IT俱乐部 MsSql SQL统计连续登陆3天用户的实现示例

SQL统计连续登陆3天用户的实现示例

发布: 2024年 6月 10日 480阅读

目录

1. 数据准备

-- 数据准备
WITH user_active_info AS (
SELECT * FROM (
    VALUES ('10001' , '2023-02-01'),('10001' , '2023-02-03')
          ,('10001' , '2023-02-04'),('10001' , '2023-02-05')
          ,('10002' , '2023-02-02'),('10002' , '2023-02-03')
          ,('10002' , '2023-02-04'),('10002' , '2023-02-05')
          ,('10002' , '2023-02-07'),('10003' , '2023-02-02')
          ,('10003' , '2023-02-03'),('10003' , '2023-02-04')
          ,('10003' , '2023-02-05'),('10003' , '2023-02-06')
          ,('10003' , '2023-02-07'),('10003' , '2023-02-08')
          ,('10004' , '2023-02-03'),('10004' , '2023-02-04')
          ,('10004' , '2023-02-06'),('10004' , '2023-02-07')
          ,('10004' , '2023-02-08'),('10004' , '2023-02-08') 
          ,('10005' , '2023-02-02'),('10005' , '2023-02-05') 
) AS user_active_info(user_id, active_date) 
)

2. 方法一: 差值计算

-- 1. 对用户数据进行分组,按照活跃日期进行排序(去重:防止有一天有多次活跃记录)
SELECT
      user_id
    , active_date
    , ROW_NUMBER() OVER(PARTITION BY user_id ORDER BY active_date) AS rn 
FROM user_active_info
GROUP BY user_id , active_date
; 

user_id	active_date	rn
10001	2023-02-01	1
10001	2023-02-03	2
10001	2023-02-04	3
10001	2023-02-05	4
10002	2023-02-02	1
10002	2023-02-03	2
10002	2023-02-04	3
10002	2023-02-05	4
10002	2023-02-07	5
…	…	…

-- 2. 使用活跃日期和排序rn进行差值计算,得到的日期如果是相等的,就说明活跃日期是连续的
SELECT
      user_id
    , active_date
    , rn 
    , DATE_SUB(active_date,rn) AS sub_date
FROM (
    SELECT
          user_id
        , active_date
        , ROW_NUMBER() OVER(PARTITION BY user_id ORDER BY active_date) AS rn 
    FROM user_active_info 
    GROUP BY user_id , active_date
    ) a
; 

user_id	active_date	rn	sub_date
10001	2023-02-01	1	2023-01-31
10001	2023-02-03	2	2023-02-01
10001	2023-02-04	3	2023-02-01
10001	2023-02-05	4	2023-02-01
10002	2023-02-02	1	2023-02-01
10002	2023-02-03	2	2023-02-01
10002	2023-02-04	3	2023-02-01
10002	2023-02-05	4	2023-02-01
10002	2023-02-07	5	2023-02-02
…	…	…	…

-- 3. 按照user_id和sub_date 进行分组求和,筛选出连续登陆天数大于3天的用户
SELECT
      user_id
    , MIN(active_date) AS begin_date
    , MAX(active_date) AS end_date
    , COUNT (1) AS login_duration
FROM (
    SELECT
          user_id
        , active_date
        , rn 
        , DATE_SUB(active_date,rn) AS sub_date
    FROM (
        SELECT
              user_id
            , active_date
            , ROW_NUMBER() OVER(PARTITION BY user_id ORDER BY active_date) AS rn 
        FROM user_active_info 
        GROUP BY user_id , active_date
    ) a
) b
GROUP BY user_id , sub_date
HAVING login_duration >= 3
; 

user_id	begin_date	end_date	login_duration
10001	2023-02-03	2023-02-05	3
10002	2023-02-02	2023-02-05	4
10003	2023-02-02	2023-02-08	7
10004	2023-02-06	2023-02-08	3

3. 方法二: lead或lag函数

-- 1. 将active_date 上抬2行,不存在默认为'0'(计算连续活跃3天以上的, 上抬2行,n天上抬n-1行)(去重:防止有一天有多次活跃记录)
SELECT
      user_id
    , active_date
    , lead(active_date , 2 , 0) OVER(PARTITION BY user_id ORDER BY active_date) AS lead_active_date 
FROM user_active_info
GROUP BY user_id , active_date

user_id	active_date	lead_active_date
10001	2023-02-01	2023-02-04
10001	2023-02-03	2023-02-05
10001	2023-02-04	0
10001	2023-02-05	0
10002	2023-02-02	2023-02-04
10002	2023-02-03	2023-02-05
10002	2023-02-04	2023-02-07
10002	2023-02-05	0
10002	2023-02-07	0
…	…	…

-- 2. 过滤筛选出, lead_active_date 与 active_date 差值为2的, 差值2 -> 连续活跃了3天
SELECT
      user_id , active_date , lead_active_date
FROM (
    SELECT
          user_id
        , active_date
        , lead(active_date , 2 , 0) OVER(PARTITION BY user_id ORDER BY active_date) AS lead_active_date
    FROM user_active_info
    GROUP BY user_id , active_date
) a 
WHERE  lead_active_date != '0'
    AND DATEDIFF(lead_active_date , active_date) = 2

user_id	active_date	lead_active_date
10001	2023-02-03	2023-02-05
10002	2023-02-02	2023-02-04
10002	2023-02-03	2023-02-05
…	…	…

-- 3. user_id 去重, 得到连续活跃天数>=3天的用户
SELECT
      user_id
FROM (
    SELECT
          user_id , active_date , lead_active_date
    FROM (
        SELECT
              user_id
            , active_date
            , lead(active_date , 2 , 0) OVER(PARTITION BY user_id ORDER BY active_date) AS lead_active_date 
        FROM user_active_info
        GROUP BY user_id , active_date
    ) a 
    WHERE  lead_active_date != '0'
        AND DATEDIFF(lead_active_date , active_date) = 2
) b
GROUP BY user_id

user_id
10001
10002
10003
10004

到此这篇关于SQL统计连续登陆3天用户的实现示例的文章就介绍到这了,更多相关SQL统计连续登陆3天用户内容请搜索IT俱乐部以前的文章或继续浏览下面的相关文章希望大家以后多多支持IT俱乐部！

本文收集自网络，不代表IT俱乐部立场，转载请注明出处。https://www.2it.club/database/mssql/11911.html

77赞

标签:sql 用户登陆示例统计