动态 LAG 函数

Question

我正在尝试获取基于动态滞后函数的值，但我不会得到。 我有一个看起来像这样的表，我想要的是来自序列号的产品的值，它比我当前的序列号高 1

+--------+---------+--------+------------------------+----------+
| Client | Product | Ser_No | Product Ser_No+1 I get | Expected |
+--------+---------+--------+------------------------+----------+
|    1   |    A    |   1    |              B         |     B    |
+--------+---------+--------+------------------------+----------+
|    1   |    B    |   2    |              C         |     C    |
+--------+---------+--------+------------------------+----------+
|    1   |    C    |   3    |              D         |     F    |
+--------+---------+--------+------------------------+----------+
|    1   |    D    |   3    |              E         |     F    |
+--------+---------+--------+------------------------+----------+
|    1   |    E    |   3    |              F         |     F    |
+--------+---------+--------+------------------------+----------+
|    1   |    F    |   4    |              null      |    null  |
+--------+---------+--------+------------------------+----------+
|    2   |    D    |   1    |              E         |     E    |
+--------+---------+--------+------------------------+----------+
|    2   |    E    |   2    |              F         |     G    |
+--------+---------+--------+------------------------+----------+
|    2   |    F    |   2    |              G         |     G    |
+--------+---------+--------+------------------------+----------+
|    2   |    G    |   3    |              H         |     H    |
+--------+---------+--------+------------------------+----------+ 
|    2   |    H    |   4    |              I         |     I    |
+--------+---------+--------+------------------------+----------+
|    2   |    I    |   5    |              null      |     null |
+--------+---------+--------+------------------------+----------+

我唯一能想到的是 Lag(product, 1)，但这并不是我现在想要的。有什么想法吗？

这就是我现在所做的：

LEAD(Product, 1) OVER(PARTITION BY Cliënt ORDER BY Ser_No ASC)

但问题是第 3、4 和 8 行。这是我需要的产品代码不是来自下一行，而是来自下一个 Ser_No

Answer 1

我可能会在这里使用lead 而不是lag：

SELECT  Client,
        Product,
        Ser_No,
        LEAD(Product) OVER(PARTITION BY Client ORDER BY Ser_No) As [Product Ser_No+1]
FROM TableName

要使用Lag，您需要order by Ser_No desc。

有了这个新的编辑，我会这样做：

WITH CTE AS
(
    SELECT  Client,
            Product,
            Ser_No,
            DENSE_RANK() OVER(PARTITION BY Client ORDER BY Ser_No) As rnk
    FROM TableName
)


SELECT  Client,
        Product,
        Ser_No,
        (
            SELECT Product 
            FROM CTE t1 
            WHERE t1.Client = t0.Client 
            AND t1.Rnk = t0.Rnk + 1
        ) As [Product Ser_No+1]
FROM CTE t0