Join created table under condition - sql-server

Question

I am creating a code to join two different tables under a certain condition. The tables look like this
(TABLE 2)
date | deal_code | originator | servicer | random |
-----------------------------------------------------
2011 | 001 | commerzbank | SPV1 | 1 |
2012 | 001 | commerzbank | SPV1 | 12 |
2013 | 001 | commerzbank | SPV1 | 7 |
2013 | 005 | unicredit | SPV2 | 7 |
and another table
(TABLE 1)
date | deal_code | amount |
---------------------------
2011 | 001 | 100 |
2012 | 001 | 100 |
2013 | 001 | 100 |
2013 | 005 | 200 |
I would like to have this as the final result
date | deal_code | amount | originator | servicer | random |
--------------------------------------------------------------
2013 | 001 | 100 | commerzbank | SPV1 | 7 |
2013 | 005 | 200 | unicredit | SPV2 | 7 |
I created the following code
select q1.deal_code, q1.date
from table1 q1
where q1.date = (SELECT MAX(t4.date)
FROM table1 t4
WHERE t4.deal_code = q1.deal_code)
that gives me:
(TABLE 3)
date | deal_code | amount |
---------------------------
2013 | 001 | 100 |
2013 | 005 | 200 |
That is the latest observation for table 1, now I would like to have the originator and servicer information given the deal_code and date. Any suggestion? I hope to have been clear enough. Thanks.

Answer 1

This should do what you are looking for. Please be careful when naming columns. Date is a reserved word and is too ambiguous to be a good name for a column.
declare #Something table
(
SomeDate int
, deal_code char(3)
, originator varchar(20)
, servicer char(4)
, random int
)
insert #Something values
(2011, '001', 'commerzbank', 'SPV1', 1)
, (2012, '001', 'commerzbank', 'SPV1', 12)
, (2013, '001', 'commerzbank', 'SPV1', 7)
, (2013, '005', 'unicredit ', 'SPV2', 7)
declare #SomethingElse table
(
SomeDate int
, deal_code char(3)
, amount int
)
insert #SomethingElse values
(2011, '001', '100')
, (2012, '001', '100')
, (2013, '001', '100')
, (2013, '005', '200')
select x.SomeDate
, x.deal_code
, x.originator
, x.servicer
, x.random
, x.amount
from
(
select s.SomeDate
, s.deal_code
, s.originator
, s.servicer
, s.random
, se.amount
, RowNum = ROW_NUMBER()over(partition by s.deal_code order by s.SomeDate desc)
from #Something s
join #SomethingElse se on se.SomeDate = s.SomeDate and se.deal_code = s.deal_code
) x
where x.RowNum = 1

Answer 2

Looks like this would work:
DECLARE #MaxYear INT;
SELECT #MaxYear = MAX(date)
FROM table1 AS t1
INNER JOIN table2 AS t2
ON t1.deal_code = t2.deal_code;
SELECT t1.date,
t1.deal_code,
t1.amount,
t2.originator,
t2.servicer,
t2.random
FROM table1 AS t1
INNER JOIN table2 AS t2
ON t1.date = #MaxYear
AND t1.deal_code = t2.deal_code;

Answer 3

I agree with Sean Lange about the date column name. His method gets around the dependency on the correlated sub-query, but at the heart of things, you really just need to add an INNER JOIN to your existing query in order to get the amount column into your result set.
select
q2.date,
q2.deal_code,
q1.amount,
q2.originator,
q2.servicer,
q2.random
from
table1 q1
join
table2 q2
on q1.date = q2.date
and q1.deal_code = q2.deal_code
where q1.date = (SELECT MAX(t4.date)
FROM table1 t4
WHERE t4.deal_code = q1.deal_code)