a.block_timestamp::date as day,

label as platform,

iff(day >= '2022-11-08', 'After FTX Collapse', 'Before FTX Collapse') as timespan,

count(distinct (a.tx_id)) as txn_count,

(sum(iff(succeeded = 1, 1, 0))/txn_count) * 100 as success_rate,

100 - success_rate as fail_rate,

sum(fee/1e9) as fee_total,

avg(fee/1e9) as fee_avg,

avg(fee/1e9) as fee_median,

sum(fee_total) over (order by day asc) as fee_total_cum,

sum(fee_avg) over (order by day asc) as fee_avg_cum

from solana.core.dim_labels

join solana.core.fact_events a

join solana.core.fact_transactions b

on address = program_id and a.tx_id = b.tx_id

where day >= current_date - 30

and label in ('jupiter', 'raydium')

group by timespan, day, label

order by day, fee_avg asc