Question

Here is the thing, I have table like this

First one

ID1 ID2 ID3

Second

ID1 ID2 ID3 month, value

What I need is to get values for three months for each combination of these 3 IDs, and I went this way:

SELECT a.ID1, a.ID2, a.ID3, b.value, c.value, d.value
FROM Table1 a

LEFT JOIN Table2 b
ON a.ID1 = b.ID1
AND a.ID2 = b.ID2
AND a.ID3 = b.ID3
WHERE month=2


LEFT JOIN Table2 c
ON a.ID1 = c.ID1
AND a.ID2 = c.ID2
AND a.ID3 = c.ID3
WHERE month=3


LEFT JOIN Table2 d
ON a.ID1 = d.ID1
AND a.ID2 = d.ID2
AND a.ID3 = d.ID3
WHERE month=4

And I got values multiplied.

Any idea why this is case, or any suggestions for other approaches?

Thanks.

Answer 1

Assuming you want the next 3 months (next month and the two following months) based on today’s date, you can do this (you can also set @base to any specific date to the get the three following months):

DECLARE @base DATETIME = GETDATE();

;WITH x AS
(
    SELECT a.ID1, a.ID2, a.ID3, 
      m = DATEADD(MONTH, b.[Month]-1, RTRIM(b.[Year])+'0101'), b.value
    FROM dbo.Table1 AS a INNER JOIN dbo.Table2 AS b 
    ON a.ID1 = b.ID1 AND a.ID2 = b.ID2 AND a.ID3 = b.ID3
), y AS
(
    SELECT ID1, ID2, ID3, m, md = DATEDIFF(MONTH, @base, m), value
    FROM x
    WHERE DATEDIFF(MONTH, @base, m) BETWEEN 1 AND 3
)
SELECT ID1, ID2, ID3, 
    BValue = SUM(CASE WHEN md = 1 THEN value END),
    CValue = SUM(CASE WHEN md = 2 THEN value END),
    DValue = SUM(CASE WHEN md = 3 THEN value END)
FROM y 
GROUP BY ID1, ID2, ID3;