Query last N related rows per row

Question

I have the following query which fetches the id of the latest N observations for each station:

SELECT id
FROM (
  SELE         


        

Answer 1

This is a good anwser only if you are not required to query up-to-date live data.

Preparation (requires postgresql 9.3)

drop materialized view test;
create materialized view test as select * from (
  SELECT station_id, id, created_at,
      row_number() OVER(
          PARTITION BY station_id
          ORDER BY created_at DESC
      ) as rn
  FROM (
      SELECT
          station_id,
          id,
          created_at
      FROM observations
  ) s
 ) q WHERE q.rn <= 100 -- use a value that will be your max limit number for further queries
ORDER BY station_id, rn DESC ;


create index idx_test on test(station_id,rn,created_at);

How to query data:

select * from test where rn<10 order by station_id,created_at;

Your original query was 281 ms on my machine and this new was 15 ms.

How to update the view with fresh data:

refresh materialized view test;

I have another solution that does not require materialized view and works with live, up-to-date data. But given that you don't need up-to-date data, this materialized view is much more efficient.