如何查找一年的先前金额的总和，包括当前金额

Question

我需要为第二个条目获取当前与该特定年份的前一笔金额的总和。

Input_table

+-----------+----------+-------------+-----------+----------+
| ID        | Name     | Date1       |Date 2     |amount    |
+-----------+----------+-------------+-----------+----------+
| 10000000  | ABC      | 11/2/2017   |**11/2/2018**  |2504      |
| 10000000  | ABC      | 12/20/2017  |**11/2/2018**  |-2174     |
| 10000000  | ABC      | 10/05/2018  |10/05/2019 |1234      |
| 10000000  | ABC      | 10/06/2019  |10/06/2020 |3456      |
+-----------+----------+-------------+-----------+----------+

以下是所需的输出：

+-----------+----------+-------------+-----------+----------+
| ID        | Name     | Date1       |Date 2     |amount    |
+-----------+----------+-------------+-----------+----------+
| 10000000  | ABC      | 11/2/2017   |**11/2/2018**  |2504      |
| 10000000  | ABC      | 12/20/2017  |**11/2/2018**  |329       |
| 10000000  | ABC      | 10/05/2018  |10/05/2019 |1234      |
| 10000000  | ABC      | 10/06/2019  |10/06/2020 |3456      |
+-----------+----------+-------------+-----------+----------+

记录3金额-1234（1234 0（该年无记录）），因为日期2不同

记录4金额- 3456 (3456 0(当年无记录))与日期2不同

如果两条记录的日期2相同，我正在查找记录的当前和以前金额值的总和。假设我们有3条记录具有相同的日期2值。然后第一条记录应该有它的实际金额值，第二条记录应该具有第一条记录金额2记录金额，而第三条记录应该拥有第一条记录数量2记录金额3记录金额。

Answer 1

我在这里做了一些相当大的假设。主要是，您希望查看当前记录的“Date2”值是否存在于上一年的“Date1”中的某个位置（例如，Date2：11/02/2018 正在查看 Date1：11/02/2017 是否存在），并且 ID 和 Name 列在筛选数据时具有某种相关性。

以下是我在SSMS中使用的内容：

-- Create sample data.

DECLARE @Data TABLE (
    ID VARCHAR(10), [Name] VARCHAR(3), Date1 DATE, Date2 DATE, Amount DECIMAL(18,2)
);

INSERT INTO @Data ( ID, [Name], Date1, Date2, Amount ) VALUES
( '10000000', 'ABC', '11/2/2017', '11/2/2018',  2504),
( '10000000', 'ABC', '12/20/2017', '11/2/2018',  -2174),
( '10000000', 'ABC', '10/05/2018', '10/05/2019',  1234),
( '10000000', 'ABC', '10/06/2019', '10/06/2020',  3456);

-- Query sample data

SELECT
    MyData.ID, MyData.[Name], MyData.Date1, MyData.Date2, MyData.Amount
    , ISNULL( LastYear.Amount, 0 ) AS PreviousYear
    , ( MyData.Amount + ISNULL( LastYear.Amount, 0 ) ) AS NewAmount
FROM @Data AS MyData
OUTER APPLY (

    SELECT
        SubData.Amount AS Amount
    FROM @Data AS SubData
    WHERE
        SubData.ID = MyData.ID
        AND SubData.[Name] = MyData.[Name]
        AND SubData.Date1 = DATEADD( yy, -1, MyData.Date2 )
        AND SubData.Date1 <> MyData.Date1   -- A weak attempt to exclude the current MyData record. Really needs a unique id.

) AS LastYear
ORDER BY
    MyData.Date2;

--退货

+----------+------+------------+------------+----------+--------------+-----------+
|    ID    | Name |   Date1    |   Date2    |  Amount  | PreviousYear | NewAmount |
+----------+------+------------+------------+----------+--------------+-----------+
| 10000000 | ABC  | 2017-11-02 | 2018-11-02 | 2504.00  | 0.00         | 2504.00   |
| 10000000 | ABC  | 2017-12-20 | 2018-11-02 | -2174.00 | 2504.00      | 330.00    |
| 10000000 | ABC  | 2018-10-05 | 2019-10-05 | 1234.00  | 0.00         | 1234.00   |
| 10000000 | ABC  | 2019-10-06 | 2020-10-06 | 3456.00  | 0.00         | 3456.00   |
+----------+------+------------+------------+----------+--------------+-----------+

理想情况下，被查询的表有一个PK，可以用来将当前记录排除在外(注意使用Date1的微弱尝试)。

使用PK（唯一id）的相同示例：

DECLARE @Data TABLE (
    ID VARCHAR(10), [Name] VARCHAR(3), Date1 DATE, Date2 DATE, Amount DECIMAL(18,2), pk_PrimaryKey INT IDENTITY(1,1) PRIMARY KEY
);

INSERT INTO @Data ( ID, [Name], Date1, Date2, Amount ) VALUES
( '10000000', 'ABC', '11/2/2017', '11/2/2018',  2504),
( '10000000', 'ABC', '12/20/2017', '11/2/2018',  -2174),
( '10000000', 'ABC', '10/05/2018', '10/05/2019',  1234),
( '10000000', 'ABC', '10/06/2019', '10/06/2020',  3456);

SELECT
    MyData.ID, MyData.[Name], MyData.Date1, MyData.Date2, MyData.Amount
    , ISNULL( LastYear.Amount, 0 ) AS PreviousYear
    , ( MyData.Amount + ISNULL( LastYear.Amount, 0 ) ) AS NewAmount
FROM @Data AS MyData
OUTER APPLY (

    SELECT
        SubData.Amount AS Amount
    FROM @Data AS SubData
    WHERE
        SubData.ID = MyData.ID
        AND SubData.[Name] = MyData.[Name]
        AND SubData.Date1 = DATEADD( yy, -1, MyData.Date2 )
        AND SubData.pk_PrimaryKey <> MyData.pk_PrimaryKey

) AS LastYear
ORDER BY
    MyData.Date1;

不管怎样，这是我根据我所掌握的一点点信息做出的最好猜测。希望这能让你朝着正确的方向前进。