PL/SQL、MySQL 或 PostgreSQL 中的日期字段中的日期之间的天数 [关闭]

Question

关闭。 这个问题需要细节或清晰。 它目前不接受答案。

---

想改进这个问题？ 通过编辑此帖子添加详细信息并澄清问题。

9 个月前关闭。

改进这个问题

我有类似的东西：

  ID  |     DATE
---------------------
  A   | '2017-02-17'  
  A   | '2018-06-01'
  A   | '2020-11-27'
  B   | '2005-01-30' 
  B   | '2019-04-18'
  C   | '2016-05-18'  
  C   | '2016-05-23'
  C   | '2017-10-01'
  C   | '2020-11-28'

我需要第三列，其中包含每个日期和最后一个 bihind 之间的天数。 就像是

  ID  |     DATE     | DATEDIFF
--------------------------------
  A   | '2017-02-17' |    
  A   | '2018-06-01' |    469
  A   | '2020-11-27' |    910
  B   | '2005-01-30' |
  B   | '2019-04-18' |   5191
  C   | '2016-05-18' | 
  C   | '2016-05-23' |      5
  C   | '2017-10-01' |    496
  C   | '2020-11-28' |   1154

Answer 1

根据您的 DBMS 和版本，使用以下查询之一：

SELECT *,
DATEDIFF(d, (LAG(date) OVER (ORDER BY M.id)), M.date) as datediffvalue
FROM MTB M

其他选项：

;    
WITH CTE
AS (
    SELECT rownum = ROW_NUMBER() OVER (
            ORDER BY p.id
            ), p.id, p.DATE
    FROM MTB p
    )
SELECT CTE.id, CTE.DATE, CASE 
        WHEN prev.id = CTE.id
            THEN DATEDIFF(d, prev.DATE, CTE.DATE)
        END AS datediffvalue
FROM CTE
LEFT JOIN CTE prev ON prev.rownum = CTE.rownum - 1

结果

+----+-------------------------+---------------+
| id | date                    | datediffvalue |
| A  | 2017-02-17 00:00:00.000 |               |
| A  | 2018-06-01 00:00:00.000 | 469           |
| A  | 2020-11-27 00:00:00.000 | 910           |
| B  | 2005-01-30 00:00:00.000 |               |
| B  | 2019-04-18 00:00:00.000 | 5191          |
| C  | 2016-05-18 00:00:00.000 |               |
| C  | 2016-05-23 00:00:00.000 | 5             |
| C  | 2017-10-01 00:00:00.000 | 496           |
| C  | 2020-11-28 00:00:00.000 | 1154          |
+----+-------------------------+---------------+

Answer 2

MySQL 8, window LAG滞后

SELECT
    t.*, DATEDIFF(date, LAG(date) OVER w) AS datediff
FROM t 
WINDOW w AS (PARTITION BY id ORDER BY date)

MySQL 8版之前

SELECT
    DATEDIFF(date, CASE WHEN @id = id THEN @lag END) AS datediff,
    @id := id AS id,
    @lag := t.date AS date
FROM t, (SELECT @lag := NULL, @id := NULL) var
ORDER BY id, date