Need help combining views; DDL included

I've solved this issue by creating 3 views but I'd rather do it in 1 SELECT
if possible.
Given my data I want to select duplicate securities based on the cusip field
in the Securities table where the cusip does not exist in the Positions
table. To rephrase I want duplicate securities that are not held.
Given my sample data I want to return one record with the cusip value 'E'
Thanks to anyone who could help.
if exists (select * from dbo.sysobjects where id =
object_id(N'[dbo].[FK_Positions_Securities]') and OBJECTPROPERTY(id,
N'IsForeignKey') = 1)
ALTER TABLE [dbo].[Positions] DROP CONSTRAINT FK_Positions_Securities
GO
if exists (select * from dbo.sysobjects where id =
object_id(N'[dbo].[Positions]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
drop table [dbo].[Positions]
GO
if exists (select * from dbo.sysobjects where id =
object_id(N'[dbo].[Securities]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
drop table [dbo].[Securities]
GO
CREATE TABLE [dbo].[Positions] (
[AccountID] [int] NOT NULL ,
[SecurityID] [int] NOT NULL ,
[Quantity] [int] NULL
) ON [PRIMARY]
GO
CREATE TABLE [dbo].[Securities] (
[SecurityID] [int] NOT NULL ,
[CUSIP] [char] (10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
) ON [PRIMARY]
GO
ALTER TABLE [dbo].[Positions] WITH NOCHECK ADD
CONSTRAINT [PK_Positions] PRIMARY KEY CLUSTERED
(
[AccountID],
[SecurityID]
) ON [PRIMARY]
GO
ALTER TABLE [dbo].[Securities] WITH NOCHECK ADD
CONSTRAINT [PK_Securities] PRIMARY KEY CLUSTERED
(
[SecurityID]
) ON [PRIMARY]
GO
ALTER TABLE [dbo].[Positions] ADD
CONSTRAINT [FK_Positions_Securities] FOREIGN KEY
(
[SecurityID]
) REFERENCES [dbo].[Securities] (
[SecurityID]
)
GO
INSERT INTO Securities (SecurityID,CUSIP) VALUES (1,'A')
INSERT INTO Securities (SecurityID,CUSIP) VALUES (2,'A')
INSERT INTO Securities (SecurityID,CUSIP) VALUES (3,'B')
INSERT INTO Securities (SecurityID,CUSIP) VALUES (4,'C')
INSERT INTO Securities (SecurityID,CUSIP) VALUES (5,'D')
INSERT INTO Securities (SecurityID,CUSIP) VALUES (6,'E')
INSERT INTO Securities (SecurityID,CUSIP) VALUES (7,'E')
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,1,10)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,2,15)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,1,20)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (5,4,10)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,3,10)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,3,10)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (3,5,10)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,2,10)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,5,10)
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,5,10)Try This:
Select CusIP, Count(*)
From Securities
Where SecurityID Not In
(Select SecurityID
From Positions)
Group By CusIP
Having Count(*) > 1
"Terri" wrote:

> I've solved this issue by creating 3 views but I'd rather do it in 1 SELEC
T
> if possible.
> Given my data I want to select duplicate securities based on the cusip fie
ld
> in the Securities table where the cusip does not exist in the Positions
> table. To rephrase I want duplicate securities that are not held.
> Given my sample data I want to return one record with the cusip value 'E'
> Thanks to anyone who could help.
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[FK_Positions_Securities]') and OBJECTPROPERTY(id,
> N'IsForeignKey') = 1)
> ALTER TABLE [dbo].[Positions] DROP CONSTRAINT FK_Positions_Securities
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Positions]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
> drop table [dbo].[Positions]
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Securities]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
> drop table [dbo].[Securities]
> GO
> CREATE TABLE [dbo].[Positions] (
> [AccountID] [int] NOT NULL ,
> [SecurityID] [int] NOT NULL ,
> [Quantity] [int] NULL
> ) ON [PRIMARY]
> GO
> CREATE TABLE [dbo].[Securities] (
> [SecurityID] [int] NOT NULL ,
> [CUSIP] [char] (10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] WITH NOCHECK ADD
> CONSTRAINT [PK_Positions] PRIMARY KEY CLUSTERED
> (
> [AccountID],
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Securities] WITH NOCHECK ADD
> CONSTRAINT [PK_Securities] PRIMARY KEY CLUSTERED
> (
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] ADD
> CONSTRAINT [FK_Positions_Securities] FOREIGN KEY
> (
> [SecurityID]
> ) REFERENCES [dbo].[Securities] (
> [SecurityID]
> )
> GO
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (1,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (2,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (3,'B')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (4,'C')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (5,'D')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (6,'E')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (7,'E')
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,1,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,2,15)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,1,20)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (5,4,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (3,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,2,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,5,10)
>
>|||What sulld happen in case that you add this to your sample data:
INSERT INTO Securities (SecurityID,CUSIP) VALUES (8,'E')
INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,8,10)
Would "E" still satisfy your request?
"Terri" wrote:

> I've solved this issue by creating 3 views but I'd rather do it in 1 SELEC
T
> if possible.
> Given my data I want to select duplicate securities based on the cusip fie
ld
> in the Securities table where the cusip does not exist in the Positions
> table. To rephrase I want duplicate securities that are not held.
> Given my sample data I want to return one record with the cusip value 'E'
> Thanks to anyone who could help.
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[FK_Positions_Securities]') and OBJECTPROPERTY(id,
> N'IsForeignKey') = 1)
> ALTER TABLE [dbo].[Positions] DROP CONSTRAINT FK_Positions_Securities
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Positions]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
> drop table [dbo].[Positions]
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Securities]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
> drop table [dbo].[Securities]
> GO
> CREATE TABLE [dbo].[Positions] (
> [AccountID] [int] NOT NULL ,
> [SecurityID] [int] NOT NULL ,
> [Quantity] [int] NULL
> ) ON [PRIMARY]
> GO
> CREATE TABLE [dbo].[Securities] (
> [SecurityID] [int] NOT NULL ,
> [CUSIP] [char] (10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] WITH NOCHECK ADD
> CONSTRAINT [PK_Positions] PRIMARY KEY CLUSTERED
> (
> [AccountID],
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Securities] WITH NOCHECK ADD
> CONSTRAINT [PK_Securities] PRIMARY KEY CLUSTERED
> (
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] ADD
> CONSTRAINT [FK_Positions_Securities] FOREIGN KEY
> (
> [SecurityID]
> ) REFERENCES [dbo].[Securities] (
> [SecurityID]
> )
> GO
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (1,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (2,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (3,'B')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (4,'C')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (5,'D')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (6,'E')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (7,'E')
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,1,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,2,15)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,1,20)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (5,4,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (3,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,2,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,5,10)
>
>|||Anyway,
If your asnwer is Yes:
Solution 1:
select
sec.CUSIP
from Positions pos right outer join Securities sec
on pos.SecurityID=sec.SecurityID
where pos.AccountID is null
group by sec.CUSIP
having count(sec.CUSIP)>1
Solution 2:
select
sec.CUSIP
from Securities sec
where sec.SecurityID not in
(
select pos.SecurityID from Positions pos
)
group by sec.CUSIP
having count(sec.CUSIP)>1
If your answer is NO:
select CUSIP
from Securities
where CUSIP not in
(
select sec.CUSIP
from Positions pos join Securities sec
on pos.SecurityID=sec.SecurityID
)
group by CUSIP
having count(CUSIP)>1
"Terri" wrote:

> I've solved this issue by creating 3 views but I'd rather do it in 1 SELEC
T
> if possible.
> Given my data I want to select duplicate securities based on the cusip fie
ld
> in the Securities table where the cusip does not exist in the Positions
> table. To rephrase I want duplicate securities that are not held.
> Given my sample data I want to return one record with the cusip value 'E'
> Thanks to anyone who could help.
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[FK_Positions_Securities]') and OBJECTPROPERTY(id,
> N'IsForeignKey') = 1)
> ALTER TABLE [dbo].[Positions] DROP CONSTRAINT FK_Positions_Securities
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Positions]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
> drop table [dbo].[Positions]
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Securities]') and OBJECTPROPERTY(id, N'IsUserTable') = 1)
> drop table [dbo].[Securities]
> GO
> CREATE TABLE [dbo].[Positions] (
> [AccountID] [int] NOT NULL ,
> [SecurityID] [int] NOT NULL ,
> [Quantity] [int] NULL
> ) ON [PRIMARY]
> GO
> CREATE TABLE [dbo].[Securities] (
> [SecurityID] [int] NOT NULL ,
> [CUSIP] [char] (10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] WITH NOCHECK ADD
> CONSTRAINT [PK_Positions] PRIMARY KEY CLUSTERED
> (
> [AccountID],
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Securities] WITH NOCHECK ADD
> CONSTRAINT [PK_Securities] PRIMARY KEY CLUSTERED
> (
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] ADD
> CONSTRAINT [FK_Positions_Securities] FOREIGN KEY
> (
> [SecurityID]
> ) REFERENCES [dbo].[Securities] (
> [SecurityID]
> )
> GO
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (1,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (2,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (3,'B')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (4,'C')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (5,'D')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (6,'E')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (7,'E')
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,1,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,2,15)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,1,20)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (5,4,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (3,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,2,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,5,10)
>
>|||Terri
See Itzik Ben-Gan's script about duplicates
CREATE TABLE #Demo (
idNo int identity(1,1),
colA int,
colB int
)
INSERT INTO #Demo(colA,colB) VALUES (1,6)
INSERT INTO #Demo(colA,colB) VALUES (1,6)
INSERT INTO #Demo(colA,colB) VALUES (2,4)
INSERT INTO #Demo(colA,colB) VALUES (3,3)
INSERT INTO #Demo(colA,colB) VALUES (4,2)
INSERT INTO #Demo(colA,colB) VALUES (3,3)
INSERT INTO #Demo(colA,colB) VALUES (5,1)
INSERT INTO #Demo(colA,colB) VALUES (8,1)
PRINT 'Table'
SELECT * FROM #Demo
PRINT 'Duplicates in Table'
SELECT * FROM #Demo
WHERE idNo IN
(SELECT B.idNo
FROM #Demo A JOIN #Demo B
ON A.idNo <> B.idNo
AND A.colA = B.colA
AND A.colB = B.colB)
PRINT 'Duplicates to Delete'
SELECT * FROM #Demo
WHERE idNo IN
(SELECT B.idNo
FROM #Demo A JOIN #Demo B
ON A.idNo < B.idNo -- < this time, not <>
AND A.colA = B.colA
AND A.colB = B.colB)
DELETE FROM #Demo
WHERE idNo IN
(SELECT B.idNo
FROM #Demo A JOIN #Demo B
ON A.idNo < B.idNo -- < this time, not <>
AND A.colA = B.colA
AND A.colB = B.colB)
PRINT 'Cleaned-up Table'
SELECT * FROM #Demo
DROP TABLE #Demo
"Terri" <terri@.cybernets.com> wrote in message
news:d3f1h9$ocg$1@.reader2.nmix.net...
> I've solved this issue by creating 3 views but I'd rather do it in 1
SELECT
> if possible.
> Given my data I want to select duplicate securities based on the cusip
field
> in the Securities table where the cusip does not exist in the Positions
> table. To rephrase I want duplicate securities that are not held.
> Given my sample data I want to return one record with the cusip value 'E'
> Thanks to anyone who could help.
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[FK_Positions_Securities]') and OBJECTPROPERTY(id,
> N'IsForeignKey') = 1)
> ALTER TABLE [dbo].[Positions] DROP CONSTRAINT FK_Positions_Securities
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Positions]') and OBJECTPROPERTY(id, N'IsUserTable') =
1)
> drop table [dbo].[Positions]
> GO
> if exists (select * from dbo.sysobjects where id =
> object_id(N'[dbo].[Securities]') and OBJECTPROPERTY(id, N'IsUserTable') =
1)
> drop table [dbo].[Securities]
> GO
> CREATE TABLE [dbo].[Positions] (
> [AccountID] [int] NOT NULL ,
> [SecurityID] [int] NOT NULL ,
> [Quantity] [int] NULL
> ) ON [PRIMARY]
> GO
> CREATE TABLE [dbo].[Securities] (
> [SecurityID] [int] NOT NULL ,
> [CUSIP] [char] (10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] WITH NOCHECK ADD
> CONSTRAINT [PK_Positions] PRIMARY KEY CLUSTERED
> (
> [AccountID],
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Securities] WITH NOCHECK ADD
> CONSTRAINT [PK_Securities] PRIMARY KEY CLUSTERED
> (
> [SecurityID]
> ) ON [PRIMARY]
> GO
> ALTER TABLE [dbo].[Positions] ADD
> CONSTRAINT [FK_Positions_Securities] FOREIGN KEY
> (
> [SecurityID]
> ) REFERENCES [dbo].[Securities] (
> [SecurityID]
> )
> GO
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (1,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (2,'A')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (3,'B')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (4,'C')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (5,'D')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (6,'E')
> INSERT INTO Securities (SecurityID,CUSIP) VALUES (7,'E')
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,1,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,2,15)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,1,20)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (5,4,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,3,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (3,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,2,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (1,5,10)
> INSERT INTO Positions (AccountID,SecurityID,Quantity ) VALUES (4,5,10)
>

Need a one to one relationship

Can anyone tell me how I can create a one to one relationship from the
following ddl? I feel like I'm close to having this working, but can't
quite get it. I think I'm not sure what I should do with the
PurchaseOrderItem.BuildID. The business rule that I am trying to infoce
is one DistributorNumber per Build as well as one DistributorNumber per
PurchaseOrder while maintaining a one to one relationship betweetn the
row in the BuildItem and PurchaseOrderItem table.
CREATE TABLE [BuildItem] (
[BuildID] int NOT NULL,
[DistributorNumber] varchar(30) DEFAULT ('') NOT NULL,
[DistributorID] int NOT NULL,
[Quantity] tinyint DEFAULT (0) NOT NULL,
[UnitGrams] decimal(7,2) DEFAULT (0) NOT NULL,
[UnitCostPrice] smallmoney DEFAULT (0) NOT NULL,
[UnitSalePrice] smallmoney DEFAULT (0) NOT NULL
)
GO
ALTER TABLE [BuildItem] ADD CONSTRAINT [PK_BuildItem]
PRIMARY KEY CLUSTERED ([BuildID], [DistributorNumber])
GO
ALTER TABLE [BuildItem] ADD CONSTRAINT [FK_BuildItem_Build]
FOREIGN KEY ([BuildID]) REFERENCES [Build] ([BuildID])
GO
CREATE TABLE [PurchaseOrderItem] (
[PurchaseOrderID] int NOT NULL,
[DistributorNumber] varchar(30) NOT NULL,
[BuildID] int,
[ItemDescription] varchar(100) DEFAULT ('') NOT NULL,
[QuantityOrdered] tinyint DEFAULT (0) NOT NULL,
[QuantityReceived] tinyint DEFAULT (0) NOT NULL,
[QuantityBackOrdered] tinyint DEFAULT (0) NOT NULL,
[UnitCost] smallmoney DEFAULT (0) NOT NULL,
)
GO
ALTER TABLE [PurchaseOrderItem] ADD CONSTRAINT [PK_PurchaseOrderItem]
PRIMARY KEY ([PurchaseOrderID], [DistributorNumber])
GO
ALTER TABLE [PurchaseOrderItem] ADD CONSTRAINT
[FK_PurchaseOrderItem_BuildItem]
FOREIGN KEY ([BuildID], [DistributorNumber]) REFERENCES [BuildItem]
([BuildID], [DistributorNumber])
GO
Regards,
Aaron1-to-1 relationships sometimes share the same PK, except that in the
"other" table it's a PK and FK
1:1 Customer to Address
Create Table Customer (
CustomerID INT IDENTITY,
CustomerName NVARCHAR(30)
PRIMARY KEY CLUSTERED CustomerID)
Create Table Address (
CustomerID INT REFERENCES Customer,
Address NVARCHAR(30)
PRIMARY KEY CustomerID )
The other way to enforce the 1:1 is to put a unique index on the "other"
table. In the last example, let's say you wanted the PK to be an
AddressID because you thought the requirements might change in the
future to allow more than one address. You could "temporarily" enforce
the 1:1 using:
Create Table Address (
AddressID INT IDENTITY,
CustomerID INT REFERENCES Customer,
Address NVARCHAR(30)
PRIMARY KEY AddressID )
Create Unique Clustered Index Address_IDX on dbo.Address(CustomerID)
David Gugick
Imceda Software
www.imceda.com

Need a kick in the right direction w/this query

Here is a link to the DDL since it would take up a lot of room here.
http://damageinc.org/DDL.html
After you use that to generate the table and some sample data, here is the
problem I need help with.
You will see the following data, if you use this query...
SELECT DISTINCT Apps.AppName, TestCases.TestCase , Tests.Attributes,
Reports.Result, Reports.ReportDate
FROM
Reports
LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
LEFT OUTER JOIN Apps ON TestCases.AppID = Apps.AppID
AppName TestCase Attribute
Result ReportDate
Test App 1 Run for 30 minutes Attribute 1 Fail
2006-06-25 19:58:31.800
Test App 1 Run for 30 minutes Attribute 1 Pass
2006-06-25 19:58:29.800
Test App 1 Run for 30 minutes Attribute 1 Pass
2006-06-25 19:58:30.800
Test App 1 Run for 45 minutes Attribute 2 Fail
2006-06-25 19:58:33.800
Test App 1 Run for 45 minutes Attribute 2 Fail
2006-06-25 19:58:34.800
Test App 1 Run for 45 minutes Attribute 2 Pass
2006-06-25 19:58:32.800
Test App 1 Run for 60 minutes Attribute 3 Fail
2006-06-25 19:58:36.863
Test App 1 Run for 60 minutes Attribute 3 Pass
2006-06-25 19:58:35.863
Test App 1 Run for 60 minutes Attribute 3 Pass
2006-06-25 19:58:37.863
What I need to get is the count most recent result per TestCase regardless
of attribute.
So I was trying to do it visually by using this query..
SELECT tc.TestCase, tc.mostRecent, tc.Result
FROM Reports rp
RIGHT JOIN
(
SELECT TestCases.TestCase ,Reports.Result, MAX ( Reports.ReportDate ) AS
mostRecent
FROM Reports
LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
WHERE ( (Tests.Attributes = 'Attribute 1') OR (Tests.Attributes =
'Attribute 2') OR (Tests.Attributes = 'Attribute 3') )
GROUP BY TestCases.TestCase, Result) tc ON (tc.mostRecent =
rp.ReportDate ) ORDER BY TestCAse
Which will give you this data...
TestCase mostRecent Result
Run for 30 minutes 2006-06-25 19:58:31.800 Fail
Run for 30 minutes 2006-06-25 19:58:30.800 Pass
Run for 45 minutes 2006-06-25 19:58:34.800 Fail
Run for 45 minutes 2006-06-25 19:58:32.800 Pass
Run for 60 minutes 2006-06-25 19:58:36.863 Fail
Run for 60 minutes 2006-06-25 19:58:37.863 Pass
Which as you can tell is giving me both the most recent pass & fail result,
where I want the most recent result regardless of Pass/Fail.
Which in turn, makes my count query wrong as well...
SELECT COUNT(DISTINCT tc.TestCase) As Count,
COUNT(DISTINCT CASE rp.Result WHEN 'Pass' THEN rp.ReportID ELSE NULL END) as
Pass,
COUNT(DISTINCT CASE rp.Result WHEN 'Fail' THEN rp.ReportID ELSE NULL END) as
Fail
FROM Reports rp
RIGHT JOIN
(
SELECT TestCases.TestCase ,Reports.Result, MAX ( Reports.ReportDate ) AS
mostRecent
FROM Reports
LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
WHERE ( (Tests.Attributes = 'Attribute 1') OR (Tests.Attributes =
'Attribute 2') OR (Tests.Attributes = 'Attribute 3') )
GROUP BY TestCases.TestCase, Result) tc ON (tc.mostRecent = rp.ReportDate )
Which will give me
Count Pass Fail
3 3 3
Instead of the result I am looking for of..
Count Pass Fail
3 1 2Select Count(*) As 'Count',
Sum(Case When Result = 'Pass' Then 1 Else 0 End) As 'Pass',
Sum(Case When Result = 'Fail' Then 1 Else 0 End) As 'Fail'
From Reports r
Where r.ReportDate In
(Select Max(r.ReportDate)
From Reports r
Inner Join TestCases tc On r.TestID = tc.TestCaseID
Inner Join Tests t On tc.TestCaseID = t.TestID
Inner Join Apps a On tc.AppID = a.AppID
Where t.Attributes In ('Attribute 1', 'Attribute 2', 'Attribute 3')
Group By a.AppName, tc.TestCase)
The above query assumes that there are no duplicates in Reports.ReportDate.
I changed your Left Outer Joins to Inner Joins since if for a Report row,
you don't have a TestCases row and a Tests row, then the Attributes column
will nave NULL in it, so the where condition would be false. So your Left
Outer Join gives the same result as an Inner Join, but Inner Joins are ofter
faster.
BTW, your DDL is not too long to post to this group (IMHO) and you will have
better luck if you just include it the post rather than a web site since
some people are reluctant to go to an unknown site.
Tom
"Lucas Graf" <lgraf2000@.comcast.net> wrote in message
news:uJYRRBNmGHA.4064@.TK2MSFTNGP02.phx.gbl...
> Here is a link to the DDL since it would take up a lot of room here.
> http://damageinc.org/DDL.html
> After you use that to generate the table and some sample data, here is the
> problem I need help with.
> You will see the following data, if you use this query...
>
> SELECT DISTINCT Apps.AppName, TestCases.TestCase , Tests.Attributes,
> Reports.Result, Reports.ReportDate
> FROM
> Reports
> LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
> LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
> LEFT OUTER JOIN Apps ON TestCases.AppID = Apps.AppID
> AppName TestCase Attribute Result
> ReportDate
> Test App 1 Run for 30 minutes Attribute 1 Fail
> 2006-06-25 19:58:31.800
> Test App 1 Run for 30 minutes Attribute 1 Pass
> 2006-06-25 19:58:29.800
> Test App 1 Run for 30 minutes Attribute 1 Pass
> 2006-06-25 19:58:30.800
> Test App 1 Run for 45 minutes Attribute 2 Fail
> 2006-06-25 19:58:33.800
> Test App 1 Run for 45 minutes Attribute 2 Fail
> 2006-06-25 19:58:34.800
> Test App 1 Run for 45 minutes Attribute 2 Pass
> 2006-06-25 19:58:32.800
> Test App 1 Run for 60 minutes Attribute 3 Fail
> 2006-06-25 19:58:36.863
> Test App 1 Run for 60 minutes Attribute 3 Pass
> 2006-06-25 19:58:35.863
> Test App 1 Run for 60 minutes Attribute 3 Pass
> 2006-06-25 19:58:37.863
>
> What I need to get is the count most recent result per TestCase regardless
> of attribute.
> So I was trying to do it visually by using this query..
> SELECT tc.TestCase, tc.mostRecent, tc.Result
> FROM Reports rp
> RIGHT JOIN
> (
> SELECT TestCases.TestCase ,Reports.Result, MAX ( Reports.ReportDate ) AS
> mostRecent
> FROM Reports
> LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
> LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
> WHERE ( (Tests.Attributes = 'Attribute 1') OR (Tests.Attributes =
> 'Attribute 2') OR (Tests.Attributes = 'Attribute 3') )
> GROUP BY TestCases.TestCase, Result) tc ON (tc.mostRecent =
> rp.ReportDate ) ORDER BY TestCAse
>
> Which will give you this data...
> TestCase mostRecent Result
> Run for 30 minutes 2006-06-25 19:58:31.800 Fail
> Run for 30 minutes 2006-06-25 19:58:30.800 Pass
> Run for 45 minutes 2006-06-25 19:58:34.800 Fail
> Run for 45 minutes 2006-06-25 19:58:32.800 Pass
> Run for 60 minutes 2006-06-25 19:58:36.863 Fail
> Run for 60 minutes 2006-06-25 19:58:37.863 Pass
> Which as you can tell is giving me both the most recent pass & fail
> result, where I want the most recent result regardless of Pass/Fail.
> Which in turn, makes my count query wrong as well...
> SELECT COUNT(DISTINCT tc.TestCase) As Count,
> COUNT(DISTINCT CASE rp.Result WHEN 'Pass' THEN rp.ReportID ELSE NULL END)
> as Pass,
> COUNT(DISTINCT CASE rp.Result WHEN 'Fail' THEN rp.ReportID ELSE NULL END)
> as Fail
> FROM Reports rp
> RIGHT JOIN
> (
> SELECT TestCases.TestCase ,Reports.Result, MAX ( Reports.ReportDate ) AS
> mostRecent
> FROM Reports
> LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
> LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
> WHERE ( (Tests.Attributes = 'Attribute 1') OR (Tests.Attributes =
> 'Attribute 2') OR (Tests.Attributes = 'Attribute 3') )
> GROUP BY TestCases.TestCase, Result) tc ON (tc.mostRecent =
> rp.ReportDate )
> Which will give me
>
> Count Pass Fail
> 3 3 3
>
> Instead of the result I am looking for of..
> Count Pass Fail
> 3 1 2
>
>|||Lucas
Thanks fro posting DDL
See if this helps you
SELECT TestCase,COUNT(CASE WHEN Result='Fail' THEN ReportDate END) AS
'Fail',
COUNT(CASE WHEN Result='Pass' THEN ReportDate END) AS 'Pass'
FROM
(
SELECT Apps.AppName, TestCases.TestCase , Tests.Attributes,
Reports.Result, Reports.ReportDate
FROM
Reports
LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
LEFT OUTER JOIN Apps ON TestCases.AppID = Apps.AppID
) AS Der GROUP BY TestCase
"Lucas Graf" <lgraf2000@.comcast.net> wrote in message
news:uJYRRBNmGHA.4064@.TK2MSFTNGP02.phx.gbl...
> Here is a link to the DDL since it would take up a lot of room here.
> http://damageinc.org/DDL.html
> After you use that to generate the table and some sample data, here is the
> problem I need help with.
> You will see the following data, if you use this query...
>
> SELECT DISTINCT Apps.AppName, TestCases.TestCase , Tests.Attributes,
> Reports.Result, Reports.ReportDate
> FROM
> Reports
> LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
> LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
> LEFT OUTER JOIN Apps ON TestCases.AppID = Apps.AppID
> AppName TestCase Attribute Result
> ReportDate
> Test App 1 Run for 30 minutes Attribute 1 Fail
> 2006-06-25 19:58:31.800
> Test App 1 Run for 30 minutes Attribute 1 Pass
> 2006-06-25 19:58:29.800
> Test App 1 Run for 30 minutes Attribute 1 Pass
> 2006-06-25 19:58:30.800
> Test App 1 Run for 45 minutes Attribute 2 Fail
> 2006-06-25 19:58:33.800
> Test App 1 Run for 45 minutes Attribute 2 Fail
> 2006-06-25 19:58:34.800
> Test App 1 Run for 45 minutes Attribute 2 Pass
> 2006-06-25 19:58:32.800
> Test App 1 Run for 60 minutes Attribute 3 Fail
> 2006-06-25 19:58:36.863
> Test App 1 Run for 60 minutes Attribute 3 Pass
> 2006-06-25 19:58:35.863
> Test App 1 Run for 60 minutes Attribute 3 Pass
> 2006-06-25 19:58:37.863
>
> What I need to get is the count most recent result per TestCase regardless
> of attribute.
> So I was trying to do it visually by using this query..
> SELECT tc.TestCase, tc.mostRecent, tc.Result
> FROM Reports rp
> RIGHT JOIN
> (
> SELECT TestCases.TestCase ,Reports.Result, MAX ( Reports.ReportDate ) AS
> mostRecent
> FROM Reports
> LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
> LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
> WHERE ( (Tests.Attributes = 'Attribute 1') OR (Tests.Attributes =
> 'Attribute 2') OR (Tests.Attributes = 'Attribute 3') )
> GROUP BY TestCases.TestCase, Result) tc ON (tc.mostRecent =
> rp.ReportDate ) ORDER BY TestCAse
>
> Which will give you this data...
> TestCase mostRecent Result
> Run for 30 minutes 2006-06-25 19:58:31.800 Fail
> Run for 30 minutes 2006-06-25 19:58:30.800 Pass
> Run for 45 minutes 2006-06-25 19:58:34.800 Fail
> Run for 45 minutes 2006-06-25 19:58:32.800 Pass
> Run for 60 minutes 2006-06-25 19:58:36.863 Fail
> Run for 60 minutes 2006-06-25 19:58:37.863 Pass
> Which as you can tell is giving me both the most recent pass & fail
> result, where I want the most recent result regardless of Pass/Fail.
> Which in turn, makes my count query wrong as well...
> SELECT COUNT(DISTINCT tc.TestCase) As Count,
> COUNT(DISTINCT CASE rp.Result WHEN 'Pass' THEN rp.ReportID ELSE NULL END)
> as Pass,
> COUNT(DISTINCT CASE rp.Result WHEN 'Fail' THEN rp.ReportID ELSE NULL END)
> as Fail
> FROM Reports rp
> RIGHT JOIN
> (
> SELECT TestCases.TestCase ,Reports.Result, MAX ( Reports.ReportDate ) AS
> mostRecent
> FROM Reports
> LEFT OUTER JOIN TestCases ON TestCases.TestCaseID = Reports.TestID
> LEFT OUTER JOIN Tests ON TestCases.TestCaseID = Tests.TestID
> WHERE ( (Tests.Attributes = 'Attribute 1') OR (Tests.Attributes =
> 'Attribute 2') OR (Tests.Attributes = 'Attribute 3') )
> GROUP BY TestCases.TestCase, Result) tc ON (tc.mostRecent =
> rp.ReportDate )
> Which will give me
>
> Count Pass Fail
> 3 3 3
>
> Instead of the result I am looking for of..
> Count Pass Fail
> 3 1 2
>
>|||Hi guys, thanks for the help so far.
I am really close now. One remaining problem, say there is no report s for
say testcase 3 (delete any reports w/the TestID of 3 in it).
Running the query will give me
Count Pass Fail
2 0 2
What i need is to show the total test cases even if there is no result(s)
for some, and the count of the ones where there are results for like so..
Count Pass Fail
3 0 2|||Select Count(*)
+ (Select Count(*) From
TestCases tc
Where Not Exists (Select 1 From Reports r Where r.TestID =
tc.TestCaseID)) As 'Count',
Sum(Case When Result = 'Pass' Then 1 Else 0 End) As 'Pass',
Sum(Case When Result = 'Fail' Then 1 Else 0 End) As 'Fail'
From Reports r
Where r.ReportDate In
(Select Max(r.ReportDate)
From Reports r
Inner Join TestCases tc On r.TestID = tc.TestCaseID
Inner Join Tests t On tc.TestCaseID = t.TestID
Inner Join Apps a On tc.AppID = a.AppID
Where t.Attributes In ('Attribute 1', 'Attribute 2', 'Attribute 3')
Group By a.AppName, tc.TestCase)
Tom
"Lucas Graf" <lgraf2000@.comcast.net> wrote in message
news:uEX$yZVmGHA.492@.TK2MSFTNGP05.phx.gbl...
> Hi guys, thanks for the help so far.
> I am really close now. One remaining problem, say there is no report s
> for say testcase 3 (delete any reports w/the TestID of 3 in it).
> Running the query will give me
> Count Pass Fail
> 2 0 2
>
> What i need is to show the total test cases even if there is no result(s)
> for some, and the count of the ones where there are results for like so..
> Count Pass Fail
> 3 0 2
>
>|||Thanks Tom "the master" Cooper!
I appreciate everyone's help as well, every time I post a question I get
another nugget of i nformation to store away.
"Tom Cooper" <tom.no.spam.please.cooper@.comcast.net> wrote in message
news:wvKdnRf8W_cuoT3ZnZ2dnUVZ_oKdnZ2d@.co
mcast.com...
> Select Count(*)
> + (Select Count(*) From
> TestCases tc
> Where Not Exists (Select 1 From Reports r Where r.TestID =
> tc.TestCaseID)) As 'Count',
> Sum(Case When Result = 'Pass' Then 1 Else 0 End) As 'Pass',
> Sum(Case When Result = 'Fail' Then 1 Else 0 End) As 'Fail'
> From Reports r
> Where r.ReportDate In
> (Select Max(r.ReportDate)
> From Reports r
> Inner Join TestCases tc On r.TestID = tc.TestCaseID
> Inner Join Tests t On tc.TestCaseID = t.TestID
> Inner Join Apps a On tc.AppID = a.AppID
> Where t.Attributes In ('Attribute 1', 'Attribute 2', 'Attribute 3')
> Group By a.AppName, tc.TestCase)
> Tom
> "Lucas Graf" <lgraf2000@.comcast.net> wrote in message
> news:uEX$yZVmGHA.492@.TK2MSFTNGP05.phx.gbl...
>|||Tom, maybe I could pick your brain one more time :)
For whatever reason on my actual tables I still can't get what I need :(
I have updated the DDL with the actual tables and a sample of table data of
what I am actually using.
http://damageinc.org/DDL.html (The DDL is too large to post, the message
gets kicked back to me)
AFter you enter that DDL
If you run the query..
SELECT Apps.AppName, Tests.TestCase, TestCases.Type
FROM
TestCases
LEFT OUTER JOIN Apps ON TestCases.AppID = Apps.ID
LEFT OUTER JOIN Tests ON TestCases.TestID = Tests.ID
WHERE
(
(
(Proj1 = '87' OR Proj2 = '87' OR Proj3 = '87' OR Proj4 = '87' OR Proj5 =
'87')
OR
(Proj1 = '88' OR Proj2 = '88' OR Proj3 = '88' OR Proj4 = '88' OR Proj5 =
'88')
)
AND (TestCases.Card = 'G71_D' OR TestCases.Card = 'G70_D' OR TestCases.Card
= 'G72_D' OR TestCases.Card = 'G73_D')
AND TestCases.OS = 'Windows Vista'
)
GROUP BY AppName, TestCase,TestCases.Type
You will see
AppName TestCase
Type
3D Mark 2003 Benchmark: 1600x1200x32 4xAA 8x Aniso D3D Benchmarks
3D Mark 2003 Benchmark: 1600x1200x32 4xAA 8x Aniso Games
3D Mark 2003 Benchmark: Default 2x AA 4x Aniso D3D
Benchmarks
3D Mark 2003 Benchmark: Default 2x AA 4x Aniso G ames
.
.
.
.
There are 97 tests there.
What I need is that 97 for the count, then the count of the most recent
pass/fail result(s) (if there is a result) for each of those tests.
Using
Select Count(*) As 'Count',
Sum(Case When Result = 'Pass' Then 1 Else 0 End) As 'Pass',
Sum(Case When Result = 'Fail' Then 1 Else 0 End) As 'Fail'
From Reports r
Where r.ReportDate In
(Select Max(r.ReportDate)
From Reports r
Inner Join TestCases tc On r.TestCaseID = tc.ID
Inner Join Tests t On tc.TestID = t.ID
Inner Join Apps a On tc.AppID = a.ID
WHERE
(
(
(tc.Proj1 = '87' OR tc.Proj2 = '87' OR tc.Proj3 = '87' OR tc.Proj4 = '87'
OR tc.Proj5 = '87')
OR
(tc.Proj1 = '88' OR tc.Proj2 = '88' OR tc.Proj3 = '88' OR tc.Proj4 = '88'
OR tc.Proj5 = '88')
)
AND (tc.Card = 'G71_D' OR tc.Card = 'G70_D' OR tc.Card = 'G72_D' OR tc.Card
= 'G73_D')
AND tc.OS = 'Windows Vista'
)
Group By a.AppName, t.TestCase)
Gives me
Count Pass Fail
25 2 23
Which is obviously wrong, since the Count should be 97.
I have to be overthinking something here...
========================================
====================================
========