Sparkbird

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

select
  ts.snapshot_date, ts.week_number, ts.listeners, ts.listeners_indexed,
  cw.artist_count, cw.p25_indexed, cw.median_indexed, cw.p75_indexed
from api_artist_timeseries ts
left join api_cohort_weekly cw
  on cw.cohort_type = $2         -- 'genre' or 'size_band'
 and cw.cohort_key = $3          -- primary_genre or size_band
 and cw.snapshot_date = ts.snapshot_date
where ts.artist_id = $1
order by ts.snapshot_date asc