select date_trunc({{period_type}}, BLOCK_TIMESTAMP) as "Date",

count(distinct from_address) as "New Users",

lag("New Users") over (order by "Date") as lag_AUs,

round(100*("New Users"-lag_AUs)/lag_AUs,2) as "New Users Change %"

from

from_address,

min(BLOCK_TIMESTAMP) as BLOCK_TIMESTAMP

from berachain.testnet.fact_transactions where TX_SUCCEEDED=true group by 1)

where BLOCK_TIMESTAMP::date>='{{start_day}}'

and date_trunc({{period_type}}, BLOCK_TIMESTAMP)<=date_trunc({{period_type}}, '{{target_day}}'::date)

group by 1

order by 1 desc