Link to home
Start Free TrialLog in
Avatar of ipaman
ipaman

asked on

Do computation in select statement using a current row column

I am writing SQL and need to write a query which returns, not only columns, but a computation on a column using the value of another column in the row.

so far I have:
query1 -> "select a, b, c, d from table1 where f = 1"

I want to do add another column to query1 (say ..e) which is the result of a select on another table where I use column 'a' from query1.

How can I accomplish this?
Avatar of apresto
apresto
Flag of Italy image

SELECT a, b, c, d, e FROM Table1 T1 INNER JOIN Table2 T2 ON T1.A = T2.ID
What is the name of the column which acts as your relationship in table2.
Infact why dont you paste your table structures here and we can go through writting you a join
Avatar of Sean Stuber
Sean Stuber

you can embed the 2nd query within the select and reference the columns from the outer query like this...


select a, b, c, d, (select table2.e+ table1.a from table2 where table2.id = table1.c)
from table1 where f = 1
Avatar of ipaman

ASKER

Below is my current query:
select id,agent_number 'Agent #',coverage_area 'Coverage Area', [Type] = CASE agent_type
WHEN 'S' THEN 'Skip'
WHEN 'R' THEN 'Recovery'
WHEN 'M' THEN 'Remarketing'
END
,city 'City',
[state] 'State',terms 'Terms', recovery_fee 'Recovery',closed_fee 'Closed',imp_client_fee 'Impound',
recovery_vol_fee 'Voluntary', imp_days_nocharge_fee 'Free Days', rating 'Rating'
from Agent
where agent_status = 1
and agent_number is not null
order by state

This is the query I need to do to create the 'New column' called '#Cases' in the query above:
(select count(distinct case_number) from [case] c, history h,
where c.id=h.case_id and
c.agent_id='current value of column 'a' above'
and  h.status=4) '# Cases'

so the relationship is agent_id in the case table to the id in the agent table...but I am only doing a count.
try this...  (assuming "a" is agent.id)
select id,agent_number 'Agent #',coverage_area 'Coverage Area', [Type] = CASE agent_type
WHEN 'S' THEN 'Skip'
WHEN 'R' THEN 'Recovery'
WHEN 'M' THEN 'Remarketing'
END
,city 'City',
[state] 'State',terms 'Terms', recovery_fee 'Recovery',closed_fee 'Closed',imp_client_fee 'Impound',
recovery_vol_fee 'Voluntary', imp_days_nocharge_fee 'Free Days', rating 'Rating',
(select count(distinct case_number) from [case] c, history h, 
where c.id=h.case_id and 
c.agent_id=Agent.id
and  h.status=4) '# Cases'
from Agent 
where agent_status = 1 
and agent_number is not null 
order by state

Open in new window

Avatar of ipaman

ASKER

the "select count(distinct case_number..." fails because it doesn't know what 'Agent' is.
If I add the agent table to the select and do a join it willl give me the intersection of the tables.
I only want to do the calculation on the other table where the current value of 'id' is used in the subquery.

not sure if i am making sense on this...
maybe your database doesn't support subqueries of that form.
You didn't specify what db you were using, I assumed Oracle, but based on your syntax, it must be something else.   If the subquery syntax isn't supported in your db then you'll need to do a join  as suggested above
Avatar of ipaman

ASKER

sorry. this is sql server.
ASKER CERTIFIED SOLUTION
Avatar of ipaman
ipaman

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial