Re:SQL should be your default choice for data engineering pi...
by srcreigh on Tuesday January 31, 2023 @01:03PM
(#34598961)
The dbt site describes this as a simple example:
select * from {{ref('really_big_table')}}
{% if incremental and target.schema == 'prod' %}
where timestamp >= (select max(timestamp) from {{this}})
{% else %}
where timestamp >= dateadd(day, -3, current_date)
{% endif %}
Re:SQL should be your default choice for data engineering pi...
by billythemaniam on Tuesday January 31, 2023 @03:16PM
(#34601285)
As someone who has written 100s (maybe 1000s at this point) of DBT models, the amount of Jinja you need is at most 1-2% of the codebase.
I know you are pitching your startup, but that's just completely untrue.