select distinct MSG_TYPE , TX_ID as Transactions

from terra.core.fact_msgs

,tx_hash as

(select

distinct tx.MSG_TYPE ,

TX_ID ,

BLOCK_TIMESTAMP ,

FEE ,

tx_succeeded

from terra.core.fact_transactions as f inner join tx

on Transactions = f.tx_id

select

distinct MSG_TYPE ,

count (distinct TX_ID) as Transaction ,

date_trunc( 'week', BLOCK_TIMESTAMP) as week ,

sum(Transaction) over (partition by MSG_TYPE order by week) as cum_Transactions ,

count (case when tx_succeeded != 'TRUE' then 1 end) as Failed_TX,

count (case when tx_succeeded = 'TRUE' then 1 end) as Success_TX,

(Success_TX / (Success_TX + Failed_TX)) * 100 as Success_Rate,

sum(FEE) as fee ,

avg (fee) as Average_Fee,

median (fee) as Median_Fee,

min (fee) as Minimum_Fee,

max (fee) as Maximum_Fee

from tx_hash

group by 1 ,3