获取活跃成员的平均每月视图 (Postgresql)

Question

我正在处理成员数据。我有负责的教练、教练对象的进入、退出状态和日期。因为有些教练可能会在一个月内毕业/离开，所以我想计算一个每日人数，然后得到每个教练的月活跃成员平均值。这意味着我需要将前几个月的所有受训者记入账户，这些受训者在当月仍然活跃。这是我的数据：

我正在考虑首先创建一个变量，以便获取每位教练的每日活跃会员数。这是我的第一种方法：

with all_years as (
    select y.year, m.month, d.day
    from generate_series(2019, 2022) as y(year)
             cross join generate_series(1, 12) as m(month)
             cross join generate_series(1, 31) as d(day) --<<*not sure how to adjust for days with less than 31 days??*
        select ay.*, coach, coachee, entry_status, entry_date, exit_reason, exit_date, sum(count) over (partition by ay.coach order by ay.year, ay.month, ay.day)
    from all_years ay
        left join table t
    on --.... *not sure what I can join on in this case*;

我对更简单的方法持开放态度，这个逻辑只是一个想法。

Answer 1

您可以cross join 列出不同教练的所有日期以生成组合，然后将表格与left join 一起带来：

select d.dt, c.coach, count(t.coach) no_coachees
from (select distinct coach from mytable) c
cross join generate_series('2019-01-01'::date, '2022-12-31'::date, '1 day':: interval) d(dt)
left join mytable t on t.coach = c.coach and t.entry_date <= d.dt and t.exit_date > d.dt
group by d.dt, c.coach

然后您可以使用另一个级别的聚合来获得每月平均值：

select date_trunc('month', d.dt) d_month, coach, avg(no_coachees) avg_coaches
from (
    select d.dt, c.coach, count(t.coach) no_coachees
    from (select distinct coach from mytable) c
    cross join generate_series('2019-01-01'::date, '2022-12-31'::date, '1 day':: interval) d(dt)
    left join mytable t on t.coach = c.coach and t.entry_date <= d.dt and t.exit_date > d.dt
    group by d.dt, c.coach
) t
group by date_trunc('month', d.dt), coach