Question

我正在尝试将重叠的起始时间戳合并为单个时间跨度。 SO上也有类似的问题here。我想为数据中的每个用户单独合并时间戳。

示例数据：

-- drop table if exists app_log;

create table app_log (
  user_id int,
  login_time timestamp,
  logout_time timestamp
);

insert into app_log values
  (1, '2014-01-01 08:00', '2014-01-01 10:00'), /* here we start */
  (1, '2014-01-01 09:10', '2014-01-01 09:59'), /* fully included in previous interval */
  (1, '2014-01-01 10:00', '2014-01-01 10:48'), /* continuing first interval */
  (1, '2014-01-01 10:40', '2014-01-01 10:49'), /* continuing previous interval */
  (1, '2014-01-01 10:55', '2014-01-01 11:00'), /* isolated interval */
  (2, '2014-01-01 09:00', '2014-01-01 11:00'), /* 2nd user is shifted by one hour */
  (2, '2014-01-01 10:10', '2014-01-01 10:59'), /* to simulate overlaps with 1st user */
  (2, '2014-01-01 11:00', '2014-01-01 11:48'), 
  (2, '2014-01-01 11:40', '2014-01-01 11:49'), 
  (2, '2014-01-01 11:55', '2014-01-01 12:00')  
;

所需结果：

  used_id  login_time       logout_time
  1        2014-01-01 08:00 2014-01-01 10:49 /* Merging first 4 lines */
  1        2014-01-01 10:55 2014-01-01 11:00 /* 5 th line is isolated */
  2        2014-01-01 09:00 2014-01-01 11:49 /* Merging lines 6-9 */
  2        2014-01-01 11:55 2014-01-01 12:00 /* last line is isolated */

我尝试使用mentioned question中提供的解决方案，但即使对于单个用户也不会返回正确的答案：

with recursive

in_data as (select login_time as d1, logout_time as d2 from app_log where user_id = 1)

, dateRanges (ancestorD1, parentD1, d2, iter) as
(
--anchor is first level of collapse
    select
        d1 as ancestorD1,
        d1 as parentD1,
        d2,
        cast(0 as int) as iter
    from in_data

--recurse as long as there is another range to fold in
    union all

    select
        tLeft.ancestorD1,
        tRight.d1 as parentD1,
        tRight.d2,
        iter + 1  as iter
    from dateRanges as tLeft join in_data as tRight
        --join condition is that the t1 row can be consumed by the recursive row
        on tLeft.d2 between tRight.d1 and tRight.d2
            --exclude identical rows
            and not (tLeft.parentD1 = tRight.d1 and tLeft.d2 = tRight.d2)
)
select
    ranges1.*
from dateRanges as ranges1
where not exists (
    select 1
    from dateRanges as ranges2
    where ranges1.ancestorD1 between ranges2.ancestorD1 and ranges2.d2
        and ranges1.d2 between ranges2.ancestorD1 and ranges2.d2
        and ranges2.iter > ranges1.iter
);

结果：

ancestord1 parentd1 d2 iter
2014-01-01 10:55:00;2014-01-01 10:55:00;2014-01-01 11:00:00;0
2014-01-01 08:00:00;2014-01-01 10:40:00;2014-01-01 10:49:00;2
2014-01-01 09:10:00;2014-01-01 10:40:00;2014-01-01 10:49:00;3

上面的查询有什么问题？我如何扩展它以获得用户的结果？在PostgreSQL中有没有更好的解决这个问题的方法？

Answer 1

我使用窗口函数和许多嵌套子查询找到了这个example of how to make a 'range aggregate'。我只是通过user_id对它进行了分区和分组，它似乎做了你想做的事情：

SELECT user_id, min(login_time) as login_time, max(logout_time) as logout_time
FROM (
    SELECT user_id, login_time, logout_time,
        max(new_start) OVER (PARTITION BY user_id ORDER BY login_time, logout_time) AS left_edge
    FROM (
        SELECT user_id, login_time, logout_time,
            CASE 
                WHEN login_time <= max(lag_logout_time) OVER (
                    PARTITION BY user_id ORDER BY login_time, logout_time 
                ) THEN NULL 
                ELSE login_time 
            END AS new_start
        FROM (
            SELECT 
                user_id, 
                login_time, 
                logout_time,
                lag(logout_time) OVER (PARTITION BY user_id ORDER BY login_time, logout_time) AS lag_logout_time
            FROM app_log
        ) AS s1
    ) AS s2
) AS s3
GROUP BY user_id, left_edge
ORDER BY user_id, min(login_time)

结果：

 user_id |     login_time      |     logout_time
---------+---------------------+---------------------
       1 | 2014-01-01 08:00:00 | 2014-01-01 10:49:00
       1 | 2014-01-01 10:55:00 | 2014-01-01 11:00:00
       2 | 2014-01-01 09:00:00 | 2014-01-01 11:49:00
       2 | 2014-01-01 11:55:00 | 2014-01-01 12:00:00
(4 rows)

首先检测每个新范围的开始（由user_id分区），然后按检测范围进行扩展和分组。我发现我必须仔细阅读那篇文章才能理解它！

文章建议通过删除最里面的子查询并更改窗口范围，可以使用Postgresql＆gt; = 9.0进行简化，但我无法使其工作。

在PostgreSQL中按用户展平相交的时间跨度

1 个答案: