More help wit Sum, Pivot and so on...

SQLServerCentral is supported by Red Gate Software Ltd.

Popular Topics

Home

Members

Calendar

Who's On

Home » SQL Server 2005 » T-SQL (SS2K5) » More help wit Sum, Pivot and so on...

More help wit Sum, Pivot and so on...

Rate Topic

Display Mode

Topic Options

Author

Message

tota00

Posted Monday, March 04, 2013 6:37 AM

Forum Newbie

Group: General Forum Members
Last Login: Friday, March 08, 2013 4:18 AM
Points: 9, Visits: 18

Hi!

First thanks for the help I got a couple of weeks ago! Smile

Now to my new proble that is to change this statement a bit as it works halfway...

SELECT *, ISNULL(Denmark,0) + ISNULL(Finland,0) + ISNULL(Lithuania,0) + ISNULL(Norway,0) + ISNULL(Sweden,0)
AS Total
FROM
(SELECT DropRate,MarketingName, Manufacturer, Country
FROM ReportedNetInfoData) s PIVOT (SUM(DropRate)
FOR Country IN ([Denmark], [Finland], [Lithuania], [Norway], [Sweden]))P

The simplified table looks like this:

DropRate Country MarketingName Manufacturer
0.000000000000 Denmark Zoarmon Intel
0.000000000000 Denmark USB316 Sierra Wireless
0.000000000000 Denmark XP3300-AR1 (P25C005AA) Sonim
0.000000000000 Denmark 700C Intermec Technologies Corp.
4.762000000000 Denmark R6230GE BlackBerry
0.000000000000 Denmark SGH-S300 Samsung
0.000000000000 Denmark Treo 600 Palm

The result from the sql-statement is something like this:
(the null values under all countries except denmark depends on no data from them yet)

MarketingName Manufacturer Denmark Finland Lithuania Norway Sweden Total
5130 XpressMusic Nokia 153.725000000000 NULL NULL NULL NULL 153.725000000000
3310 Nokia 151.236000000000 NULL NULL NULL NULL 151.236000000000
1100 Nokia 151.206000000000 NULL NULL NULL NULL 151.206000000000
iPhone 4 Apple 144.371000000000 NULL NULL NULL NULL 144.371000000000
1208 Nokia 133.742000000000 NULL NULL NULL NULL 133.742000000000
iPhone 4S Apple 130.181000000000 NULL NULL NULL NULL 130.181000000000
X1-01 Nokia 129.931000000000 NULL NULL NULL NULL 129.931000000000
SGH-E250 Samsung 118.778000000000 NULL NULL NULL NULL 118.778000000000
6300 Nokia 112.905000000000 NULL NULL NULL NULL 112.905000000000
101, 1010 Nokia 110.319000000000 NULL NULL NULL NULL 110.319000000000

This is fine. I have the total drop rate however I am also looking for the average drop rate per country as well as the average drop rate for the total. (Total rate / # of countries per model)

Ie. IPhone 4 occurs Three times with droprate 1 an 2 and 3 for sweden and 2 and 4 and 6 for Norway then it should produce the average 2 for sweden and 4 for norway and total should be showing (countries + countries) divided by number of countries like (2+4)/2 in this case because it shouldn´t divide by countries that return null values when summaring and Dividing the total.

Hope I didn+t confuse you to much!

Thank´s in advance // TT

Post #1426212

Phil Parkin

Posted Monday, March 04, 2013 6:41 AM

SSCarpal Tunnel

Group: General Forum Members
Last Login: Yesterday @ 5:12 AM
Points: 4,226, Visits: 9,458

Please follow the link in my signature to get details of how to post sample DDL, data and desired results to get the best response to your question.

____________________________________________________________________________________________

Help us to help you. For better, quicker and more focused answers to your questions, consider following the advice in this link:

http://www.sqlservercentral.com/articles/Best+Practices/61537/

If you are asking for help and your post does not contain a question, you should expect responses which do not contain any answers. Put a question mark in there somewhere - it's not rocket science.

Post #1426216

tota00

Posted Monday, March 04, 2013 6:51 AM

Forum Newbie

Group: General Forum Members
Last Login: Friday, March 08, 2013 4:18 AM
Points: 9, Visits: 18

Phil Parkin (3/4/2013)

Please follow the link in my signature to get details of how to post sample DDL, data and desired results to get the best response to your question.

I´m sorry and will post a script for generating a table containing testdata! Sorry again!

// TT

Post #1426219

Phil Parkin

Posted Monday, March 04, 2013 6:54 AM

SSCarpal Tunnel

Group: General Forum Members
Last Login: Yesterday @ 5:12 AM
Points: 4,226, Visits: 9,458

tota00 (3/4/2013)

Phil Parkin (3/4/2013)

Please follow the link in my signature to get details of how to post sample DDL, data and desired results to get the best response to your question.

I´m sorry and will post a script for generating a table containing testdata! Sorry again!

// TT

You're pretty new here, so no need for apologies.

You will find that you get far more responses - with working and tested code - if you take the time to post your question in such a way.

____________________________________________________________________________________________

Help us to help you. For better, quicker and more focused answers to your questions, consider following the advice in this link:

http://www.sqlservercentral.com/articles/Best+Practices/61537/

If you are asking for help and your post does not contain a question, you should expect responses which do not contain any answers. Put a question mark in there somewhere - it's not rocket science.

Post #1426220

dwain.c

Posted Monday, March 04, 2013 6:21 PM

SSCrazy

Group: General Forum Members
Last Login: Yesterday @ 12:42 AM
Points: 2,338, Visits: 3,158

tota00 (3/4/2013)

This is fine. I have the total drop rate however I am also looking for the average drop rate per country as well as the average drop rate for the total. (Total rate / # of countries per model)

Ie. IPhone 4 occurs Three times with droprate 1 an 2 and 3 for sweden and 2 and 4 and 6 for Norway then it should produce the average 2 for sweden and 4 for norway and total should be showing (countries + countries) divided by number of countries like (2+4)/2 in this case because it shouldn´t divide by countries that return null values when summaring and Dividing the total.

Hope I didn+t confuse you to much!

Thank´s in advance // TT

You need to calculate your averages using a GROUP BY and then add these back to your PIVOT.

Here's a start on your sample data:

DECLARE @Phones TABLE
    (DropRate DECIMAL(38,12)
    ,Country VARCHAR(10)
    ,MarketingName VARCHAR(50)
    ,Manufacturer VARCHAR(50))

INSERT INTO @Phones
SELECT 0.000000000000,'Denmark','Zoarmon','Intel'
UNION ALL SELECT 0.000000000000,'Denmark','USB316','Sierra Wireless'
UNION ALL SELECT 0.000000000000,'Denmark','XP3300-AR1 (P25C005AA)','Sonim'
UNION ALL SELECT 0.000000000000,'Denmark','700C','Intermec Technologies Corp.'
UNION ALL SELECT 4.000000000000,'Denmark','R6230GE','BlackBerry'
UNION ALL SELECT 0.000000000000,'Denmark','SGH-S300','Samsung'
UNION ALL SELECT 0.000000000000,'Denmark','Treo 600','Palm'
UNION ALL SELECT 0.000000000000,'Sweden','Zoarmon','Intel'
UNION ALL SELECT 0.000000000000,'Sweden','USB316','Sierra Wireless'
UNION ALL SELECT 0.000000000000,'Sweden','XP3300-AR1 (P25C005AA)','Sonim'
UNION ALL SELECT 0.000000000000,'Sweden','700C','Intermec Technologies Corp.'
UNION ALL SELECT 2.000000000000,'Sweden','R6230GE','BlackBerry'
UNION ALL SELECT 0.000000000000,'Sweden','SGH-S300','Samsung'
UNION ALL SELECT 0.000000000000,'Sweden','Treo 600','Palm'
UNION ALL SELECT 0.000000000000,'Finland','Zoarmon','Intel'
UNION ALL SELECT 0.000000000000,'Finland','USB316','Sierra Wireless'
UNION ALL SELECT 0.000000000000,'Finland','XP3300-AR1 (P25C005AA)','Sonim'
UNION ALL SELECT 0.000000000000,'Finland','700C','Intermec Technologies Corp.'
UNION ALL SELECT 2.000000000000,'Finland','R6230GE','BlackBerry'
UNION ALL SELECT 0.000000000000,'Finland','SGH-S300','Samsung'
UNION ALL SELECT 0.000000000000,'Finland','Treo 600','Palm'

SELECT Country, Manufacturer, MarketingName, AvgDropRate=AVG(DropRate)
FROM @Phones
GROUP BY Country, Manufacturer, MarketingName

No loops! No CURSORs! No RBAR! Hoo-uh!

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?

Need to UNPIVOT? Why not CROSS APPLY VALUES instead?
Since random numbers are too important to be left to chance, let's generate some!
Are you too recursively challenged?
Splitting strings based on patterns can be fast!

Post #1426551

tota00

Posted Tuesday, March 05, 2013 12:21 AM

Forum Newbie

Group: General Forum Members
Last Login: Friday, March 08, 2013 4:18 AM
Points: 9, Visits: 18

You need to calculate your averages using a GROUP BY and then add these back to your PIVOT.

Here's a start on your sample data: "snip"

SELECT Country, Manufacturer, MarketingName, AvgDropRate=AVG(DropRate)
FROM @Phones
GROUP BY Country, Manufacturer, MarketingName

Thank´s dwain! Will start with this Smile

Best // Thomas

Post #1426613

tota00

Posted Tuesday, March 05, 2013 2:19 AM

Forum Newbie

Group: General Forum Members
Last Login: Friday, March 08, 2013 4:18 AM
Points: 9, Visits: 18

Okey!
I have come a bit on the way to a solution and have included a script for creating a test-table with sample dummy data in it!

The result from the sql-statement now works fine except the total sum result that has to be divided by the number of columns that do not return null values or is greater then zero so in this case with the sample data it should be divided by 3 but how do I solve that?

Sql statement:

SELECT *, ISNULL(Denmark,0) + ISNULL(Finland,0) + ISNULL(Lithuania,0) + ISNULL(Norway,0) + ISNULL(Sweden,0)  AS Total 
FROM (SELECT DropRate,MarketingName, Manufacturer, Country FROM tblTest) s PIVOT (AVG(DropRate)
FOR Country IN ([Denmark], [Finland], [Lithuania], [Norway], [Sweden]))P

Generates output:
MarketingName Manufacturer Denmark Finland Lithuania Norway Sweden Total
iPhone 3G Apple 1.408000000000 1.485000000000 0.000000000000 NULL 2.625000000000 5.518000000000

What I need is:
Should be: (difference is the division by three in this example)

SELECT *, (ISNULL(Denmark,0) + ISNULL(Finland,0) + ISNULL(Lithuania,0) + ISNULL(Norway,0) + ISNULL(Sweden,0)) / 3  AS Total 
FROM (SELECT DropRate,MarketingName, Manufacturer, Country FROM tblTest) s PIVOT (AVG(DropRate)
FOR Country IN ([Denmark], [Finland], [Lithuania], [Norway], [Sweden]))P

That generates output:
MarketingName Manufacturer Denmark Finland Lithuania Norway Sweden Total
iPhone 3G Apple 1.408000000000 1.485000000000 0.000000000000 NULL 2.625000000000 1.839333333333

But from time over time the amount of null or zero values in the columns will differ so I need intelligence in the sql-statement that takes care of this and this is above my knowledge for the moment :-(

And of course this is very simplified and I have my reason to include a country that is not represented in the sample data and so on Wink

Here you go with the sample table.

/****** Object:  Table [dbo].[tblTest]    Script Date: 2013-03-05 09:58:12 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
CREATE TABLE [dbo].[tblTest](
	[DropRate] [decimal](38, 12) NULL,
	[Country] [nchar](10) NULL,
	[MarketingName] [nvarchar](500) NULL,
	[Manufacturer] [nvarchar](200) NULL
) ON [PRIMARY]

GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.547000000000 AS Decimal(38, 12)), N'Denmark   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.274000000000 AS Decimal(38, 12)), N'Denmark   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.476000000000 AS Decimal(38, 12)), N'Finland   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.494000000000 AS Decimal(38, 12)), N'Finland   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(0.000000000000 AS Decimal(38, 12)), N'Lithuania ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.513000000000 AS Decimal(38, 12)), N'Denmark   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.431000000000 AS Decimal(38, 12)), N'Denmark   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(2.625000000000 AS Decimal(38, 12)), N'Sweden    ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.370000000000 AS Decimal(38, 12)), N'Denmark   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.384000000000 AS Decimal(38, 12)), N'Denmark   ', N'iPhone 3G', N'Apple')
GO
INSERT [dbo].[tblTest] ([DropRate], [Country], [MarketingName], [Manufacturer]) VALUES (CAST(1.337000000000 AS Decimal(38, 12)), N'Denmark   ', N'iPhone 3G', N'Apple')
GO

Post #1426652

tota00

Posted Tuesday, March 05, 2013 3:52 AM

Forum Newbie

Group: General Forum Members
Last Login: Friday, March 08, 2013 4:18 AM
Points: 9, Visits: 18

Problem solved! Smile

Here is the sql-statement that did the trick!

SELECT *, (ISNULL(Denmark,0) + ISNULL(Finland,0) + ISNULL(Lithuania,0) + ISNULL(Norway,0) + ISNULL(Sweden,0)) /
(CASE WHEN ISNULL(Denmark,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Finland,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Lithuania,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Norway,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Sweden,0)>0 THEN 1 ELSE 0 END) AS Total
FROM (SELECT DropRate,MarketingName, Manufacturer, Country FROM tblTest) s PIVOT (AVG(DropRate)
FOR Country IN ([Denmark], [Finland], [Lithuania], [Norway], [Sweden]))P

Many thanks to Visakh Murukes for the help!

Best // Thomas

Post #1426710

John_P

Posted Tuesday, March 05, 2013 8:57 AM

Forum Newbie

Group: General Forum Members
Last Login: Monday, May 13, 2013 1:02 PM
Points: 6, Visits: 37

Looking at Thomas' select statement made me think of one of my favorite functions, that (IMHO) doesn't get enough use: SIGN

Now, this only works here because there are no negative numbers for each of the countries. SIGN returns one of four values:
SIGN(X) = -1 when X < 0
SIGN(X) = 0 when X = 0
SIGN(X) = 1 when X = 1
SIGN(NULL) = NULL

The bottom part of the average calculation:
(CASE WHEN ISNULL(Denmark,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Finland,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Lithuania,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Norway,0)>0 THEN 1 ELSE 0 END +
CASE WHEN ISNULL(Sweden,0)>0 THEN 1 ELSE 0 END)

Can be replace with the following:
(SIGN(ISNULL(Denmark,0)) +SIGN(ISNULL(Finland,0)) + SIGN(ISNULL(Lithuania,0)) + SIGN(ISNULL(Norway,0)) + SIGN(ISNULL(Sweden,0))

Good Luck! John

Post #1426877

tota00

Posted Wednesday, March 06, 2013 12:35 AM

Forum Newbie

Group: General Forum Members
Last Login: Friday, March 08, 2013 4:18 AM
Points: 9, Visits: 18

John_P (3/5/2013)

Thank you John!
This was interesting.
By the way my solution above had to be changed a little bit to take care of division by zero problem (when all droprates in countries is null or equal to zero) but handled that by adding a -0.0000001 dummy value to the division operator.
As I only output a rounded 6 decimal value to the user it doesn´t affect anything.
(ugly workaround I know, but easy ;-(

Thank´s again! // Thomas

Post #1427210

« Prev Topic | Next Topic »

Permissions