Table A drops rows in table B in SQL JOIN - sql

Question

I have two tables which I need joined. I have two select statements with the same columns except OrderValue and SalesTarget:
SELECT
b.TrnYear, b.TrnMonth, b.Branch, b.Salesperson, b.OrderValue
FROM
dbo.vw_jab_Consolidated_Orders as b
INNER JOIN
dbo.vw_jab_SalTargets as a ON a.Sequence2 = b.Salesperson
WHERE
b.TrnYear = '2017'
ORDER BY
a.TrnMonth
SELECT
a.TrnYear, a.TrnMonth, a.Sequence1, a.Sequence2, a.SalesTarget
FROM
dbo.vw_jab_SalTargets as a
WHERE
a.TrnYear = '2017' AND a.SequenceType = 'BR'
GROUP BY
a.TrnYear, a.TrnMonth, a.Sequence1, a.Sequence2, a.SalesTarget
My problem is that as there are months where there are no OrderValues, the SalesTarget value is being filtered out when I join the two tables as follows:
SELECT
a.TrnYear, SUM(a.SalesTarget) as SalesTarget, SUM(b.InvoicedSales) AS InvoicedSales
FROM
(SELECT
TrnYear, SUM(SalesTarget) AS SalesTarget
FROM
dbo.vw_jab_SalTargets
WHERE
(SequenceType = 'BR')
GROUP BY
SequenceType, TrnYear) AS A
LEFT JOIN
(SELECT
TrnYear, SUM(ActualSales) AS InvoicedSales
FROM
dbo.vw_jab_Consolidated_Sales
GROUP BY
TrnYear) AS b ON a.TrnYear = b.TrnYear
WHERE
a.TrnYear = '2017'
GROUP BY
a.TrnYear

Answer 1

You want a right outer join, instead of an inner join. For a right outer join, all columns of the right table are kept even if there is not a matching value in the left table.
Try this:
SELECT a.TrnYear, a.TrnMonth, SUM(a.SalesTarget) as SalesTarget, SUM(b.InvoicedSales) AS InvoicedSales
FROM (SELECT TrnYear, TrnMonth, SUM(SalesTarget) AS SalesTarget
FROM dbo.vw_jab_SalTargets
WHERE (SequenceType = 'BR')
GROUP BY SequenceType, TrnYear, TrnMonth) AS a RIGHT OUTER JOIN
(SELECT TrnYear, TrnMonth, SUM(ActualSales) AS InvoicedSales
FROM dbo.vw_jab_Consolidated_Sales
GROUP BY TrnYear, TrnMonth) AS b ON a.TrnYear = b.TrnYear and a.TrnMonth=b.TrnMonth
WHERE a.TrnYear = '2017'
GROUP BY a.TrnYear, a.TrnMonth
ORDER BY TrnMonth

Answer 2

You could switch to a RIGHT JOIN, but then if you have no InvoicedSales, you won't get any SalesTargets. I'd use a FULL OUTER JOIN like so:
SELECT COALESCE(a.TrnYear,b.TrnYear) AS TrnYear
,SUM(a.SalesTarget) AS SalesTarget
,SUM(b.InvoicedSales) AS InvoicedSales
FROM (
SELECT TrnYear
,SUM(SalesTarget) AS SalesTarget
FROM dbo.vw_jab_SalTargets
WHERE (SequenceType = 'BR')
GROUP BY SequenceType
,TrnYear
) AS a
FULL OUTER JOIN (
SELECT TrnYear
,SUM(ActualSales) AS InvoicedSales
FROM dbo.vw_jab_Consolidated_Sales
GROUP BY TrnYear
) AS b
ON a.TrnYear = b.TrnYear
WHERE a.TrnYear = '2017'
OR b.TrnYear = '2017'
GROUP BY COALESCE(a.TrnYear,b.TrnYear)
The COALESCE() is necessary so that the column will always return the TrnYear.