SQL Server query to generate dynamic columns - sql-server

I have two tables - one is state and other one is for Job titles.
I want to write a query which will output me something like this:-
Job titles State name1 State name2
Job title1 200 300
Job title2 500 600
How can I write this query in SQL Server.

I have no idea what your schema looks like, but it sounds like you want to change it to have three tables: JobTitle, State, and JobTitle_State_Salary. That way you're not repeating either job titles or states in order to tie to salary.
However, addressing the problem as written (and making the assumption that salary travels with state), something like this should do the trick:
WITH [CTE] AS
(
SELECT [Title], [State], [Salary]
FROM [JobTitle]
INNER JOIN [StateSalary]
ON [JobTitle].[ID] = [StateSalary].[JobTitleID]
)
SELECT
[Title], [State name1], [State name2]
FROM
[CTE]
PIVOT
(
MAX([Salary])
FOR [State] IN ([State name1], [State name2])
) AS [P]
SQLFiddle example here.

As zimdanen said, it would be difficult to write the query without knowing the exact table structure.
I assumed for structure that JobTitle table has JobTitleId and JobTitle fields, and that *state_salary* table has JobTitleId, State, and Salary as fields.
USING PIVOT TABLE
SELECT * FROM (
SELECT A.JOB_TITLE,B.STATE,B.SALARY FROM dbo.JOB_TITLE A INNER JOIN dbo.STATE_SALARY B
ON A.JOB_TITLE_ID = B.JOB_TITLE_ID)AS SOURCE_TABLE PIVOT
(SUM(SALARY) FOR STATE IN (STATE1,STATE2)) AS PivotTable
Without Pivot Table
SELECT A.JOB_TITLE,SUM(CASE WHEN B.STATE = 'State1' THEN B.SALARY ELSE 0 END) STATE1, SUM(CASE WHEN B.STATE = 'State2' THEN B.SALARY ELSE 0 END) STATE2
FROM dbo.JOB_TITLE A INNER JOIN dbo.STATE_SALARY B
ON A.JOB_TITLE_ID = B.JOB_TITLE_ID
GROUP BY A.JOB_TITLE

If you are using sqlserver 2005 or greater you can try using pivot table

Related

Repeat Customers with multiple purchases on the same day counts a 1

I am trying to wrap my head around this problem. I was asked to create a report that show repeat customers in our database.
One of the requirements is if a customer has more than 1 order on a specific date, it would only count as 1.
Then if they have more than 1 purchase date, they would then count as a repeat customer.
Searching on here, I found this which works for finding the Customers with more then 1 purchase on a specific purchase date.
SELECT DISTINCT s.[CustomerName], s.PurchaseDate
FROM Reports.vw_Repeat s WHERE s.PurchaseDate <> ''
GROUP BY s.[CustomerName] , cast(s.PurchaseDate as date)
HAVING COUNT(*) > 1;
This MSSQL code works like it should, by showing customers who had more than 1 purchase on the same date.
My problem is what would the best approach be to Join this into another query (this is where i need help) that then shows a complete repeat customer list where customers with more than 1 purchase would be returned.
I am using MSSQL. Any help would be greatly appreciated.
You're close, you need to move distinct into your having clause because you want to include only customers that have more than 1 distinct purchase date.
Also, only group by the customer id because the different dates have to be part of the same group for count distinct to work.
SELECT s.[CustomerName], COUNT(distinct cast(s.PurchaseDate as date))
FROM Reports.vw_Repeat s WHERE s.PurchaseDate <> ''
GROUP BY s.[CustomerName]
HAVING COUNT(distinct cast(s.PurchaseDate as date)) > 1;
If you want to pass a parameter to a query and join the result, that's what table-valued functions are for. When you join it, you use CROSS APPLY or OUTER APPLY instead of an INNER JOIN or a LEFT JOIN.
Also, I think this goes without saying, but when you check if PurchaseDate is empty:
WHERE s.PurchaseDate <> ''
Could be issues there... it implies it's a varchar field instead of a datetime (yes?) and doesn't handle null values. You might, at least, want to replace that with ISNULL(s.PurchaseDate, '') <> ''. If it's actually a datetime, use IS NOT NULL instead of <> ''.
(Edited to add sample data and DDL statements. I recommend adding these to SQL posts to assist answerers. Also, I made purchasedate a varchar instead of a datetime because of the string comparison in the query.)
https://technet.microsoft.com/en-us/library/ms191165(v=sql.105).aspx
CREATE TABLE company (company_name VARCHAR(25))
INSERT INTO company VALUES ('Company1'), ('Company2')
CREATE TABLE vw_repeat (customername VARCHAR(25), purchasedate VARCHAR(25), company VARCHAR(25))
INSERT INTO vw_repeat VALUES ('Cust1', '11/16/2017', 'Company1')
INSERT INTO vw_repeat VALUES ('Cust1', '11/16/2017', 'Company1')
INSERT INTO vw_repeat VALUES ('Cust2', '11/16/2017', 'Company2')
CREATE FUNCTION [dbo].tf_customers
(
#company varchar(25)
)
RETURNS TABLE AS RETURN
(
SELECT s.[CustomerName], cast(s.PurchaseDate as date) PurchaseDate
FROM vw_Repeat s
WHERE s.PurchaseDate <> '' AND s.Company = #company
GROUP BY s.[CustomerName] , cast(s.PurchaseDate as date)
HAVING COUNT(*) > 1
)
GO
SELECT *
FROM company c
CROSS APPLY tf_customers(c.company_name)
First thanks to everyone for the help.
#MaxSzczurek suggested I use table-valued functions. After looking into this more, I ended up using just a temporary table first to get the DISTINCT purchase dates for each Customer. I then loaded that into another temp table RIGHT JOINED to the main table. This gave me the result I was looking for. Its a little(lot) ugly, but it works.

Multiple Select against one CTE

I have a CTE query filtering a table Student
Student
(
StudentId PK,
FirstName ,
LastName,
GenderId,
ExperienceId,
NationalityId,
CityId
)
Based on a lot filters (multiple cities, gender, multiple experiences (1, 2, 3), multiple nationalites), I create a CTE by using dynamic sql and joining the student table with a user defined tables (CityTable, NationalityTable,...)
After that I have to retrieve the count of student by each filter like
CityId City Count
NationalityId Nationality Count
Same thing the other filter.
Can I do something like
;With CTE(
Select
FROM Student
Inner JOIN ...
INNER JOIN ....)
SELECT CityId,City,Count(studentId)
FROm CTE
GROUP BY CityId,City
SELECT GenderId,Gender,Count
FROM CTE
GROUP BY GenderId,Gender
I want to something like what LinkedIn is doing with search(people search,job search)
http://www.linkedin.com/search/fpsearch?type=people&keywords=sales+manager&pplSearchOrigin=GLHD&pageKey=member-home
It's so fast and do the same thing.
You can not use multiple select but you can use more than one CTE like this.
WITH CTEA
AS
(
SELECT 'Coulmn1' A,'Coulmn2' B
),
CETB
AS
(
SELECT 'CoulmnX' X,'CoulmnY' Y
)
SELECT * FROM CTEA, CETB
For getting count use RowNumber and CTE some think like this.
ROW_NUMBER() OVER ( ORDER BY COLUMN NAME )AS RowNumber,
Count(1) OVER() AS TotalRecordsFound
Please let me know if you need more information on this.
Sample for your reference.
With CTE AS (
Select StudentId, S.CityId, S.GenderId
FROM Student S
Inner JOIN CITY C
ON S.CityId = C.CityId
INNER JOIN GENDER G
ON S.GenderId = G.GenderId)
,
GENDER
AS
(
SELECT GenderId
FROM CTE
GROUP BY GenderId
)
SELECT * FROM GENDER, CTE
It is not possible to get multiple result sets from a single CTE.
You can however use a table variable to cache some of the information and use it later instead of issuing the same complex query multiple times:
declare #relevantStudent table (StudentID int);
insert into #relevantStudent
select s.StudentID from Students s
join ...
where ...
-- now issue the multiple queries
select s.GenderID, count(*)
from student s
join #relevantStudent r on r.StudentID = s.StudentID
group by s.GenderID
select s.CityID, count(*)
from student s
join #relevantStudent r on r.StudentID = s.StudentID
group by s.CityID
The trick is to store only the minimum required information in the table variable.
As with any query whether this will actually improve performance vs. issuing the queries independently depends on many things (how big the table variable data set is, how complex is the query used to populate it and how complex are the subsequent joins/subselects against the table variable, etc.).
Do a UNION ALL to do multiple SELECT and concatenate the results together into one table.
;WITH CTE AS(
SELECT
FROM Student
INNER JOIN ...
INNER JOIN ....)
SELECT CityId,City,Count(studentId),NULL,NULL
FROM CTE
GROUP BY CityId,City
UNION ALL
SELECT NULL,NULL,NULL,GenderId,Gender,Count
FROM CTE
GROUP BY GenderId,Gender
Note: The NULL values above just allow the two results to have matching columns, so the results can be concatenated.
I know this is a very old question, but here's a solution I just used. I have a stored procedure that returns a PAGE of search results, and I also need it to return the total count matching the query parameters.
WITH results AS (...complicated foo here...)
SELECT results.*,
CASE
WHEN #page=0 THEN (SELECT COUNT(*) FROM results)
ELSE -1
END AS totalCount
FROM results
ORDER BY bar
OFFSET #page * #pageSize ROWS FETCH NEXT #pageSize ROWS ONLY;
With this approach, there's a small "hit" on the first results page to get the count, and for the remaining pages, I pass back "-1" to avoid the hit (I assume the number of results won't change during the user session). Even though totalCount is returned for every row of the first page of results, it's only computed once.
My CTE is doing a bunch of filtering based on stored procedure arguments, so I couldn't just move it to a view and query it twice. This approach allows avoid having to duplicate the CTE's logic just to get a count.

How to create RowNum column in SQL Server?

In Oracle we have "rownum".
What can I do in SQL Server?
In SQL Server 2005 (and 2008) you can use the ROW_NUMBER function, coupled with the OVER clause to determine the order in which the rows should be counted.
Update
Hmm. I don't actually know what the Oracle version does. If it's giving you a unique number per row (across the entire table), then I'm not sure there's a way to do that in SQL Server. SQL Server's ROW_NUMBER() only works for the rows returned in the current query.
If you have an id column, you can do this:
select a.*,
(select count(*) from mytable b where b.id <= a.id) as rownum
from mytable a
order by id;
Of course, this only works where you're able to order rownums in the same (or opposite) order as the order of the ids.
If you're selecting a proper subset of rows, of course you need to apply the same predicate to the whole select and to the subquery:
select a.*,
(select count(*) from table b where b.id <= a.id and b.foo = 'X') as rownum
from table a where a.foo = 'X'
order by id;
Obviously, this is not particularly efficient.
Based on my understanding, you'd need to use ranking functions and/or the TOP clause. The SQL Server features are specific, the Oracle one combines the 2 concepts.
The ranking function is simple: here is why you'd use TOP.
Note: you can't WHERE on ROWNUMBER directly...
'Orable:
select
column_1, column_2
from
table_1, table_2
where
field_3 = 'some value'
and rownum < 5
--MSSQL:
select top 4
column_1, column_2
from
table_1, table_2
where
field_3 = 'some value'

Is recursion good in SQL Server?

I have a table in SQL server that has the normal tree structure of Item_ID, Item_ParentID.
Suppose I want to iterate and get all CHILDREN of a particular Item_ID (at any level).
Recursion seems an intuitive candidate for this problem and I can write an SQL Server function to do this.
Will this affect performance if my table has many many records?
How do I avoid recursion and simply query the table? Please any suggestions?
With the new MS SQL 2005 you could use the WITHkeyword
Check out this question and particularly this answer.
With Oracle you could use CONNECT BY keyword to generate hierarchical queries (syntax).
AFAIK with MySQL you'll have to use the recursion.
Alternatively you could always build a cache table for your records parent->child relationships
As a general answer, it is possible to do some pretty sophisticated stuff in SQL Server that normally needs recursion, simply by using an iterative algorithm. I managed to do an XHTML parser in Transact SQL that worked surprisingly well. The the code prettifier I wrote was done in a stored procedure. It aint elegant, it is rather like watching buffalo doing Ballet. but it works .
Are you using SQL 2005?
If so you can use Common Table Expressions for this. Something along these lines:
;
with CTE (Some, Columns, ItemId, ParentId) as
(
select Some, Columns, ItemId, ParentId
from myTable
where ItemId = #itemID
union all
select a.Some, a.Columns, a.ItemId, a.ParentId
from myTable as a
inner join CTE as b on a.ParentId = b.ItemId
where a.ItemId <> b.ItemId
)
select * from CTE
The problem you will face with recursion and performance is how many times it will have to recurse to return the results. Each recursive call is another separate call that will have to be joined into the total results.
In SQL 2k5 you can use a common table expression to handle this recursion:
WITH Managers AS
(
--initialization
SELECT EmployeeID, LastName, ReportsTo
FROM Employees
WHERE ReportsTo IS NULL
UNION ALL
--recursive execution
SELECT e.employeeID,e.LastName, e.ReportsTo
FROM Employees e INNER JOIN Managers m
ON e.ReportsTo = m.employeeID
)
SELECT * FROM Managers
or another solution is to flatten the hierarchy into a another table
Employee_Managers
ManagerId (PK, FK to Employee table)
EmployeeId (PK, FK to Employee table)
All the parent child relation ships would be stored in this table, so if Manager 1 manages Manager 2 manages employee 3, the table would look like:
ManagerId EmployeeId
1 2
1 3
2 1
This allows the hierarchy to be easily queried:
select * from employee_managers em
inner join employee e on e.employeeid = em.employeeid and em.managerid = 42
Which would return all employees that have manager 42. The upside will be greater performance, but downside is going to be maintaining the hierarchy
Joe Celko has a book (<- link to Amazon) specifically on tree structures in SQL databases. While you would need recursion for your model and there would definitely be a potential for performance issues there, there are alternative ways to model a tree structure depending on what your specific problem involves which could avoid recursion and give better performance.
Perhaps some more detail is in order.
If you have a master-detail relationship as you describe, then won't a simple JOIN get what you need?
As in:
SELECT
SOME_FIELDS
FROM
MASTER_TABLE MT
,CHILD_TABLE CT
WHERE CT.PARENT_ID = MT.ITEM_ID
You shouldn't need recursion for children - you're only looking at the level directly below (i.e. select * from T where ParentId = #parent) - you only need recursion for all descendants.
In SQL2005 you can get the descendants with:
with AllDescendants (ItemId, ItemText) as (
select t.ItemId, t.ItemText
from [TableName] t
where t.ItemId = #ancestorId
union
select sub.ItemId, sub.ItemText
from [TableName] sub
inner join [TableName] tree
on tree.ItemId = sub.ParentItemId
)
You don't need recursion at all....
Note, I changed columns to ItemID and ItemParentID for ease of typing...
DECLARE #intLevel INT
SET #intLevel = 1
INSERT INTO TempTable(ItemID, ItemParentID, Level)
SELECT ItemID, ItemParentID, #intLevel
WHERE ItemParentID IS NULL
WHILE #intLevel < #TargetLevel
BEGIN
SET #intLevel = #intLevel + 1
INSERT INTO TempTable(ItemID, ItemParentID, Level)
SELECt ItemID, ItemParentID, #intLevel
WHERE ItemParentID IN (SELECT ItemID FROM TempTable WHERE Level = #intLevel-1)
-- If no rows are inserted then there are no children
IF ##ROWCOUNT = 0
BREAK
END
SELECt ItemID FROM TempTable WHERE Level = #TargetLevel

SQL Server: Examples of PIVOTing String data

Trying to find some simple SQL Server PIVOT examples. Most of the examples that I have found involve counting or summing up numbers. I just want to pivot some string data. For example, I have a query returning the following.
Action1 VIEW
Action1 EDIT
Action2 VIEW
Action3 VIEW
Action3 EDIT
I would like to use PIVOT (if even possible) to make the results like so:
Action1 VIEW EDIT
Action2 VIEW NULL
Action3 VIEW EDIT
Is this even possible with the PIVOT functionality?
Remember that the MAX aggregate function will work on text as well as numbers. This query will only require the table to be scanned once.
SELECT Action,
MAX( CASE data WHEN 'View' THEN data ELSE '' END ) ViewCol,
MAX( CASE data WHEN 'Edit' THEN data ELSE '' END ) EditCol
FROM t
GROUP BY Action
Table setup:
CREATE TABLE dbo.tbl (
action VARCHAR(20) NOT NULL,
view_edit VARCHAR(20) NOT NULL
);
INSERT INTO dbo.tbl (action, view_edit)
VALUES ('Action1', 'VIEW'),
('Action1', 'EDIT'),
('Action2', 'VIEW'),
('Action3', 'VIEW'),
('Action3', 'EDIT');
Your table:
SELECT action, view_edit FROM dbo.tbl
Query without using PIVOT:
SELECT Action,
[View] = (Select view_edit FROM tbl WHERE t.action = action and view_edit = 'VIEW'),
[Edit] = (Select view_edit FROM tbl WHERE t.action = action and view_edit = 'EDIT')
FROM tbl t
GROUP BY Action
Query using PIVOT:
SELECT [Action], [View], [Edit] FROM
(SELECT [Action], view_edit FROM tbl) AS t1
PIVOT (MAX(view_edit) FOR view_edit IN ([View], [Edit]) ) AS t2
Both queries result:
If you specifically want to use the SQL Server PIVOT function, then this should work, assuming your two original columns are called act and cmd. (Not that pretty to look at though.)
SELECT act AS 'Action', [View] as 'View', [Edit] as 'Edit'
FROM (
SELECT act, cmd FROM data
) AS src
PIVOT (
MAX(cmd) FOR cmd IN ([View], [Edit])
) AS pvt
From http://blog.sqlauthority.com/2008/06/07/sql-server-pivot-and-unpivot-table-examples/:
SELECT CUST, PRODUCT, QTY
FROM Product) up
PIVOT
( SUM(QTY) FOR PRODUCT IN (VEG, SODA, MILK, BEER, CHIPS)) AS pvt) p
UNPIVOT
(QTY FOR PRODUCT IN (VEG, SODA, MILK, BEER, CHIPS)
) AS Unpvt
GO
Well, for your sample and any with a limited number of unique columns, this should do it.
select
distinct a,
(select distinct t2.b from t t2 where t1.a=t2.a and t2.b='VIEW'),
(select distinct t2.b from t t2 where t1.a=t2.a and t2.b='EDIT')
from t t1
With pivot_data as
(
select
action, -- grouping column
view_edit -- spreading column
from tbl
)
select action, [view], [edit]
from pivot_data
pivot ( max(view_edit) for view_edit in ([view], [edit]) ) as p;
I had a situation where I was parsing strings and the first two positions of the string in question would be the field names of a healthcare claims coding standard. So I would strip out the strings and get values for F4, UR and UQ or whatnot. This was great on one record or a few records for one user. But when I wanted to see hundreds of records and the values for all usersz it needed to be a PIVOT. This was wonderful especially for exporting lots of records to excel. The specific reporting request I had received was "every time someone submitted a claim for Benadryl, what value did they submit in fields F4, UR, and UQ. I had an OUTER APPLY that created the ColTitle and the value fields below
PIVOT(
min(value)
FOR ColTitle in([F4], [UR], [UQ])
)

Resources