Query editor
with cohort as (select trader from query_8149881where block_time >= timestamp '2025-06-01'group by 1 having count(*) >= 3)select d.blockchain, count(*) as trades, count(distinct d.tx_from) as users, approx_percentile(d.amount_usd, 0.5) as median_size, sum(d.amount_usd) as vol_usd, count(distinct case when d.block_date >= date '2026-05-01' then d.tx_from end) as users_ult_3mfrom dex.trades djoin cohort c on c.trader = d.tx_fromwhere d.block_date > date '2025-10-15'group by 1order by 3 desclimit 20