══════╩═════════════╩══════╩═════════════╝

Answer 1

只需使用条件聚合：

SELECT CustomerNum,
       SUM(CASE WHEN TransType = 'Cash' THEN 1 ELSE 0 END) AS cash,
       SUM(CASE WHEN TransType = 'Cash' THEN amount ELSE 0 END) AS cash_amount,
       SUM(CASE WHEN TransType = 'Card' THEN 1 ELSE 0 END) AS card,
       SUM(CASE WHEN TransType = 'Card' THEN amount ELSE 0 END) AS card_amount
FROM transactions
GROUP BY CustomerNum;

Answer 2

考虑将两个透视查询用作派生表：

SELECT t1.customernum, t1.Cash, t2.Cash As [Cash Amount],
       t1.Card, t2.Card As [Card Amount]

FROM

    (SELECT * FROM
         (SELECT customernum, TransType
          FROM transactions) AS src1
     PIVOT
     (
      COUNT(TransType) FOR TransType IN ([Cash],[Card])
     )
     AS pvt1) As t1

INNER JOIN

   (SELECT * FROM
        (SELECT customernum, TransType, Amount
         FROM transactions) AS src2
    PIVOT
    (
     SUM(Amount) FOR TransType IN ([Cash],[Card])
    )
    AS pvt2) As t2

ON t1.customernum = t2.customernum

SQLFiddle

SQL查询中的第二个Pivot

2 个答案: