TSQL to recurse based on a condition - sql

I have a query to get list of members in a group as below
;WITH CTE (GroupName, GroupMember, isMemberGroup)
AS (SELECT ag.name,
agm.Member,
agm.MemberIsGroup
FROM tb1 ag
LEFT JOIN tb2 agm
ON ag.ID = agm.ID
WHERE ag.name = 'somegroupame')
SELECT *
FROM CTE
Group Name Group Member IsMemberGroup
Admin John 0
Admin Sam 0
Admin GDBA 1
xyz Dan 0
xyz GXy 1
I want to write a query to get members of the sub group as well if IsMemberGroup is 1. Please guide me how to achieve that.
The expected result is to get a list of members including members of sub groups for a give group. The recursion should happen for all the sub groups' groups as well if any:
Resultant: Admin
Group Group Member
Admin John
Admin Sam
Admin(GDBA) Mike
Admin(GDBA) June
Admin(GDBA/Bcksdmin)Mark

You could solve this problem using a recursive CTE to generate your group member hierarchy.
This example using the following data:
/* Let's create some sample data to experiment with.
*/
DECLARE #Sample TABLE
(
GroupName VARCHAR(50),
GroupMember VARCHAR(50),
IsMemberGroup BIT
)
;
INSERT INTO #Sample
(
GroupName,
GroupMember,
IsMemberGroup
)
VALUES
('Admin', 'John', 0),
('Admin', 'Sam', 0),
('Admin', 'GDBA', 1),
('GDBA', 'Mike', 0),
('GDBA', 'June', 0),
('GDBA', 'Bcksdmin', 1),
('Bcksdmin', 'Mark', 0)
;
The query uses the CTE to calculate the group hierarchy. This is then joined back to the original data, like so:
/* Using a recursive CTE we can build the group name
* hierarchy.
*/
WITH [Group] AS
(
/* Anchor query returns all top level teams, ie
* those that do not appear in the group member
* field.
*/
SELECT
s1.GroupName,
CAST(s1.GroupName AS VARCHAR(MAX)) AS Breadcrumb
FROM
#Sample AS s1
LEFT OUTER JOIN #Sample AS s2 ON s2.GroupMember = s1.GroupName
WHERE
s2.GroupName IS NULL
GROUP BY
s1.GroupName
UNION ALL
/* Using recursion, find all children.
* (Sub groups are identified by the IsMemberGroup flag).
*/
SELECT
s.GroupMember AS GroupName,
g.Breadcrumb + '/' + s.GroupMember
FROM
[Group] AS g
INNER JOIN #Sample AS s ON s.GroupName = g.GroupName
WHERE
s.IsMemberGroup = 1
)
SELECT
g.Breadcrumb,
s.GroupMember
FROM
#Sample AS s
INNER JOIN [Group] AS g ON g.GroupName = s.GroupName
WHERE
s.IsMemberGroup = 0
;
Although this solution works I would suggest you revisit the table design, if possible. The column GroupMember is pulling double duty by storing both teams and team members. Splitting this into Group (with parent lookup column) and GroupMember tables would simplify the task. Each new table would only need to describe one real world object, making it clearer where new columns should be created.

Related

More efficient way of doing multiple joins to the same table and a "case when" in the select

At my organization clients can be enrolled in multiple programs at one time. I have a table with a list of all of the programs a client has been enrolled as unique rows in and the dates they were enrolled in that program.
Using an External join I can take any client name and a date from a table (say a table of tests that the clients have completed) and have it return all of the programs that client was in on that particular date. If a client was in multiple programs on that date it duplicates the data from that table for each program they were in on that date.
The problem I have is that I am looking for it to only return one program as their "Primary Program" for each client and date even if they were in multiple programs on that date. I have created a hierarchy for which program should be selected as their primary program and returned.
For Example:
1.)Inpatient
2.)Outpatient Clinical
3.)Outpatient Vocational
4.)Outpatient Recreational
So if a client was enrolled in Outpatient Clinical, Outpatient Vocational, Outpatient Recreational at the same time on that date it would only return "Outpatient Clinical" as the program.
My way of thinking for doing this would be to join to the table with the previous programs multiple times like this:
FROM dbo.TestTable as TestTable
LEFT OUTER JOIN dbo.PreviousPrograms as PreviousPrograms1
ON TestTable.date = PreviousPrograms1.date AND PreviousPrograms1.type = 'Inpatient'
LEFT OUTER JOIN dbo.PreviousPrograms as PreviousPrograms2
ON TestTable.date = PreviousPrograms2.date AND PreviousPrograms2.type = 'Outpatient Clinical'
LEFT OUTER JOIN dbo.PreviousPrograms as PreviousPrograms3
ON TestTable.date = PreviousPrograms3.date AND PreviousPrograms3.type = 'Outpatient Vocational'
LEFT OUTER JOIN dbo.PreviousPrograms as PreviousPrograms4
ON TestTable.date = PreviousPrograms4.date AND PreviousPrograms4.type = 'Outpatient Recreational'
and then do a condition CASE WHEN in the SELECT statement as such:
SELECT
CASE
WHEN PreviousPrograms1.name IS NOT NULL
THEN PreviousPrograms1.name
WHEN PreviousPrograms1.name IS NULL AND PreviousPrograms2.name IS NOT NULL
THEN PreviousPrograms2.name
WHEN PreviousPrograms1.name IS NULL AND PreviousPrograms2.name IS NULL AND PreviousPrograms3.name IS NOT NULL
THEN PreviousPrograms3.name
WHEN PreviousPrograms1.name IS NULL AND PreviousPrograms2.name IS NULL AND PreviousPrograms3.name IS NOT NULL AND PreviousPrograms4.name IS NOT NULL
THEN PreviousPrograms4.name
ELSE NULL
END as PrimaryProgram
The bigger problem is that in my actual table there are a lot more than just four possible programs it could be and the CASE WHEN select statement and the JOINs are already cumbersome enough.
Is there a more efficient way to do either the SELECTs part or the JOIN part? Or possibly a better way to do it all together?
I'm using SQL Server 2008.
You can simplify (replace) your CASE by using COALESCE() instead:
SELECT
COALESCE(PreviousPrograms1.name, PreviousPrograms2.name,
PreviousPrograms3.name, PreviousPrograms4.name) AS PreviousProgram
COALESCE() returns the first non-null value.
Due to your design, you still need the JOINs, but it would be much easier to read if you used very short aliases, for example PP1 instead of PreviousPrograms1 - it's just a lot less code noise.
You can simplify the Join by using a bridge table containing all the program types and their priority (my sql server syntax is a bit rusty):
create table BridgeTable (
programType varchar(30),
programPriority smallint
);
This table will hold all the program types and the program priority will reflect the priority you've specified in your question.
As for the part of the case, that will depend on the number of records involved. One of the tricks that I usually do is this (assuming programPriority is a number between 10 and 99 and no type can have more than 30 bytes, because I'm being lazy):
Select patient, date,
substr( min(cast(BridgeTable.programPriority as varchar) || PreviousPrograms.type), 3, 30)
From dbo.TestTable as TestTable
Inner Join dbo.BridgeTable as BridgeTable
Left Outer Join dbo.PreviousPrograms as PreviousPrograms
on PreviousPrograms.type = BridgeTable.programType
and TestTable.date = PreviousPrograms.date
Group by patient, date
You can achieve this using sub-queries, or you could refactor it to use CTEs, take a look at the following and see if it makes sense:
DECLARE #testTable TABLE
(
[id] INT IDENTITY(1, 1),
[date] datetime
)
DECLARE #previousPrograms TABLE
(
[id] INT IDENTITY(1,1),
[date] datetime,
[type] varchar(50)
)
INSERT INTO #testTable ([date])
SELECT '2013-08-08'
UNION ALL SELECT '2013-08-07'
UNION ALL SELECT '2013-08-06'
INSERT INTO #previousPrograms ([date], [type])
-- a sample user as an inpatient
SELECT '2013-08-08', 'Inpatient'
-- your use case of someone being enrolled in all 3 outpation programs
UNION ALL SELECT '2013-08-07', 'Outpatient Recreational'
UNION ALL SELECT '2013-08-07', 'Outpatient Clinical'
UNION ALL SELECT '2013-08-07', 'Outpatient Vocational'
-- showing our workings, this is what we'll join to
SELECT
PPP.[date],
PPP.[type],
ROW_NUMBER() OVER (PARTITION BY PPP.[date] ORDER BY PPP.[Priority]) AS [RowNumber]
FROM (
SELECT
[type],
[date],
CASE
WHEN [type] = 'Inpatient' THEN 1
WHEN [type] = 'Outpatient Clinical' THEN 2
WHEN [type] = 'Outpatient Vocational' THEN 3
WHEN [type] = 'Outpatient Recreational' THEN 4
ELSE 999
END AS [Priority]
FROM #previousPrograms
) PPP -- Previous Programs w/ Priority
SELECT
T.[date],
PPPO.[type]
FROM #testTable T
LEFT JOIN (
SELECT
PPP.[date],
PPP.[type],
ROW_NUMBER() OVER (PARTITION BY PPP.[date] ORDER BY PPP.[Priority]) AS [RowNumber]
FROM (
SELECT
[type],
[date],
CASE
WHEN [type] = 'Inpatient' THEN 1
WHEN [type] = 'Outpatient Clinical' THEN 2
WHEN [type] = 'Outpatient Vocational' THEN 3
WHEN [type] = 'Outpatient Recreational' THEN 4
ELSE 999
END AS [Priority]
FROM #previousPrograms
) PPP -- Previous Programs w/ Priority
) PPPO -- Previous Programs w/ Priority + Order
ON T.[date] = PPPO.[date] AND PPPO.[RowNumber] = 1
Basically we have our deepest sub-select giving all PreviousPrograms a priority based on type, then our wrapping sub-select gives them row numbers per date so we can select only the ones with a row number of 1.
I am guessing you would need to include a UR Number or some other patient identifier, simply add that as an output to both sub-selects and change the join.

SQL groupby multiple columns

Tablename: EntryTable
ID CharityName Title VoteCount
1 save the childrens save them 1
2 save the childrens saving childrens 3
3 cancer research support them 10
Tablename: ContestantTable
ID FirstName LastName EntryId
1 Neville Vyland 1
2 Abhishek Shukla 1
3 Raghu Nandan 2
Desired output
CharityName FullName
save the childrens Neville Vyland
Abhishek Shukla
cancer research Raghu Nandan
I tried
select LOWER(ET.CharityName) AS CharityName,COUNT(CT.FirstName) AS Total_No_Of_Contestant
from EntryTable ET
join ContestantTable CT
on ET.ID = CT.ID
group by LOWER(ET.CharityName)
Please advice.
Please have a look at this sqlfiddle.
Have a try with this query:
SELECT
e.CharityName,
c.FirstName,
c.LastName,
sq.my_count
FROM
EntryTable e
INNER JOIN ContestantTable c ON e.ID = c.EntryId
INNER JOIN (
SELECT EntryId, COUNT(*) AS my_count FROM ContestantTable GROUP BY EntryId
) sq ON e.ID = sq.EntryId
I assumed you actually wanted to join with ContestantTable's EntryId column. It made more sense to me. Either way (joining my way or yours) your sample data is faulty.
Apart from that, you didn't want repeating CharityNames. That's not the job of SQL. The database is just there to store and retrieve the data. Not to format it nicely. You want to work with the data on application layer anyways. Removing repeating data doesn't make this job easier, it makes it worse.
Most people do not realize that T-SQL has some cool ranking functions that can be used with grouping. Many things like reports can be done in T-SQL.
The first part of the code below creates two local temporary tables and loads them with data for testing.
The second part of the code creates the report. I use two common table expressions (CTE). I could have used two more local temporary tables or table variables. It really does not matter with this toy example.
The cte_RankData has two columns RowNum and RankNum. If RowNum = RankNum, we are on the first instance of charity. Print out the charity name and the total number of votes. Otherwise, print out blanks.
The name of the contestant and votes for that contestant are show on the detail lines. This is a typical report with sub totals show at the top.
I think this matches the report output that you wanted. I ordered the contestants by most votes descending.
Sincerely
John Miner
www.craftydba.com
--
-- Create the tables
--
-- Remove the tables
drop table #tbl_Entry;
drop table #tbl_Contestants;
-- The entries table
Create table #tbl_Entry
(
ID int,
CharityName varchar(25),
Title varchar(25),
VoteCount int
);
-- Add data
Insert Into #tbl_Entry values
(1, 'save the childrens', 'save them', 1),
(2, 'save the childrens', 'saving childrens', 3),
(3, 'cancer research', 'support them', 10)
-- The contestants table
Create table #tbl_Contestants
(
ID int,
FirstName varchar(25),
LastName varchar(25),
EntryId int
);
-- Add data
Insert Into #tbl_Contestants values
(1, 'Neville', 'Vyland', 1),
(2, 'Abhishek', 'Shukla', 1),
(3, 'Raghu', 'Nandan', 2);
--
-- Create the report
--
;
with cte_RankData
as
(
select
ROW_NUMBER() OVER (ORDER BY E.CharityName ASC, VoteCount Desc) as RowNum,
RANK() OVER (ORDER BY E.CharityName ASC) AS RankNum,
E.CharityName as CharityName,
C.FirstName + ' ' + C.LastName as FullName,
E.VoteCount
from #tbl_Entry E inner join #tbl_Contestants C on E.ID = C.ID
),
cte_SumData
as
(
select
E.CharityName,
sum(E.VoteCount) as TotalCount
from #tbl_Entry E
group by E.CharityName
)
select
case when RowNum = RankNum then
R.CharityName
else
''
end as rpt_CharityName,
case when RowNum = RankNum then
str(S.TotalCount, 5, 0)
else
''
end as rpt_TotalVotes,
FullName as rpt_ContestantName,
VoteCount as rpt_Votes4Contestant
from cte_RankData R join cte_SumData S
on R.CharityName = S.CharityName

Is there a better way to get the last level of groups from a table?

I have a Group table and within that table, there is a ParentId column that denotes a groups parent within the Group table. The purpose is to build a dynamic menu from these groups. I know I can loop and grab the last child and construct a result set, but I'm curious if there's a more SQL-y way of accomplishing this.
The table has Id, ParentId and Title fields of int, int and varchar.
Basically, a hierarchy may be constructed this way (People is the base group):
People -> Male -> Boy
-> Man
-> Female
I want to grab the last child(ren) of each branch. So, {Boy, Man, Female} in this case.
As I mentioned, getting that info isn't a problem. I'm just looking for a better way of getting it without having to write a bunch of unions and loops where I can basically change the base group and traverse the entire hierarchy outward, dynamically. I'm not really a Db guy, so I don't know if there's a slick way of doing this or not.
To get the leaf levels for one of many hierarchies, you can use a Recursive Common Table Expressions (CTEs) to enumerate the hierarchy, and then check which members are not the parent of another group to filter to the leaves:
Declare #RootID int = 1
;with cte as (
select
Id,
ParentId,
Title
From
Groups
Where
Id = #RootID
Union All
Select
g.Id,
g.ParentId,
g.Title
From
cte c
Inner Join
Groups g
On c.Id = g.ParentID
)
Select
*
From
cte g
Where
Not Exists (
Select
'x'
From
Groups g2
Where
g2.ParentID = g.Id
);
You can also do this with a left join rather than a not exists
http://sqlfiddle.com/#!6/8f1aa/9
Since you are using SQL Server 2012 you could take advantage of hierarchyid; here is an example following Laurence's schema:
CREATE TABLE Groups
(
Id INT NOT NULL
PRIMARY KEY
, Title VARCHAR(20)
, HID HIERARCHYID
)
INSERT INTO Groups
VALUES ( 1, 'People', '/' ),
( 2, 'Male', '/1/' ),
( 3, 'Female', '/2/' ),
( 4, 'Boy', '/1/1/' ),
( 5, 'Man', '/1/2/' );
SELECT Id
, Title
FROM Groups
WHERE HID NOT IN ( SELECT HID.GetAncestor(1)
FROM Groups
WHERE HID.GetAncestor(1) IS NOT NULL )
http://sqlfiddle.com/#!6/00330/1/0
Results:
ID TITLE
3 Female
4 Boy
5 Man

SQL Server 2005 recursive query with loops in data - is it possible?

I've got a standard boss/subordinate employee table. I need to select a boss (specified by ID) and all his subordinates (and their subrodinates, etc). Unfortunately the real world data has some loops in it (for example, both company owners have each other set as their boss). The simple recursive query with a CTE chokes on this (maximum recursion level of 100 exceeded). Can the employees still be selected? I care not of the order in which they are selected, just that each of them is selected once.
Added: You want my query? Umm... OK... I though it is pretty obvious, but - here it is:
with
UserTbl as -- Selects an employee and his subordinates.
(
select a.[User_ID], a.[Manager_ID] from [User] a WHERE [User_ID] = #UserID
union all
select a.[User_ID], a.[Manager_ID] from [User] a join UserTbl b on (a.[Manager_ID]=b.[User_ID])
)
select * from UserTbl
Added 2: Oh, in case it wasn't clear - this is a production system and I have to do a little upgrade (basically add a sort of report). Thus, I'd prefer not to modify the data if it can be avoided.
I know it has been a while but thought I should share my experience as I tried every single solution and here is a summary of my findings (an maybe this post?):
Adding a column with the current path did work but had a performance hit so not an option for me.
I could not find a way to do it using CTE.
I wrote a recursive SQL function which adds employeeIds to a table. To get around the circular referencing, there is a check to make sure no duplicate IDs are added to the table. The performance was average but was not desirable.
Having done all of that, I came up with the idea of dumping the whole subset of [eligible] employees to code (C#) and filter them there using a recursive method. Then I wrote the filtered list of employees to a datatable and export it to my stored procedure as a temp table. To my disbelief, this proved to be the fastest and most flexible method for both small and relatively large tables (I tried tables of up to 35,000 rows).
this will work for the initial recursive link, but might not work for longer links
DECLARE #Table TABLE(
ID INT,
PARENTID INT
)
INSERT INTO #Table (ID,PARENTID) SELECT 1, 2
INSERT INTO #Table (ID,PARENTID) SELECT 2, 1
INSERT INTO #Table (ID,PARENTID) SELECT 3, 1
INSERT INTO #Table (ID,PARENTID) SELECT 4, 3
INSERT INTO #Table (ID,PARENTID) SELECT 5, 2
SELECT * FROM #Table
DECLARE #ID INT
SELECT #ID = 1
;WITH boss (ID,PARENTID) AS (
SELECT ID,
PARENTID
FROM #Table
WHERE PARENTID = #ID
),
bossChild (ID,PARENTID) AS (
SELECT ID,
PARENTID
FROM boss
UNION ALL
SELECT t.ID,
t.PARENTID
FROM #Table t INNER JOIN
bossChild b ON t.PARENTID = b.ID
WHERE t.ID NOT IN (SELECT PARENTID FROM boss)
)
SELECT *
FROM bossChild
OPTION (MAXRECURSION 0)
what i would recomend is to use a while loop, and only insert links into temp table if the id does not already exist, thus removing endless loops.
Not a generic solution, but might work for your case: in your select query modify this:
select a.[User_ID], a.[Manager_ID] from [User] a join UserTbl b on (a.[Manager_ID]=b.[User_ID])
to become:
select a.[User_ID], a.[Manager_ID] from [User] a join UserTbl b on (a.[Manager_ID]=b.[User_ID])
and a.[User_ID] <> #UserID
You don't have to do it recursively. It can be done in a WHILE loop. I guarantee it will be quicker: well it has been for me every time I've done timings on the two techniques. This sounds inefficient but it isn't since the number of loops is the recursion level. At each iteration you can check for looping and correct where it happens. You can also put a constraint on the temporary table to fire an error if looping occurs, though you seem to prefer something that deals with looping more elegantly. You can also trigger an error when the while loop iterates over a certain number of levels (to catch an undetected loop? - oh boy, it sometimes happens.
The trick is to insert repeatedly into a temporary table (which is primed with the root entries), including a column with the current iteration number, and doing an inner join between the most recent results in the temporary table and the child entries in the original table. Just break out of the loop when ##rowcount=0!
Simple eh?
I know you asked this question a while ago, but here is a solution that may work for detecting infinite recursive loops. I generate a path and I checked in the CTE condition if the USER ID is in the path, and if it is it wont process it again. Hope this helps.
Jose
DECLARE #Table TABLE(
USER_ID INT,
MANAGER_ID INT )
INSERT INTO #Table (USER_ID,MANAGER_ID) SELECT 1, 2
INSERT INTO #Table (USER_ID,MANAGER_ID) SELECT 2, 1
INSERT INTO #Table (USER_ID,MANAGER_ID) SELECT 3, 1
INSERT INTO #Table (USER_ID,MANAGER_ID) SELECT 4, 3
INSERT INTO #Table (USER_ID,MANAGER_ID) SELECT 5, 2
DECLARE #UserID INT
SELECT #UserID = 1
;with
UserTbl as -- Selects an employee and his subordinates.
(
select
'/'+cast( a.USER_ID as varchar(max)) as [path],
a.[User_ID],
a.[Manager_ID]
from #Table a
where [User_ID] = #UserID
union all
select
b.[path] +'/'+ cast( a.USER_ID as varchar(max)) as [path],
a.[User_ID],
a.[Manager_ID]
from #Table a
inner join UserTbl b
on (a.[Manager_ID]=b.[User_ID])
where charindex('/'+cast( a.USER_ID as varchar(max))+'/',[path]) = 0
)
select * from UserTbl
basicaly if you have loops like this in data you'll have to do the retreival logic by yourself.
you could use one cte to get only subordinates and other to get bosses.
another idea is to have a dummy row as a boss to both company owners so they wouldn't be each others bosses which is ridiculous. this is my prefferd option.
I can think of two approaches.
1) Produce more rows than you want, but include a check to make sure it does not recurse too deep. Then remove duplicate User records.
2) Use a string to hold the Users already visited. Like the not in subquery idea that didn't work.
Approach 1:
; with TooMuchHierarchy as (
select "User_ID"
, Manager_ID
, 0 as Depth
from "User"
WHERE "User_ID" = #UserID
union all
select U."User_ID"
, U.Manager_ID
, M.Depth + 1 as Depth
from TooMuchHierarchy M
inner join "User" U
on U.Manager_ID = M."user_id"
where Depth < 100) -- Warning MAGIC NUMBER!!
, AddMaxDepth as (
select "User_ID"
, Manager_id
, Depth
, max(depth) over (partition by "User_ID") as MaxDepth
from TooMuchHierarchy)
select "user_id", Manager_Id
from AddMaxDepth
where Depth = MaxDepth
The line where Depth < 100 is what keeps you from getting the max recursion error. Make this number smaller, and less records will be produced that need to be thrown away. Make it too small and employees won't be returned, so make sure it is at least as large as the depth of the org chart being stored. Bit of a maintence nightmare as the company grows. If it needs to be bigger, then add option (maxrecursion ... number ...) to whole thing to allow more recursion.
Approach 2:
; with Hierarchy as (
select "User_ID"
, Manager_ID
, '#' + cast("user_id" as varchar(max)) + '#' as user_id_list
from "User"
WHERE "User_ID" = #UserID
union all
select U."User_ID"
, U.Manager_ID
, M.user_id_list + '#' + cast(U."user_id" as varchar(max)) + '#' as user_id_list
from Hierarchy M
inner join "User" U
on U.Manager_ID = M."user_id"
where user_id_list not like '%#' + cast(U."User_id" as varchar(max)) + '#%')
select "user_id", Manager_Id
from Hierarchy
The preferrable solution is to clean up the data and to make sure you do not have any loops in the future - that can be accomplished with a trigger or a UDF wrapped in a check constraint.
However, you can use a multi statement UDF as I demonstrated here: Avoiding infinite loops. Part One
You can add a NOT IN() clause in the join to filter out the cycles.
This is the code I used on a project to chase up and down hierarchical relationship trees.
User defined function to capture subordinates:
CREATE FUNCTION fn_UserSubordinates(#User_ID INT)
RETURNS #SubordinateUsers TABLE (User_ID INT, Distance INT) AS BEGIN
IF #User_ID IS NULL
RETURN
INSERT INTO #SubordinateUsers (User_ID, Distance) VALUES ( #User_ID, 0)
DECLARE #Distance INT, #Finished BIT
SELECT #Distance = 1, #Finished = 0
WHILE #Finished = 0
BEGIN
INSERT INTO #SubordinateUsers
SELECT S.User_ID, #Distance
FROM Users AS S
JOIN #SubordinateUsers AS C
ON C.User_ID = S.Manager_ID
LEFT JOIN #SubordinateUsers AS C2
ON C2.User_ID = S.User_ID
WHERE C2.User_ID IS NULL
IF ##RowCount = 0
SET #Finished = 1
SET #Distance = #Distance + 1
END
RETURN
END
User defined function to capture managers:
CREATE FUNCTION fn_UserManagers(#User_ID INT)
RETURNS #User TABLE (User_ID INT, Distance INT) AS BEGIN
IF #User_ID IS NULL
RETURN
DECLARE #Manager_ID INT
SELECT #Manager_ID = Manager_ID
FROM UserClasses WITH (NOLOCK)
WHERE User_ID = #User_ID
INSERT INTO #UserClasses (User_ID, Distance)
SELECT User_ID, Distance + 1
FROM dbo.fn_UserManagers(#Manager_ID)
INSERT INTO #User (User_ID, Distance) VALUES (#User_ID, 0)
RETURN
END
You need a some method to prevent your recursive query from adding User ID's already in the set. However, as sub-queries and double mentions of the recursive table are not allowed (thank you van) you need another solution to remove the users already in the list.
The solution is to use EXCEPT to remove these rows. This should work according to the manual. Multiple recursive statements linked with union-type operators are allowed. Removing the users already in the list means that after a certain number of iterations the recursive result set returns empty and the recursion stops.
with UserTbl as -- Selects an employee and his subordinates.
(
select a.[User_ID], a.[Manager_ID] from [User] a WHERE [User_ID] = #UserID
union all
(
select a.[User_ID], a.[Manager_ID]
from [User] a join UserTbl b on (a.[Manager_ID]=b.[User_ID])
where a.[User_ID] not in (select [User_ID] from UserTbl)
EXCEPT
select a.[User_ID], a.[Manager_ID] from UserTbl a
)
)
select * from UserTbl;
The other option is to hardcode a level variable that will stop the query after a fixed number of iterations or use the MAXRECURSION query option hint, but I guess that is not what you want.

Simple Query to Grab Max Value for each ID

OK I have a table like this:
ID Signal Station OwnerID
111 -120 Home 1
111 -130 Car 1
111 -135 Work 2
222 -98 Home 2
222 -95 Work 1
222 -103 Work 2
This is all for the same day. I just need the Query to return the max signal for each ID:
ID Signal Station OwnerID
111 -120 Home 1
222 -95 Work 1
I tried using MAX() and the aggregation messes up with the Station and OwnerID being different for each record. Do I need to do a JOIN?
Something like this? Join your table with itself, and exclude the rows for which a higher signal was found.
select cur.id, cur.signal, cur.station, cur.ownerid
from yourtable cur
where not exists (
select *
from yourtable high
where high.id = cur.id
and high.signal > cur.signal
)
This would list one row for each highest signal, so there might be multiple rows per id.
You are doing a group-wise maximum/minimum operation. This is a common trap: it feels like something that should be easy to do, but in SQL it aggravatingly isn't.
There are a number of approaches (both standard ANSI and vendor-specific) to this problem, most of which are sub-optimal in many situations. Some will give you multiple rows when more than one row shares the same maximum/minimum value; some won't. Some work well on tables with a small number of groups; others are more efficient for a larger number of groups with smaller rows per group.
Here's a discussion of some of the common ones (MySQL-biased but generally applicable). Personally, if I know there are no multiple maxima (or don't care about getting them) I often tend towards the null-left-self-join method, which I'll post as no-one else has yet:
SELECT reading.ID, reading.Signal, reading.Station, reading.OwnerID
FROM readings AS reading
LEFT JOIN readings AS highersignal
ON highersignal.ID=reading.ID AND highersignal.Signal>reading.Signal
WHERE highersignal.ID IS NULL;
In classic SQL-92 (not using the OLAP operations used by Quassnoi), then you can use:
SELECT g.ID, g.MaxSignal, t.Station, t.OwnerID
FROM (SELECT id, MAX(Signal) AS MaxSignal
FROM t
GROUP BY id) AS g
JOIN t ON g.id = t.id AND g.MaxSignal = t.Signal;
(Unchecked syntax; assumes your table is 't'.)
The sub-query in the FROM clause identifies the maximum signal value for each id; the join combines that with the corresponding data row from the main table.
NB: if there are several entries for a specific ID that all have the same signal strength and that strength is the MAX(), then you will get several output rows for that ID.
Tested against IBM Informix Dynamic Server 11.50.FC3 running on Solaris 10:
+ CREATE TEMP TABLE signal_info
(
id INTEGER NOT NULL,
signal INTEGER NOT NULL,
station CHAR(5) NOT NULL,
ownerid INTEGER NOT NULL
);
+ INSERT INTO signal_info VALUES(111, -120, 'Home', 1);
+ INSERT INTO signal_info VALUES(111, -130, 'Car' , 1);
+ INSERT INTO signal_info VALUES(111, -135, 'Work', 2);
+ INSERT INTO signal_info VALUES(222, -98 , 'Home', 2);
+ INSERT INTO signal_info VALUES(222, -95 , 'Work', 1);
+ INSERT INTO signal_info VALUES(222, -103, 'Work', 2);
+ SELECT g.ID, g.MaxSignal, t.Station, t.OwnerID
FROM (SELECT id, MAX(Signal) AS MaxSignal
FROM signal_info
GROUP BY id) AS g
JOIN signal_info AS t ON g.id = t.id AND g.MaxSignal = t.Signal;
111 -120 Home 1
222 -95 Work 1
I named the table Signal_Info for this test - but it seems to produce the right answer.
This only shows that there is at least one DBMS that supports the notation. However, I am a little surprised that MS SQL Server does not - which version are you using?
It never ceases to surprise me how often SQL questions are submitted without table names.
WITH q AS
(
SELECT c.*, ROW_NUMBER() OVER (PARTITION BY id ORDER BY signal DESC) rn
FROM mytable
)
SELECT *
FROM q
WHERE rn = 1
This will return one row even if there are duplicates of MAX(signal) for a given ID.
Having an index on (id, signal) will greatly improve this query.
with tab(id, sig, sta, oid) as
(
select 111 as id, -120 as signal, 'Home' as station, 1 as ownerId union all
select 111, -130, 'Car', 1 union all
select 111, -135, 'Work', 2 union all
select 222, -98, 'Home', 2 union all
select 222, -95, 'Work', 1 union all
select 222, -103, 'Work', 2
) ,
tabG(id, maxS) as
(
select id, max(sig) as sig from tab group by id
)
select g.*, p.* from tabG g
cross apply ( select top(1) * from tab t where t.id=g.id order by t.sig desc ) p
We can do using self join
SELECT T1.ID,T1.Signal,T2.Station,T2.OwnerID
FROM (select ID,max(Signal) as Signal from mytable group by ID) T1
LEFT JOIN mytable T2
ON T1.ID=T2.ID and T1.Signal=T2.Signal;
Or you can also use the following query
SELECT t0.ID,t0.Signal,t0.Station,t0.OwnerID
FROM mytable t0
LEFT JOIN mytable t1 ON t0.ID=t1.ID AND t1.Signal>t0.Signal
WHERE t1.ID IS NULL;
select a.id, b.signal, a.station, a.owner from
mytable a
join
(SELECT ID, MAX(Signal) as Signal FROM mytable GROUP BY ID) b
on a.id = b.id AND a.Signal = b.Signal
SELECT * FROM StatusTable
WHERE Signal IN (
SELECT A.maxSignal FROM
(
SELECT ID, MAX(Signal) AS maxSignal
FROM StatusTable
GROUP BY ID
) AS A
);
select
id,
max_signal,
owner,
ownerId
FROM (
select * , rank() over(partition by id order by signal desc) as max_signal from table
)
where max_signal = 1;