I have posted this in the Experts Forum so I have put replies to questions in here as well (Hence the long post!!). Hope someone here can help!!
I am having trouble with a T-SQL query. I have three tables in a join and I need to limit the results returned by query using a group by.
All of the fields in the result group are the same except for two fields, a date field and a varchar field. What I want to achieve is to return the row that has the latest date, and I need the varchar field as that is the information I am after. The problem is if I include the varchar field in the group by, it is returned as a separate group, if I do not included it in the group by, Server: Msg 8120 (aggregate fn / group by) error occurs!
Any ideas how to get around this?
Sample code and results below:
Query:
select distinct s.id, [active-from], code, s.desc, [scheme], [market-id], [market-id]+'CODE'
from [coded] c join sec s on s.id = c.id
join [mkt-security] m on m.id = s.id
where ([market-id] = [scheme] or [market-id]+'CODE' = [scheme])
Group by s.id, [active-from], code, s.desc, [scheme], [market-id], [market-id]+'CODE'
i.e. those with the latest [active-from] date, but I must have the corresponding code value.
I have tried a correlated sub query on the [active-from] field, but it is possible to have different id’s with the same [active-from] date so it did not work.
The primary key for the [coded] table is the combination of [active-from], code and [scheme]. The tables are truncated and then imported back into the SQL database from a Progress database daily, so restructuring the table(s) is not a possibility.
I have rewritten the query to "AND [ACTIVE_FROM] in (SELECT MAX([ACTIVE_FROM]) FROM CODED AND MAX([CODE]) = S.[CODE] AND MAX([SCHEME]) = S.[SCHEME]", (Which I think is the T-SQL for you query), but it returned the error "An aggregate may not appear in the WHERE clause unless it is in a subquery contained in a HAVING clause or a select list, and the column being aggregated is an outer reference."
Also I think the query need to use a Group by as the results I need returned are those with the latest [ACTIVE_FROM] date WITHIN the group of [ID]and [MARKET-ID]. I can't think of any conditions where I can filter the results in the where clause.
The [ ] in the T-SQL is to identify table names as some of the table names (from the Progress db) use characters (the "-") that are unsupported for table names in T-SQL.
I am using the group by as I want to return specific groups of results ( [ID] and [MARKET-ID] ), that have the latest active-from date. I have made a change to that posted before to illustrate:
I am trying to count a column field in a single table and return two count values as one record set using group by.
field1 = group by (department) nvarachar field2 = count (closed) datetime
I have tried using derived tables with no luck getting the desired result.
field2 is a datetime field as indicated I want a count for two conditions
1. WHERE field2 is null 2. WHERE field2 is not null
End Results would like this ====== Department | OpenItems | ClosedItems Department1 | 32 | 24 Departmnet2 | 87 | 46 Department3 | 42 | 76
=======
I got it *almost* working with derived tables, but the group by function was not putting the department as one single row. I was getting multiple rows for departments.
I realize this is probably a simple answer and I am making this a lot harder than it actually is....
I'm new to MSSQL 2005 and want to get a summary of a log table. I want to count all the rows for each date based on a DATETIME field called 'post_date' that holds the date and time of each record's creation.
this is the best I can come up with:
Code:
SELECT DISTINCT(LEFT(post_date,11)) AS post_date, COUNT(DISTINCT(LEFT(post_date,11))) AS total_posts FROM log_directory_contacts GROUP BY post_date
The results show each date but the count column ('total_posts') returns '1' for every row even when I know their are more than 1 record on that date.
If a user connects to SQL Server through say QA using a trusted connection. What would the function be to return the NT username? I tried this (I am a system admin) and it returned dbo. I was looking for my NTusername which I can see when I query processes
I created a report using RS with a SQL stored procedure (with date range parameters and a cursor to build the data) as the data source. When I execute the proc in SQL it returns over 300 records. When i go the the data tab on my report view the data, the proc and the preview only returns the first row. What am i doing wrong? I am new to RS.
I have a Stored Procedure that I am calling from ADO 2.7 in Visual Basic. The process works fine when the stored procedure has records to return. However , when the recordset is return emtpy ADO does not recognize the recordset as being open and empty. For example: VB Code to open recordset: Set rstClass1 = New ADODB.Recordset With rstClass1 .CursorLocation = adUseServer Set .ActiveConnection = DBConn .Source = strRecordSource .LockType = adLockOptimistic .CursorType = adOpenKeyset .Open If Not .EOF Or Not .BOF Then
The program bombs on the If Not .EOF statement with Error 3704 - "Operation is not allowed when the object is closed.
When I run the Stored Procedure In QA using an ID that I know will NOT return any records the result window returns nothing. I noticed that on some of my other SP's the result window will at least return the column titles. These SP's also do not produce the 3704 error in ADO.
Is there a command or a certain method to construct my SP so that is will return enough for ADO to know that the recordset did open its just empty?
The SP is ugly looking but I will post if need be.
Hi,I'm trying to build a query that get only one row from a group ofrows, but I need the values from that row and not the results of onefunction group.I need one row for each idRef, with column2=2 and the bigger column1id |idRef | column1 | column21 1 0 12 1 1 23 1 2 14 2 0 15 2 1 26 2 2 17 2 3 2For these, I will take the rows with id=2 and id=7.Thank you, and sory for my english.
Okay, I am sure this is an easy question, but for some reason I cannot wrap my brain around it. I have a table that has data similar to the following:
IPAddress VisitDate URLVisited
192.154.21.554 9/18/2007 http://www.microsoft.com
192.154.21.554 9/19/2007 http://www.google.com
164.21.124.23 9/19/2007 http://www.microsoft.com
192.154.21.554 9/20/2007 http://www.yahoo.com
What I am trying to do is use a select query to find the most recent visit per IP address (but I also want the other row data like URL visited). I have something like this that is finding the most recent visit per IP address:
SELECT IPAddress, MAX(VisitDate) AS MaxDateVisited FROM VisitorTable GROUP BY IPAddress
How do I also pull in the URL that is associated with the MAX date visited? Thank you!
Hi, I am trying to get the first row of what might be a group of any size of rows within an id that may contain a specific field. For eg
Row A = NoteID 1, FK_JobID 1, UnRead
Row B = NoteID 2, FK_JobID 1, UnRead
Row C = NoteID 3, FK_JobID 1, UnRead
I need the sql to return to just one Job (row) even though the job as 3 UnRead fields. But its returning 3 because its only doing what I'm asking. What I need it to do is just get the one Job (row) where any of the notes = UnRead.
I tried using Top 1, but that will only ever return one row and since I need it to return more than one job (row) it won't work.
Heres my attempt
DECLARE @UserID INT
SET @UserID = 4
SELECT User_Notes.BeenRead, Master_Jobs.By_Who, Master_Jobs.Next_Action, Master_Jobs.Due_Time, Master_Jobs.Due_Date, Master_Jobs.Contact, Master_Jobs.Job_Title, Master_Jobs.JobID FROM User_Notes INNER JOIN Note ON User_Notes.FK_UN_NoteID = Note.NoteID INNER JOIN Master_Jobs ON Note.FK_JobID = Master_Jobs.JobID
WHERE Note.FK_UserID = User_Notes.FK_UN_UserID AND BeenRead = 'UnRead'
creating a query to group identical rows into one and placing corresponding data in appropriate columns for a table named items, what I mean is that I have a table structured as below
As you can see each room is assigned and id and once the room changes the id starts from 1 again. I don't want the 2nd row to appear because 9am - 1pm is in 8am - 5pm. How do I remove this row?
My first post on the forum, I wish I had found this place sooner, looks to be full of good advice and knowledgeable posters.
I have tried searching and looking at the FAQ's but couldn't find an answer. So here goes, all help greatly appreciated.
SQL Server 2005
Table (many columns left out for simplicity)
ID - unique. HdrID - key to the header record. PTtimestamp - date, date/time the row was written to DB. PType - integer, representing various states of the row. etc etc etc (another 15 columns of data)....
Typically there is approximately 250 records per HdrID.
I am trying to do an SQL UPDATE without using my usual solution of writing vb code!
I want to update the PType of latest row (as per timestamp) for each HdrID to -9999.
I built a report with one field as a group. I want to count the number of rows in each field so I can add it to the group field or somw where in the report.
How can I count how many rows do I have in each group?
Hi everyone: I guess this should be a simple question but have not been able to find the answer, does anyone know how to make a SQL Sentence to return at least one row when counting? SELECT COUNT(Id_Field), Field2 FROM Table1 WHERE Code_Field = 1 GROUP BY Field2 This will return 0 rows when the where criteria is not matched by any record on the Table1, but I would like to have one row counting 0 rows, in stead it returns no rows at all. Thanks for any help.
hi, i have a stored procedure SELECT UserName AS Visitor, COUNT(VisitID) AS TotalVisit FROM UserVisits WHERE (ProductID = @ProductID) AND (AnonimIP IS NULL) GROUP BY UserName UNION SELECT AnonimIP AS Visitor, COUNT(VisitID) AS TotalVisit FROM UserVisits AS UserVisits_1 WHERE (ProductID = @ProductID) AND (UserName IS NULL) GROUP BY AnonimIP this will return something like: zuperboy90 - 4 visits ANONIMOUS - 6 visits 85.104.103 - 2 visits etc how can i count the rows returned in both selections (4+6+2 = 12) ? thank you
I have a table, Table1 with 3 columns as follows: colItemKey, colGrouping1, colGrouping2. colItemKey is the primary key. Say colGrouping1 has 4 different types: Grp1A, Grp1B, Grp1C and Grp1D and colGrouping2 has 5 as follows: Grp2A, Grp2B, Grp2C, Grp2D and Grp2E. How do I setup my select so that the result set is as follows:
I have a report with a table and it has 58 rows, in some of the rows i am displaying the data like this
Roth contribution (Heading)
data for the roth contribution.
So depending on the options checked sometimes the heading Roth contribution is at the end of the page and the data is the next page. so how can i group 2 table rows together.
I have tried clicking on the 2 rows and grouping it, but nothing seems to appear in the report..
I'm no expert at SQL, I can't figure out how to solve that problem.
The following statement:
SELECT TeamMemberId, max(MonthId) as MonthId FROM Position INNER JOIN [Month] m INNER JOIN [Year] y ON m.YearId = y.Id AND (y.Id = @yearId OR y.Id < @yearId) ON m.Id = Position.MonthId GROUP BY TeamMemberId
Returns TeamMemberId MonthId 1 14 6 20
which is exactly what I want: The greates MonthId for each TeamMember in a specific year.
But I need more data for this postition, like the PositionName.
But the statement:
SELECT TeamMemberId, max(MonthId) as MonthId, PositionName FROM Position INNER JOIN [Month] m INNER JOIN [Year] y ON m.YearId = y.Id AND (y.Id = 2 OR y.Id < 2) ON m.Id = Position.MonthId GROUP BY TeamMemberId, PositionName
Returns TeamMemberId MonthId PositionName 1 2 ***. d. Geschäftsführung 6 20 ***. d. Geschäftsführung 1 14 CEO 6 16 CEO
The red rows are the ones I wanna get. How can I achieve that only the max. MonthId rows are returned, even when I need to select more columns? Any help is much appreciated.
I have dataset ready having sales rep ID  and other column like customer ID, company name, Email adress and transaction month as shown above how can i make a  row group by sales rep id and each tab (when we export report to excel)  I mean one tab for A sales repID , another tab for B...I did row group  parent group and group by sales repID but not working as expected.
I have data that is grouped by a code number. One of the code numbers have over 600 rows, while other code numbers have around 10 to 20 rows within it.
When I run the report the code number that has over 600 rows gets split over 2 pages while the other code numbers each get their own page.
How do I make it so that when I run the report the code number that has over 600 rows gets all displayed on the 1 page instead of being split over 2?
How could one do this? I understand you could use the COUNT() function, but I'm not sure which object's visibility would best support this. All that I've tried (subtotal area, group visibility) do not seem to work.
If you change the visible property on the subtotal textbox that RS adds, it will only 'blank out' the area where the subtotal row would have been - this doesn't achieve the desired effect of saving space.
Hello Friends, I have created a report using SSRS and in that report I am using group rows in one of the matrix. When I tried to display the subtotal of that group row by using the SSRS in-built feature the subtotal column is coming at the last of the matrix columns but I want it at the front .
so it will be like this,
First the total should be displayed then the group members value.
I have a table Transaction that looks something like the following :
TransactionIDÂ Â Â Â Â Â Â Â Â Â
Currency        Credit             Debit 1     USD     500             0 2     Afcu          6000            0
[Code] ....
I write query like this
select SUM(credit)-SUM(Debit)as [Balance] ,Source from Transaction group by Source
And it came like
Balance Source          1500                       USD 6000                       Afcu 6800                       INR 7000                       Pfc -200                        AUD
But I also want to add Afcu , Pfc with USD and want output like
Balance Source 14500Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â USD 6800Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â INR -200Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â AUD
How can I add a group number to the following query?
For example, I want to be able to have all rows that have Category = 'Field Sales' and Division = 'CA BDM' to be given a unique group number (GN):
RN          ReportDate Category            Division            TotalBalance -------------------- ---------- ------------------------------ ------------------------------ --------------------- 1           2015-06-08 Field Sales           CA BDM             299743154.3912 2           2015-06-07 Field Sales           CA BDM             299765954.0354 3           2015-06-01 Field Sales           CA BDM             297902654.4172 1           2015-06-08 Key Accounts          Life Office           49954981.74 2           2015-06-07 Key Accounts          Life Office           50016989.22 3           2015-06-01 Key Accounts          Life Office           50169967.26 4           2015-05-31 Key Accounts          Life Office           50169918.01
Becomes
GN  RN          ReportDate Category            Division         TotalBalance -------------------------- ---------- ------------------------------ ------------------------------ --------------------- 1    1           2015-06-08 Field Sales           CA BDM           299743154.3912 1    2           2015-06-07 Field Sales           CA BDM           299765954.0354 1    3           2015-06-01 Field Sales           CA BDM           297902654.4172 2    1           2015-06-08 Key Accounts          Life Office           49954981.74 2    2           2015-06-07 Key Accounts          Life Office           50016989.22 2    3           2015-06-01 Key Accounts          Life Office           50169967.26 2    4           2015-05-31 Key Accounts          Life Office           50169918.01
i.e. each combination of Category+Division results in a new GN.
The query is:
selectROW_NUMBER() over (partition by Category, Division order by ReportDate desc) 'RN' , ReportDate , Category , Division , sum(BalanceGBP) as 'TotalBalance' FROM FlowsAndOpenings group by ReportDate, Category, Division order by Category, Division, RN
I am using Sql Server 2008 R2.I have a existing query that basically says
Select Top 50 Subscriber_ID, Â Member_Name, Group_ID from my_table order by rand(checksum(newid()))
However the client now wants to have at least two from each group_id. There are 17 different groups. When I run this as is I get about six of the 17 groups in the results. How can I change this to get at least two results from each group_id?
