Weeekly

select
  snapshot_date, week_number, listeners, playcount,
  listener_delta, listener_pct_change, listeners_indexed
from api_artist_timeseries
where artist_id = $1
order by snapshot_date asc

with peers as (
  select similar_artist_id as artist_id
  from api_artist_similar
  where artist_id = $1
),
peer_agg as (
  select
    ts.snapshot_date,
    count(*) as artist_count,
    percentile_cont(0.25) within group (order by ts.listeners_indexed) as p25_indexed,
    percentile_cont(0.5)  within group (order by ts.listeners_indexed) as median_indexed,
    percentile_cont(0.75) within group (order by ts.listeners_indexed) as p75_indexed
  from api_artist_timeseries ts
  join peers p on p.artist_id = ts.artist_id
  group by ts.snapshot_date
)
select ts.*, pa.artist_count, pa.p25_indexed, pa.median_indexed, pa.p75_indexed
from api_artist_timeseries ts
left join peer_agg pa on pa.snapshot_date = ts.snapshot_date
where ts.artist_id = $1
order by ts.snapshot_date asc