在 Postgresql 上配对顺序事件

Question

我们正在 table 上记录用户在我们的 iPad 应用程序上进行的主要操作流程。每个流都有一个开始（标记为已开始）和一个标记为已取消或已完成的结束，并且不应有任何重叠事件。

用户的一组已开始、已取消或已完成的流程如下所示：

user_id             timestamp                   event_text      event_num
info@cafe-test.de   2016-10-30 00:08:00.966+00  Flow Started    0
info@cafe-test.de   2016-10-30 00:08:15.58+00   Flow Cancelled  2
info@cafe-test.de   2016-10-30 00:08:15.581+00  Flow Started    0
info@cafe-test.de   2016-10-30 00:34:44.134+00  Flow Finished   1
info@cafe-test.de   2016-10-30 00:42:26.102+00  Flow Started    0
info@cafe-test.de   2016-10-30 00:42:49.276+00  Flow Cancelled  2
info@cafe-test.de   2016-10-30 00:42:49.277+00  Flow Started    0
info@cafe-test.de   2016-10-30 00:59:47.337+00  Flow Cancelled  2
info@cafe-test.de   2016-10-30 00:59:47.337+00  Flow Started    0
info@cafe-test.de   2016-10-30 00:59:47.928+00  Flow Cancelled  2

我们要计算已取消和已完成的流程平均持续多长时间。为此，我们需要将事件 Started 与 Canceled 或 Finished 配对。以下代码可以做到这一点，但无法解决我们遇到的以下数据质量问题：

当客户想要在结束正在进行的流程 (Flow1) 之前开始新流程（我们称之为流程 2）时，我们会在拍摄新流程的开始事件时拍摄取消事件。所以Flow1 Cancelled=Flow2 Started。但是，当我们使用 window 函数进行排序时，实际上属于不同流的有序事件之间的 lead/lag 会匹配。通过使用此代码：

WITH track_scf AS (SELECT user_id, timestamp, event_text, CASE WHEN event_text LIKE '%Started%' THEN 0 when event_text like '%Cancelled%' then 2 ELSE 1 END AS event_num FROM tracks ORDER BY 2, 4 desc ) SELECT user_id, CASE WHEN event_num=0 then timestamp end as start,CASE WHEN LEAD(event_num, 1) OVER (PARTITION BY user_id ORDER BY timestamp,event_num) <> 0 THEN LEAD(timestamp, 1) OVER (PARTITION BY user_id ORDER BY timestamp,event_num) END as end, CASE WHEN LEAD(event_num, 1) OVER (PARTITION BY user_id ORDER BY timestamp,event_num) <> 0 THEN LEAD(event_num, 1) OVER (PARTITION BY user_id ORDER BY timestamp,event_num) END as action FROM track_scf

我们得到这个结果：

user_id             start                       end                         action
info@cafe-test.de   2016-10-30 00:08:00.966+00  2016-10-30 00:08:15.58+00   2
info@cafe-test.de   2016-10-30 00:08:15.581+00  2016-10-30 00:34:44.134+00  1
info@cafe-test.de   2016-10-30 00:42:26.102+00  2016-10-30 00:42:49.276+00  2
info@cafe-test.de   2016-10-30 00:42:49.277+00  NULL                        NULL
info@cafe-test.de   2016-10-30 00:59:47.337+00  2016-10-30 00:59:47.337+00  2
info@cafe-test.de   NULL                        2016-10-30 00:59:47.928+00  2

但我们应该得到这个：

user_id             start                       end                         action
info@cafe-test.de   2016-10-30 00:08:00.966+00  2016-10-30 00:08:15.58+00   2
info@cafe-test.de   2016-10-30 00:08:15.581+00  2016-10-30 00:34:44.134+00  1
info@cafe-test.de   2016-10-30 00:42:26.102+00  2016-10-30 00:42:49.276+00  2
info@cafe-test.de   2016-10-30 00:42:49.277+00  2016-10-30 00:59:47.337+00  2
info@cafe-test.de   2016-10-30 00:59:47.337+00  2016-10-30 00:59:47.928+00  2

我需要如何更改代码才能正确配对？

Answer 1

select      user_id       
           ,"start"                       
           ,"end"                         
           ,"action"

from       (select      user_id
                       ,timestamp                 as "start"
                       ,lead (event_num)   over w as "action"
                       ,lead ("timestamp") over w as "end"
                       ,event_num

            from        tracks t

            window      w as (partition by user_id order by "timestamp",event_num desc)
            ) t

where       t.event_num = 0
;

在 Postgresql 上配对顺序事件

Pairing Sequential Events on Postgresql

sql

postgresql

datetime

intervals

window-functions