December 12, 2013 at 8:10 am
I have my sql tables and query as shown below :
CREATE TABLE #ABC([Year] INT, [Month] INT, Stores INT);
CREATE TABLE #DEF([Year] INT, [Month] INT, SalesStores INT);
CREATE TABLE #GHI([Year] INT, [Month] INT, Products INT);
INSERT #ABC VALUES (2013,1,1);
INSERT #ABC VALUES (2013,1,2);
INSERT #ABC VALUES (2013,2,3);
INSERT #DEF VALUES (2013,1,4);
INSERT #DEF VALUES (2013,1,5);
INSERT #DEF VALUES (2013,2,6);
INSERT #GHI VALUES (2013,1,7);
INSERT #GHI VALUES (2013,1,8);
INSERT #GHI VALUES (2013,2,9);
INSERT #GHI VALUES (2013,3,10);
My current query is
I have @Year and @Month as parameters , both integers , example @Year = '2013' , @Month = '11'
SELECT T.[Year],
T.[Month]
-- select the sum for each year/month combination using a correlated subquery (each result from the main query causes another data retrieval operation to be run)
,
(SELECT SUM(Stores)
FROM #ABC
WHERE [Year] = T.[Year]
AND [Month] = T.[Month]) AS [Sum_Stores],
(SELECT SUM(SalesStores)
FROM #DEF
WHERE [Year] = T.[Year]
AND [Month] = T.[Month]) AS [Sum_SalesStores],
(SELECT SUM(Products)
FROM #GHI
WHERE [Year] = T.[Year]
AND [Month] = T.[Month]) AS [Sum_Products]
FROM (
-- this selects a list of all possible dates.
SELECT [Year],
[Month]
FROM #ABC where [Year] = @Year and [Month] = @Month
UNION
SELECT [Year],
[Month]
FROM #DEF where [Year] = @Year and [Month] = @Month
UNION
SELECT [Year],
[Month]
FROM #GHI where [Year] = @Year and [Month] = @Month) AS T;
Which returns
+------+-------+------------+-----------------+--------------+
| Year | Month | Sum_Stores | Sum_SalesStores | Sum_Products |
+------+-------+------------+-----------------+--------------+
| 2013 | | | | |
| 2013 | | | | |
| 2013 | | | | |
+------+-------+------------+-----------------+--------------+
What I want to do is to add more columns to the query which show the difference from the last month. as shown below. Example : The Diff beside the Sum_Stores shows the difference in the Sum_Stores from last month to this month.
Something like this :
+------+-------+------------+-----------------+-----|-----|---+-----------------
| Year | Month | Sum_Stores |Diff | Sum_SalesStores |Diff | Sum_Products |Diff|
+------+-------+------------+-----|------------+----|---- |----+--------------|
| 2013 | | | | | | | |
| 2013 | | | | | | | |
| 2013 | | | | | | | |
+------+-------+------------+-----|------------+--- |-----|----+---------| ----
Can anyone tell me how I can modify this to achive my goal.
December 12, 2013 at 8:29 am
Have you looked at the new LAG and LEAD windowing functions?
John
December 12, 2013 at 8:52 am
First of all, you should avoid using integers to store dates, even if they are just partial as month and year. You get in lots of trouble when calculating change of years (in this case).
Here's a possible solution but something else might be better.
--Parameters
DECLARE @Year int = 2013,
@Month int = 2;
--Variables for previous dates
DECLARE@PrevYear int,
@PrevMonth int;
SELECT @PrevYear = YEAR( DATEADD( MONTH, @Month - 2, CAST( CAST( @Year AS CHAR(4)) AS datetime))),
@PrevMonth = MONTH( DATEADD( MONTH, @Month - 2, CAST( CAST( @Year AS CHAR(4)) AS datetime)))
SELECT @Year AS [Year],
@Month AS [Month],
MAX( CASE WHEN [Month] = @Month AND [Description] = 'Stores' THEN Value END) AS [SumStores],
MAX( CASE WHEN [Month] = @Month AND [Description] = 'Stores' THEN Value END) -
MAX( CASE WHEN [Month] = @PrevMonth AND [Description] = 'Stores' THEN Value END) AS [Diff],
MAX( CASE WHEN [Month] = @Month AND [Description] = 'SalesStores' THEN Value END) AS [SumStores],
MAX( CASE WHEN [Month] = @Month AND [Description] = 'SalesStores' THEN Value END) -
MAX( CASE WHEN [Month] = @PrevMonth AND [Description] = 'SalesStores' THEN Value END) AS [Diff],
MAX( CASE WHEN [Month] = @Month AND [Description] = 'Products' THEN Value END) AS [SumStores],
MAX( CASE WHEN [Month] = @Month AND [Description] = 'Products' THEN Value END) -
MAX( CASE WHEN [Month] = @PrevMonth AND [Description] = 'Products' THEN Value END) AS [Diff]
FROM (
-- this selects a list of all possible dates.
SELECT [Year],
[Month],
SUM(Stores) Value,
'Stores' Description
FROM #ABC
where ([Year] = @Year and [Month] = @Month)
OR ([Year] = @PrevYear and [Month] = @PrevMonth)
GROUP BY [Year], [Month]
UNION
SELECT [Year],
[Month],
SUM(SalesStores),
'SalesStores' Description
FROM #DEF
where ([Year] = @Year and [Month] = @Month)
OR ([Year] = @PrevYear and [Month] = @PrevMonth)
GROUP BY [Year], [Month]
UNION
SELECT [Year],
[Month],
SUM(Products),
'Products' Description
FROM #GHI
where ([Year] = @Year and [Month] = @Month)
OR ([Year] = @PrevYear and [Month] = @PrevMonth)
GROUP BY [Year], [Month]) AS T
;
December 12, 2013 at 6:01 pm
John Mitchell-245523 (12/12/2013)
Have you looked at the new LAG and LEAD windowing functions?John
I agree with John that LAG would be pretty good for this:
WITH Aggregates AS
(
SELECT [Year], [Month]
,SumStores=ISNULL(SUM(Stores), 0)
,SumSalesStores=ISNULL(SUM(SalesStores), 0)
,SumProducts=ISNULL(SUM(Products), 0)
,[Date]=DATEADD(month, [Month]-1, CAST([Year] AS CHAR(4)))
FROM
(
SELECT [Year], [Month], Stores, SalesStores=NULL, Products=NULL
FROM #ABC
UNION ALL
SELECT [Year], [Month], NULL, SalesStores, NULL
FROM #DEF
UNION ALL
SELECT [Year], [Month], NULL, NULL, Products
FROM #GHI
) a
GROUP BY [Year], [Month]
)
SELECT [Year], [Month]
,SumStores
,DiffStores=CASE
WHEN DATEDIFF(month, [Date], LastMonth) = -1
THEN SumStores-LastStores
WHEN LastMonth IS NULL THEN NULL
ELSE SumStores END
,SumSalesStores
,DiffSalesStores=CASE
WHEN DATEDIFF(month, [Date], LastMonth) = -1
THEN SumStores-LastSalesStores
WHEN LastMonth IS NULL THEN NULL
ELSE SumStores END
,SumProducts
,DiffProducts=CASE
WHEN DATEDIFF(month, [Date], LastMonth) = -1
THEN SumStores-LastProducts
WHEN LastMonth IS NULL THEN NULL
ELSE SumStores END
FROM
(
SELECT [Year], [Month], [Date], SumStores, SumSalesStores, SumProducts
,LastMonth=LAG([Date], 1) OVER (ORDER BY [Date])
,LastStores=LAG(SumStores, 1) OVER (ORDER BY [Date])
,LastSalesStores=LAG(SumStores, 1) OVER (ORDER BY [Date])
,LastProducts=LAG(SumStores, 1) OVER (ORDER BY [Date])
FROM Aggregates
) a;
That should be close to what you need.
My thought question: Have you ever been told that your query runs too fast?
My advice:
INDEXing a poor-performing query is like putting sugar on cat food. Yeah, it probably tastes better but are you sure you want to eat it?
The path of least resistance can be a slippery slope. Take care that fixing your fixes of fixes doesn't snowball and end up costing you more than fixing the root cause would have in the first place.
Need to UNPIVOT? Why not CROSS APPLY VALUES instead?[/url]
Since random numbers are too important to be left to chance, let's generate some![/url]
Learn to understand recursive CTEs by example.[/url]
[url url=http://www.sqlservercentral.com/articles/St
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply