address,

address_name,

project_name

from polygon.core.dim_labels

where project_name like '%opensea%' or project_name like '%nft%'

),

nft as (select

block_timestamp,

tx_hash,

origin_from_address as from_users,

origin_to_address as to_users,

raw_amount,

raw_amount / POW(10,18) AS adjusted_amount,

contract_address,

from polygon.core.fact_token_transfers

join labels ON labels.address = polygon.core.fact_token_transfers.contract_address)

select

DATE(block_timestamp) as dayz,

project_name,

count(distinct from_users) as interact_users,

count(distinct tx_hash) as txs,

sum(adjusted_amount) as tokenvolume,

sum(interact_users) over (partition by project_name order by dayz asc rows between unbounded preceding and current row) as users_cumulative,

sum(txs) over (partition by project_name order by dayz asc rows between unbounded preceding and current row) as txs_cumulative

from nft

group by 1, 2