Question

Given a bunch of records (which represent checkins in my app) with a timestamp field, what would be a good way to determine the current streak of consecutive checkins?

In other words, with the checkins sorted by checkin time descending, how many records are there until a user missed a day?

Currently I’m using this technique:

SELECT distinct(uca.created_at::date) as created_at
    FROM user_challenge_activities as uca INNER JOIN user_challenges as uc
    ON user_challenge_id = uc.ID WHERE uc.user_id = #{user.id}
    order by (uca.created_at::date) DESC;

…where I cast the checkin timestamps to a date (to end up with e.g. 2012-03-20), then in code, go through the records and increment a counter until the date between the record and the next record is greater than 1 day.

However, this approach seems clumsy to me, and it seems like the sort of thing that Postgres would excel at.

So is there in fact a better way to accomplish this?

Answer 1

with t as (
    SELECT distinct(uca.created_at::date) as created_at
    FROM user_challenge_activities as uca 
    INNER JOIN user_challenges as uc ON user_challenge_id = uc.ID 
    WHERE uc.user_id = #{user.id}
    )
select count(*)
from t
where t.create_at > (
    select d.d
    from generate_series('2010-01-01'::date, CURRENT_DATE, '1 day') d(d)
    left outer join t on t.created_at = d.d::date
    where t.created_at is null
    order by d.d desc
    limit 1
)