Graph of top four products each month for last 6 months?
(Yeah I know I typ'od the title)
I need create a chart which displays the four highest selling products for each month in the last 6 months to be displayed in a report.
Ive managed to get all the products for each month for the last six months but im unsure how to filter so it displays only the top four for each month.
Ive done some programming but I thiinnk im going about this wrongly.
This is my code so far -
DECLARE @StartDate SMALLDATETIME -- Variables for the dashboard
-
DECLARE @EndDate SMALLDATETIME
-
-
SET @StartDate = DATEADD(s,0,DATEADD(mm, DATEDIFF(m,0,GETDATE())-6,0)) -- get todays month minus 6
-
SET @EndDate = GETDATE() -- get todays date
-
-
----
-
-
DECLARE @WhereDateStart SMALLDATETIME
-
DECLARE @WhereDateEnd SMALLDATETIME
-
-
SET @WhereDateStart = CONVERT(SMALLDATETIME,'01/' + LTRIM(STR(MONTH(@StartDate))) + '/' + LTRIM(STR(YEAR(@StartDate))))
-
-
SET @WhereDateEnd = DATEADD(s,-1,DATEADD(mm, DATEDIFF(m,0,@EndDate)+1,0))-1
-
-- we want each date/month minus six months for filter
-
-
SELECT -- Set up columns and graph output.
-
-
LEFT(Products.ProductId,14) AS ChartLabel,
-
Products.ProductId,
-
CONVERT(INT, SUM(ManufacturingOrders.QuantityMade)) AS QuantityMade
-
,
-
CONVERT(CHAR(10),
-
LTRIM(STR(YEAR(ManufacturingOrders.ReleaseDate))) + '/' +
-
LTRIM(STR(MONTH(ManufacturingOrders.ReleaseDate)))) AS ReleaseDate -- trim dates
-
FROM
-
ManufacturingOrders
-
INNER JOIN Products ON ManufacturingOrders.Product = Products.Product -- Get the product type
-
INNER JOIN Classifications ON Products.Classification = Classifications.Classification -- Get the Product number
-
WHERE
-
-
Products.ProductId IN (
-
-
SELECT Products.ProductId
-
-- sales code
-
FROM Products INNER JOIN Classifications ON Products.Classification = Classifications.Classification
-
-
WHERE Classifications.ClassificationId = 'BOOK'
-
)
-
-
AND ManufacturingOrders.ReleaseDate BETWEEN @WhereDateStart AND @WhereDateEnd
-
AND ManufacturingOrders.SystemType = 'B'
-
AND QuantityMade > 0
-
-
GROUP BY
-
-
Products.ProductId
-
,
-
YEAR(ManufacturingOrders.ReleaseDate)
-
,
-
MONTH(ManufacturingOrders.ReleaseDate)
-
ORDER BY
-
ReleaseDate
-
,
-
ProductId
-
This produces ChartLabel,ProductId,QunatityMade,ReleaseDate columns.
but its everything for each month, I just want the top 4 for each month to be sent to the graph.
Sorry, too lazy to read your code so I'll post something similar :)
Try doing something like this: -
select * from
-
(
-
select
-
rank = row_number() over (partition by month order by numberofsolditems desc)
-
month, numberofsolditems
-
from
-
(your subquery) q) z
-
where rank <= 4
-
Happy Coding!!!
~~ CK
12 2782
Sorry, too lazy to read your code so I'll post something similar :)
Try doing something like this: -
select * from
-
(
-
select
-
rank = row_number() over (partition by month order by numberofsolditems desc)
-
month, numberofsolditems
-
from
-
(your subquery) q) z
-
where rank <= 4
-
Happy Coding!!!
~~ CK
NeoPa 32,497
Expert Mod 16PB
What do you need help with?
Have you tried looking through it to see how you can apply these techniques to your situation?
Well I got this far last night but it doesnt work.
It produces a "Msg 8120, Level 16, State 1, Line 14
Column 'Products.ProductId' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause.
"
WHich is being caused by the inclusion of the SELECT product.productsId
But the function being proposed is one which I don't fully understand.
Specifically these dot functions such as the ".ss" I can only assume is to identify specific variables. I've never seen them before in any of the code I've researched (Remember I'm only about 16 hours into programming this language).
I have a database which contains information about stock across multiple tables and the point is to locate the top 4 best selling items on a per month basis over the last 6 months then plot them on a barograph. So what I am trying to do is even more complex for me than I am showing here. I can handle the graphics part I just need to figureout how to do what in any normal programming language would be a nested loop. -
DECLARE @StartDate SMALLDATETIME
-
DECLARE @EndDate SMALLDATETIME
-
-
SET @StartDate = DATEADD(s,0,DATEADD(mm, DATEDIFF(m,0,GETDATE())-6,0))
-
SET @EndDate = GETDATE()
-
-
-
DECLARE @WhereDateStart SMALLDATETIME
-
DECLARE @WhereDateEnd SMALLDATETIME
-
-
SET @WhereDateStart = CONVERT(SMALLDATETIME,'01/' + LTRIM(STR(MONTH(@StartDate))) + '/' + LTRIM(STR(YEAR(@StartDate))))
-
SET @WhereDateEnd = DATEADD(s,-1,DATEADD(mm, DATEDIFF(m,0,@EndDate)+1,0))-1
-
-
SELECT * FROM
-
(
-
SELECT
-
Products.ProductId,
-
CONVERT(INT, SUM(ManufacturingOrders.QuantityMade)) AS Quantity,
-
CONVERT(CHAR(10),
-
LTRIM(STR(YEAR(ManufacturingOrders.ReleaseDate))) + '/' +
-
LTRIM(STR(MONTH(ManufacturingOrders.ReleaseDate)))) AS ReleaseDate,
-
-
ROW_NUMBER()OVER (ORDER BY ReleaseDate,Quantity DESC)AS RowNumber,
-
RANK() OVER (ORDER BY ReleaseDate) AS Rank
-
-
FROM ManufacturingOrders
-
-
INNER JOIN Products ON ManufacturingOrders.Product = Products.Product
-
INNER JOIN Classifications ON Products.Classification = Classifications.Classification
-
-
WHERE
-
-
Classifications.ClassificationId = 'STOCK 332'
-
AND ManufacturingOrders.ReleaseDate BETWEEN @WhereDateStart
-
AND @WhereDateEnd AND ManufacturingOrders.SystemType = 'C'
-
-
-
Group By ReleaseDate,Quantity)ss
-
-
WHERE ss.RowNumber = ss.Rank or ss.RowNumber = ss.Rank+1
-
Conversely if I simplyfy it as much as possible and take CK9663's original code and simply insert what I think are the relevent fields producing this version; -
SELECT * FROM
-
(
-
SELECT
-
Products.ProductId,
-
CONVERT(INT, SUM(ManufacturingOrders.QuantityMade)) AS Quantity, -- calculate total sales
-
-
RANK = ROW_NUMBER() OVER (PARTITION BY MONTH ORDER BY Quantity DESC)
-
MONTH, Quantity
-
-
FROM ManufacturingOrders
-
-
INNER JOIN Products ON ManufacturingOrders.Product = Products.Product
-
INNER JOIN Classifications ON Products.Classification = Classifications.Classification
-
-
(
-
Classifications.ClassificationId = 'stock item 443'
-
AND ManufacturingOrders.ReleaseDate BETWEEN @WhereDateStart
-
AND @WhereDateEnd AND ManufacturingOrders.SystemType = 'C') q) z
-
WHERE RANK <= 4
-
It produces; "Msg 102, Level 15, State 1, Line 7
Incorrect syntax near 'MONTH'."
since the only thing I changed in that line of code was the quantity variable, I'm somwhat puzzled as to why there is a syntax error. All of the commands in that line (Except "Quantity") are reserved words.
NeoPa 32,497
Expert Mod 16PB
Let's take a bunch of steps backwards. I do not even see how your attempts match the suggestion, but as you say you're new to this.
A step at a time then, do you understand subqueries ( Subqueries in SQL) in SQL?
There are two levels of subquery in the proposed solution but, although there is a sort of one in one of your examples, it doesn't appear you have what is suggested in the original.
Well I do now, though my guess work was reasonably accurate. It provides a seconed layer of filtering to the original query.
I guess the two levels of sub query is the (sub query here) and the final line in the original example.
"where rank <= 4 "
NeoPa 32,497
Expert Mod 16PB ck9663: - select * from
-
(
-
select
-
rank = row_number() over (partition by month order by numberofsolditems desc)
-
month, numberofsolditems
-
from
-
(your subquery) q) z
-
where rank <= 4
Pretty well yes. Let me explain :
In line #7 (your subquery) q specifies your query and the result set is renamed (ALIASed) to [q]. This is equivalent to (your subquery) AS q.
From there the next level query (named [z] now) would be lines #2 through #7. - (
-
SELECT [Rank] = ROW_NUMBER() OVER (PARTITION BY [Month] ORDER BY [NumberOfSoldItems] DESC)
-
[Month]
-
, [NumberOfSoldItems]
-
FROM (your subquery) AS q) AS z
I think possibly an extra comma (,) may be required before [Month].
This implies that your query must provide data for [Month] and [NumberOfSoldItems].
Anyway, if we now call this [QueryZ] then we are left with : - SELECT *
-
FROM (QueryZ) AS [z]
-
WHERE [Rank] <= 4
Does that all start to make a little more sense?
Yes that makes a lot more sense as soon as you mentioned the "AS" keyword.
The "z" routeine I can leave as is, but the 'q' query needs to specify the particulatr brand of product from a related search.
This is the relationship that needs to be parsed - FROM ManufacturingOrders
-
-
INNER JOIN Products ON ManufacturingOrders.Product = Products.Product
-
INNER JOIN Classifications ON Products.Classification = Classifications.Classification
Which goes inside the query q brackets?
and continues with ; - Classifications.ClassificationId = 'stock item 443'
-
AND ManufacturingOrders.ReleaseDate BETWEEN @WhereDateStart
-
AND @WhereDateEnd AND ManufacturingOrders.SystemType = 'C'
?
I'm not stupid but the syntax of this language is freakish in comparision with those that I have played with in the past. I guess it doesnt help to only have been supplied with one book which is Sams teach yourself Microsoft SQL Server T-SQL.
Any suggestions for a quick rummage in amazon.co.uk for a better book than this. It is ... well erm.. brief.. :)
NeoPa 32,497
Expert Mod 16PB MicroBlitz:
Yes that makes a lot more sense as soon as you mentioned the "AS" keyword.
I so agree. It's an optional keyword and I always use it myself as I feel clarity is so important, especially when reading someone-else's code.
Generally, I find that where subqueries become nested and complicated it is worthwhile to develop each in turn, starting from the innermost and proceeding to the overall. Clearly it's important to have a good understanding of what precisely is required at each level, but once the innermost one is developed and working you can think of it in its own little black box (within parentheses) and know that it does its job. The last SQL of post #9 for instance, is easy to work on if you know you can rely on (QueryZ) to do all that's required of it. MicroBlitz:
I guess it doesnt help to only have been supplied with one book which is Sams teach yourself Microsoft SQL Server T-SQL.
Any suggestions for a quick rummage in amazon.co.uk for a better book than this. It is ... well erm.. brief.. :)
Not really, but I'm finding out all I need from the Context-Sensitive Help system. I find it is invaluable to me. I'm in a position not far different from yours in fact. I may well have a few years of Jet SQL experience, but T-SQL I'm picking up as I go along (after dipping in to T-SQL 2000 at a very basic level once or twice over the last few years). Not a day goes by at this time without my nose deeply buried in the Help system. It answers almost all of my questions well. If it weren't an electronic resource it would certainly be well thumbed by now, and I've only been even looking at this in the last month or two.
Anyway, good luck with your learning. Whichever way you find works best.
This is the final version I added to the code. It's not the same method as the suggestion above but I managed to get that way working after I got the report to function properly. If I can find the method detailed here I will post it for comparison - Microblitz -
DECLARE @Classification bigint
-
DECLARE @StartDate smalldatetime
-
DECLARE @EndDate smalldatetime
-
DECLARE @MonthEndDate smalldatetime
-
DECLARE @TempDate smalldatetime
-
-
SET @Classification = (SELECT Classification FROM Classifications WHERE Classifications.ClassificationId = 'STOCKITEM')
-
SET @StartDate = CONVERT(smalldatetime,'01 ' + RIGHT(CONVERT(nvarchar,getdate(),106),8))
-
SET @EndDate = DATEADD(d,-1,@StartDate)
-
SET @StartDate = DATEADD(m,-6,@StartDate)
-
-
CREATE TABLE #DETAILS
-
(
-
SalesYear int,
-
SalesMonth int,
-
ProductId nvarchar(100),
-
Quantity decimal(17,5)
-
)
-
-
SET @TempDate = @StartDate
-
-
WHILE @TempDate < @EndDate
-
BEGIN
-
SET @MonthEndDate = DATEADD(d,-1,DATEADD(m,1,@TempDate))
-
-
-
INSERT INTO #DETAILS
-
(
-
SalesYear,
-
SalesMonth,
-
ProductId,
-
Quantity
-
)
-
SELECT TOP 4
-
YEAR(@TempDate),
-
DATEPART(m,@TempDate),
-
Products.ProductId,
-
SUM(SalesInvoiceItems.Quantity) AS QuantityInvoiced
-
FROM SalesInvoiceItems
-
INNER JOIN SalesInvoices ON SalesInvoiceItems.SalesInvoice = SalesInvoices.SalesInvoice
-
AND dbo.wfn_GetSimpleDate(SalesInvoices.EffectiveDate) BETWEEN @TempDate AND @MonthEndDate
-
AND SalesInvoices.SystemType = 'F'
-
AND SalesInvoices.SourceType = 'I'
-
INNER JOIN Products ON SalesInvoiceItems.Product = Products.Product
-
AND Products.Classification = @Classification
-
GROUP BY
-
Products.ProductId
-
ORDER BY
-
QuantityInvoiced DESC
-
-
-
SET @TempDate = DATEADD(m,1,@TempDate)
-
END
-
-
SELECT *
-
FROM #DETAILS
-
-
DROP TABLE #DETAILS
-
The alternative code using the technique as explained above by ck9663 for comparison. -
DECLARE @Classification bigint
-
SET @Classification = (SELECT Classification FROM Classifications WHERE Classifications.ClassificationId = 'STOCKITEM')
-
-
select * from
-
(
-
select
-
Dtbl.SalesYear
-
Dtbl.SalesMonth,
-
Dtbl.ProductId,
-
row_number() over (partition by SalesYear,SalesMonth order by QuantityInvoiced desc) AS Ranking,
-
Dtbl.QuantityInvoiced
-
from
-
(
-
SELECT
-
YEAR(SalesInvoices.EffectiveDate) AS SalesYear,
-
DATEPART(m,SalesInvoices.EffectiveDate) AS SalesMonth,
-
Products.ProductId,
-
SUM(SalesInvoiceItems.Quantity) AS QuantityInvoiced
-
FROM SalesInvoiceItems
-
INNER JOIN SalesInvoices ON SalesInvoiceItems.SalesInvoice = SalesInvoices.SalesInvoice
-
AND dbo.wfn_GetSimpleDate(SalesInvoices.EffectiveDate) BETWEEN '01 Mar 2010' AND '31 Aug 2010'
-
AND SalesInvoices.SystemType = 'F'
-
AND SalesInvoices.SourceType = 'I'
-
INNER JOIN Products ON SalesInvoiceItems.Product = Products.Product
-
AND Products.Classification = @Classification
-
GROUP BY
-
YEAR(SalesInvoices.EffectiveDate),
-
DATEPART(m,SalesInvoices.EffectiveDate),
-
Products.ProductId
-
) AS Dtbl
-
) As DTbl2
-
where Ranking <= 4
Post your reply Sign in to post your reply or Sign up for a free account.
Similar topics
16 posts
views
Thread by sandy |
last post: by
|
1 post
views
Thread by Finlay |
last post: by
|
9 posts
views
Thread by Robin Tucker |
last post: by
| | | | |
11 posts
views
Thread by Harlin Seritt |
last post: by
| | | | | | | | | | | | |