如何编写显示基于时间戳的细分的PostgreSQL查询

Question

在我的PostgreSQL数据库中，我有一个包含以下字段的邀请表：

Invitations: id, created_at, completed_at

我正在编写一个PostgreSQL查询，该查询按每周同类群组细分数据，并在created_at的1天之内（在1-2天至3天或更长时间之间）显示完成时间的细分。完成时间基于字段completed_at

所需输出：

weeks_ago | start | end | Total | % Completed in Day 1 | % Completed in D2 | % Completed 3 or days later
0 | 2020-02-04 | 2020-02-11 | ?% | ?% | ?% 
1 | 2020-01-28 | 2020-02-03 | ?% | ?% | ?% 

[这是我到目前为止的内容：

SELECT  TRUNC(DATE_PART('day', CURRENT_DATE - i.created_at )/7)  AS weeks_ago,
                date(min(i.created_at)) AS "start",
                date(max(i.created_at)) AS "end",
                count(DISTINCT i.id) AS "total",
FROM invitations i
GROUP BY weeks_ago
ORDER BY weeks_ago ASC;

鉴于上面的查询，如何添加以下内容？

• 在created_at字段的1天内完成的邀请的百分比
• 在created_at字段的1-2天内完成的邀请的百分比
• 3天或更长时间后完成的邀请百分比

谢谢

Answer 1

嗯。 。 。这似乎是条件聚合：

select date_trunc('week', i.created_at) as week_start,
       date_trunc('week', i.created_at) + interval '6 day' as week_end,       
       floor( (date_trunc('week', current_date) - date_trunc('week', i.created_at) / 7) as weeks ago,
       avg( (i.completed_at <= i.started_at + interval '1 day')::int ) as percent_1,
       avg( (i.completed_at <= i.started_at + interval '2 day' and i.started_at > i.start_at + interval '1' day >)::int ) as percent_2,
       avg( (i.completed_at > i.started_at + interval '2 day')::int ) as percent_3plus
from invitations i
group by week_start, week_end, weeks_ago
order by weeks_ago asc;