Snowflake - invalid identifier 'LEVEL' - sql

Question

When running this query, I'm getting an error in Snowflake: invalid identifier 'LEVEL'
Also, I had some help putting parts of the query together, and I'm confused by origination of reference to 'Q'. In the two 'from' clauses in the query, I don't understand where Q is pulling from - I don't have a table called 'Q' to pull from.
WITH Q AS (SELECT LEVEL Q_LEVEL FROM DUAL A CONNECT BY PRIOR LEVEL <= 36),
Q1 AS (
select
Q.Q_LEVEL Q_LEVEL
, v_dept_history_adj.associate_id,
v_dept_history_adj.home_department_code,
v_dept_history_adj.position_effective_date
, max(position_effective_date)
OVER(PARTITION BY v_dept_history_adj.associate_id) AS most_recent_record
from datawarehouse.srctable, Q
where v_dept_history_adj.position_effective_date <=
last_day(date_from_parts(year(current_date()),
month(current_date())-Q.Q_LEVEL,1),month))
select
associate_id
, position_effective_date
, home_department_code,
most_recent_record
, (last_day(date_from_parts(year(current_date())
,month(current_date())-Q_LEVEL,1),month)) AS month
FROM Q1
where position_effective_date = most_recent_record
order by month desc, position_effective_date desc

Answer 1

So in Oracle
SELECT LEVEL
FROM DUAL A
CONNECT BY PRIOR LEVEL <= 36
would give the rows 1 -> 36, in Snowflake need to use:
select row_number() over(order by null) as q_level
from table(generator(ROWCOUNT=>36));
last_day(date_from_parts(year(current_date()),
month(current_date())-Q.Q_LEVEL,1),month)
is the same as:
last_day(dateadd(month, -q.q_level, CURRENT_DATE), month)
at which point that should be moved into the Q CTE for read ability:
WITH Q AS (
select
row_number() over(order by null) as q_level,
last_day(dateadd(month, -q_level, CURRENT_DATE), month) as last_day_month
from table(generator(ROWCOUNT=>36))
), Q1 AS (
select
q.last_day_month
,v_dept_history_adj.associate_id
,v_dept_history_adj.home_department_code
,v_dept_history_adj.position_effective_date
,max(position_effective_date) OVER (PARTITION BY v_dept_history_adj.associate_id) AS most_recent_record
from datawarehouse.srctable
join Q
on v_dept_history_adj.position_effective_date <= q.last_day_month
)
select
associate_id
,position_effective_date
,home_department_code
,most_recent_record
,last_day_month AS month
FROM Q1
where position_effective_date = most_recent_record
order by month desc, position_effective_date desc
So in the question, it was asked "where does Q come from" Q is the first of the CTE's or Common Table Expression, with the the two blocks of
WITH cte_name (
<valid sql selection>
)
The above SQL from logical perspective could be as:
with cte_q as (
select
row_number() over(order by null) as q_level,
last_day(dateadd(month, -q_level, CURRENT_DATE), month) as last_day_month
from table(generator(ROWCOUNT=>36))
), cte_q1 as (
select
q.last_day_month
,v_dept_history_adj.associate_id
,v_dept_history_adj.home_department_code
,v_dept_history_adj.position_effective_date
,max(position_effective_date) over(partition by v_dept_history_adj.associate_id) AS most_recent_record
from datawarehouse.srctable
join cte_q
on v_dept_history_adj.position_effective_date <= q.last_day_month
)
select
associate_id
,position_effective_date
,home_department_code
,most_recent_record
,last_day_month AS month
from cte_q1
where position_effective_date = most_recent_record
order by month desc, position_effective_date desc
OR cte_q could be just a sub-query:
with cte_q1 as (
select
q.last_day_month
,v_dept_history_adj.associate_id
,v_dept_history_adj.home_department_code
,v_dept_history_adj.position_effective_date
,max(position_effective_date) over(partition by v_dept_history_adj.associate_id) AS most_recent_record
from datawarehouse.srctable
join (
select
row_number() over(order by null) as q_level,
last_day(dateadd(month, -q_level, CURRENT_DATE), month) as last_day_month
from table(generator(ROWCOUNT=>36))
) as q
on v_dept_history_adj.position_effective_date <= q.last_day_month
)
select
associate_id
,position_effective_date
,home_department_code
,most_recent_record
,last_day_month AS month
from cte_q1
where position_effective_date = most_recent_record
order by month desc, position_effective_date desc
and then the cte_q1 can also be just a sub-query, but the "main" query is just filtering results, so that can be pushed into a QUALIFY clause:
select
,v_dept_history_adj.associate_id
,v_dept_history_adj.position_effective_date
,v_dept_history_adj.home_department_code
,max(position_effective_date) over(partition by v_dept_history_adj.associate_id) AS most_recent_record
q.last_day_month as month
from datawarehouse.srctable
join (
select
row_number() over(order by null) as q_level,
last_day(dateadd(month, -q_level, CURRENT_DATE), month) as last_day_month
from table(generator(ROWCOUNT=>36))
) as q
on v_dept_history_adj.position_effective_date <= q.last_day_month
qualify position_effective_date = most_recent_record
order by month desc, position_effective_date desc
The major problem I see is that v_dept_history_adj seems like a view name, but it does not appear in the original tables/cte list, so I am not sure what the real code is really doing.