Subtracting two columns from two different table with different data - sql-server

Question

i had this problem where i got two tables with different values.
TABLE 1:
Description Qty Amount
Proc 1 1 100
Proc 2 1 50
Proc 3 1 60
TABLE 2:
Description Payment
Proc 1 60
Proc 1 30
Proc 2 20
Proc 3 60
Proc 2 20
So, the result should be like:
Description Balance
Proc 1 10
Proc 2 40
Proc 3 0
How will i do this with select query? thanks.
I Already tried this code but it should first get the distinct rows of TABLE1 and sum distinct values from TABLE2 before subtracting the two tables.
SELECT FEESList.[Fee Description], sum(StudentBILLING.Quantity* StudentBILLING.Total- isnull(StudReceipts.Amount,0)) as Balance
FROM StudentBILLING INNER JOIN FEESList ON StudentBILLING.FeeID = FEESList.FeeID INNER JOIN SREGStudentInformation ON StudentBILLING.StudentID = SREGStudentInformation.ID INNER JOIN
SemesterList ON StudentBILLING.SemesterID = SemesterList.SemID INNER JOIN SchoolYear ON StudentBILLING.SYid = SchoolYear.[SY ID] FULL JOIN StudReceipts ON FEESList.FeeID = StudReceipts.FeeID
WHERE (SREGStudentInformation.[Student ID] = #StudentNumber) AND (SemesterList.[Sem.] = #Sem) AND (SchoolYear.[School Year] = #SchoolYear) AND
(FEESList.[Type of Fee] = 2) AND (StudentBILLING.Quantity* StudentBILLING.Total- ISNULL(StudReceipts.Amount,0))>0
GROUP BY FEESList.[Fee Description]
The result is from this;
Description Amount Payments
Proc 1 100 60
Proc 1 100 30
Proc 2 50 20
Proc 2 50 20
Proc 3 60 60
To this result which is not supposed to be.
Description Amount
Proc 1 100
Proc 2 50
Can you help me with this?

Answer 1

You could try below with sample data which you have provided
First, it will generated row numbers by using row_number() function partition by Description and Payment column which has duplicate payment made & sum() if have duplicate payments made
;WITH cte AS (
SELECT *,
ROW_NUMBER() OVER(PARTITION BY [Description],
Payment ORDER BY [Description]) rn
FROM <TABLE 2>)
SELECT t.[Description],
CASE
WHEN COUNT(DISTINCT c.rn) > 1 THEN SUM(c.Payment)
ELSE MAX(t.amount) - SUM(c.Payment)
END [Balance]
FROM cte c
JOIN <TABLE 1> t ON t.[Description] = c.[Description]
GROUP BY t.[Description];
Result :
Description Balance
Proc 1 10
Proc 2 40
Proc 3 0
Note, But you could also check with your sample data what if one Description has both Payments made one is duplicate and other is unique ??

Answer 2

You're maths isn't correct
Proc 2 = Tabl1 50, Tab2 20+ 20 = balance = 10
DECLARE #tab1 TABLE ([Description] NVARCHAR(15), Qty INT, Amount INT )
INSERT INTO #tab1
([Description], Qty, Amount)
SELECT 'Proc 1',1,100 UNION ALL
SELECT 'Proc 2',1,50 UNION ALL
SELECT 'Proc 3',1,60
DECLARE #tab2 TABLE ([Description] NVARCHAR(15), Payment INT)
INSERT INTO #tab2
([Description], Payment)
SELECT 'Proc 1', 60 UNION ALL
SELECT 'Proc 1', 30 UNION ALL
SELECT 'Proc 2', 20 UNION ALL
SELECT 'Proc 3', 60 UNION ALL
SELECT 'Proc 2', 20
SELECT
T.[Description]
, Balance = (T.Qty * T.Amount) - X.Payment
FROM #tab1 T
INNER JOIN
(
SELECT
T.[Description]
, Payment = SUM(T.Payment)
FROM #tab2 T
GROUP BY T.[Description]
) X ON X.[Description] = T.[Description]

Answer 3

Left joining from a derived table will solve your issue. Alternatively you could have done it the slightly longer route by doing aggregation in a cte and then a join
select t1.description
,t1.amount - t2.sum as balance
from table1 t1
left join
(select description
,sum(payment) as sum
from table2
group by description) t2 on t1.description = t2.description
Just some food for thought:
You want to use a left join instead of an inner join. Left join says if it's in this first table and not second, show null. Inner join says show it only if it is in BOTH tables. And I'm guessing you want to show the balance even if a payment hasn't been made (just a guess but that's up to your report).
Here's the rextester sample you can play with.