we have a handful of developers and each of us are responsible for laying out and creating our own database backends. This often leads to inconsistencies in table and column structures.One obvious situation that comes up often is whether or not the other developers are building in history into their primary tables, using history/archive tables or (usually in cases of helper tables) no historical data at all.
My thought on how to alleviate this a little was to suggest that we all build a IS_DELETED computed column into our tables so that someone else trying to work with their data doesn't have to play the guessing game. In most cases, this column would just be running date comparisons on an Expiration Date and either checking to see if it's in the future (usually 12/31/9999) or NULL.
I have read that computed columns can be a performance hit if used/returned unnecessarily but is that also the case on fields where their main use would be filtering? It just seems that the calculation that the computed column is doing would be necessary for the WHERE anywhere so it seems like a wash ... and worth the benefit of not having to decipher someone else's work.
How can I make a statement that will return the 10 most frequently occuring values in a column?
I have no idea if that is even possible, if you have an idea on how I could do that I would really appreciate it.
Im trying to make a page that would show some statistics on a table I have.
Im also trying to make something that would show the count of the number of records inserted in the last 24 hours, week, month, year etc. The table has a column called "DateInserted" as SmallDate, right now i can use a Where DateInserted > '20080310' to get the count, but its not dynamic, is there anyone to merge all these results into one row with each column being a diffrent time period?
I know this a lot of questions, but I would really appreciate any pointers.
I have a table with 3 computed columns, the first two reference a function for there value. The last computed column should be total of the other computed columns, however when trying to write the formula for this column SQL Server rejects it.
Am i correct in thinking it is not possible to reference other computed columns in a computed column's forumla.
My current work around is to call the functions again that the other computed columns use to generate the total, though this seems to me like a performance issue.
I have a Table Having Date,Opening,Addition,Sale values where opening value comes in the very first row other times it is zero.
In ssrs how can i have a report showing closing value = Opening+Addition-Sale  in current row (it is simple for 1st row ). this closing be the opening value in next row and same formula to be continued...
I have an ordering database with several tables that store data of orders belonging to a wide variety of clients. There is a generic report that I need to run which outputs the same data elements. However the criteria to select these orders will vary widely between each client. For e.g.
i) for client# 1 it could be all orders that are still open after 30 days of placing an order
iii) for client# 3 it could be a combination of different things (all orders in West Region that are in hold status for more than 10 days + all orders in Eastern Region that are in shipping and are expected to be delivered in the next 2 days + all completed orders for the rest of the regions).
( OrderRegion = 'West' AND OrderStatus = 'Hold' AND GetDate() - OrderHoldDate >= 10
[code].....
I want to have a stored procedure that selects all data and dynamically attach the where condition at the end for filtering. This way I wouldn't have to worry about any additions/changes that are made to the selection criteria. I can build an interface for admins who can use the UI to maintain the selection criteria and not worry about any code changes to accommodate it. I would like to design a table that holds this criteria. At this point in time, I am thinking of using key value pairs (Column Name, Column Value) but I am not sure how to implement multiple logical operators.
Maybe I am missing something very obvious, but I couldn't do it: begin trangocreate table foo (f1 int not null,f2 int not null,f3 as (f1 + f2) not null primary key clustered)gorollback trango This returns:Server: Msg 8183, Level 16, State 1, Line 8Only UNIQUE or PRIMARY KEY constraints are allowed on computed columns.
I want to create computed column in table. Suppose I have three physical column A,B and C I want to create compute column with computed column. A+B= X X+C=Z Is it possible.
I created a index on a computed column. I did not see any improvementin performance with a join to this column and also my inserts andupdates to this table are failing. Any ideas?Chender
i have to use a datetime field in all the tables in a database as a updated timestamp. i.e. : whenever an update happens to a row in a table, this column called LASTUPDATED has to be updated with current date time.Is there any way to implement this without using the trigger ? can i use COMPUTED column to acheive this ?
I have a SQL table that maintains a field on the status of a report being completed. I have in the record the date the report is due (DateDue) I also have a field called DaysLate which I have set to be a calculated field with formula: DATEDIFF(dd, DateDue, GETDATE()) Thsi works but when the report is *not* late I'd like this to be null is there I way I can do this conditional calculation in a calculated field? Regards Cvive
I'm having a problem with a computed field in a table. I have a stored procedure that inserts a row into a table and returns the id and the computed value. The computed colmn is returned as a decimal (29,6) but for some reason the value is returned with no decimals (confirmed by the Profiler). The value is calculated and displayed in the database properly with decimals. Also, Is Persisted is OFF.
Below is the stored procedure, computed column foruma (both give same result), and the profiler trace.
-- Get the identity value SET @PurchaseOrderDetailId = SCOPE_IDENTITY()
-- Select computed columns into output parameters SELECT @LineTotal = [LineTotal] FROM [Purchasing].[PurchaseOrderDetail] WHERE [PurchaseOrderDetailID] = @PurchaseOrderDetailId
Computed Column Formula ================================================== ============== isnull(CONVERT(DECIMAL (29,6),[OrderQty]*([UnitPrice]*[PackingQty])),0.000000) or isnull([OrderQty]*([UnitPrice]*[PackingQty]),0.000000)
Profiler Trace ================================================== ============== declare @p1 int set @p1=115 declare @p16 numeric(29,0) <- should be numeric(29,6) set @p16=5 <- value should be 5.259200 exec Purchasing.PurchaseOrderDetail_Insert @PurchaseOrderDetailId=@p1 output,@PurchaseOrderId=68,@OrderQty=4,@VendorProd uctId=28,@Description=N'93678975 - GL-2222',@UnitPrice=0.657400,@PackingMethod=N'Bags (2)',@PackingQty=2.0000,@DueDateValue=NULL,@Modifi edDate=''2008-05-19 15:06:37:610'',@LineTotal=@p16 output select @p1, @p16
I have a question about Computed Column Specification which you can specify as a formula for each column inside a table.
I have now columns named Age and Class.
Classes are "Kids" (ID #1) , "Junior" (ID #2) and "Senior" (ID #3) Kids, which is for age of 6 till 12 Junior, which is for 12 till 16 Senior, 16 and above.
I have already searched for hours (I really did) on the internet for a solution, but ended with more questions because of the complicated solutions.
Now the Age is shown as a result of a formule of DOB (Date of Birth column), now I want the exact same thing, but the age must specify which Class the user is in.
Example, when I add a user with the birthdate 25/03/1988 (DD/MM/YYYY) he/she gets 24 as age.
With this formula : (datediff(year,[age],getdate()))
Now I want that the user gets "Senior" as Class (same table).
Senior is ID 3 in this case.
Now I do know how Case, When and Then works, but the validation fails. After reading some forums I understood that I should use a create function method. I am not really experienced with creating functions. Also the coding looks more different as I am used to. How to link the Computed Column to a created formula.
i have a table called "Loans" where i need to compute a column i.e. NoofDays based on which other calculation like interest calculation needs to be done. my query goes like this
"select datediff(dd,VDate, MDate) as NoOfDays ,NoOfDays * Principal * Rate /100 * 365 from Loans".
if i run the above query it says
"Invalid column name 'NoOfDays'".
this executes fine if i use Access but not in SQL Server.
can anybody say what might be the problem and how i can solve it.
hi i want to create a table that has a computed column like this : create table resources( id int identity(0,1) primary key, currentDate int not null, currentMonth int default 0, monthBefore as (currentMonth - (select top 1 currentMonth from resources where (currentDate - resources.currentDate) = 1)))
as u can see, monthBefore is computed column, and i want get currentMonth value of previous month, for this work, i define a column as currentDate that hold only year+month (like 971,082,083,...) and by this expression, i want to get currentMonth of previous record (previous month), but the following error has shown me : Subqueries are not allowed in this context. Only scalar expressions are allowed. how to solve this problem to get currentMonth of previous record ? thanks
I have downloaded the SQL Server Books online and found the section on Computed columns. In my small banking program I have columns named Deposit/Withdrawel and Balance. If I am reading this right I need to set the Computed Column Specification of my Balance Column in order to perform the calculation. My problem now is that I have no idea how to word the formula and other than telling me what a formula is the SQL Server Books online is no help whatsoever, so any help would be awesome.
If i've misunderstood what I am meant to do somebody please please tell me. Thanks
how do you change the allow null on a column to be 0, so instead of having null when empty, it would be 0, I tried default value or binding to 0 to -1, and to 1 and in 0 or -1, it stays null, in 1 it puts the 1 for example if is quantity, it puts as if I would have 1 quantity, so it's either null, or whathever number I put visides -1 or 0 how can I make it be 0 as defult? I'm using it to add and substract with computed column, formula; and the problem is that if it's the first time it's used it does not add or substract, becsause it does not add null with a number, if it has 0 than it works, whata I had to do is add it manually, but of course it doesn't suppoesed to be that way any suggestions will be appriciated:
If i want to split a computed column into two or more columns based onthe the length (its a varchar) of the computed column, how often willsql server determine what the computed column is? roughly along thelines ofselect comp_col,'comp_col 2'=case when datalength(comp_col)<=100 then comp_colelse left(comp_col,99) + '~' end,datalength(comp_col)from aTableAs you can see, in this scenario we use the computed coulumn,comp_col, in a few places, so does SQL server need to calculate thiseach time? I'm playing with this on the basis that it does and thustrying to shift the computed column out to a variable and thenmanipulte and return from their, but that has its own problems whenyou throw in additional parameters (trying to join table udf's) so ifSQL server is smart enough to not calculate the column each time Iwould save a lot of hassle?Cheers Dave
Can I create an index on a variation of a column that isn't actually inthe table?I have a ParcelNumber column, with values like123 AB-67012345ABC 000-00-04012-345-67AP34567890The blanks and non-alphanumeric characters cause problems with users,because sometimes they're there, and sometimes they aren't. So I wouldlike to create an index based on this column, with the non-alphanumericcharacters squeezed out. Of course I can add such a column to thetable and index it, but I'm wondering if it can be done withoutactually adding the column.Thanks,Jim
Hello,I want to assign a column a computed value, which is the multiplicationof a value from the table within and a value from another table.How can I do that?Say the current table is A, column1; and the other table is B, column3.What should I write as formula?I tried someting like;column1 * (SELECT column3 FROM B WHERE A.ID = B.ID)but it didn't work.
hi i want to create a table that has a computed column like this :
Code Snippetcreate table resources(id int identity(0,1) primary key,currentDate int not null,currentMonth int default 0,monthBefore as (currentMonth - (select top 1 currentMonth from resources where (currentDate - resources.currentDate) = 1)))
as u can see, monthBefore is computed column, and i want get currentMonth value of previous month, for this work, i define a column as currentDate that hold only year+month (like 971,082,083,...) and by this expression, i want to get currentMonth of previous record (previous month), but the following error has shown me :
Code SnippetSubqueries are not allowed in this context. Only scalar expressions are allowed. how to solve this problem to get currentMonth of previous record ? thanks
I have question about indexed and not indexed Persisted columns on sql server 2005. It's a bug?
First?, my version of SQL Server is Microsoft SQL Server 2005 - 9.00.3186.00 (Intel X86) Aug 11 2007 03:13:58 Copyright (c) 1988-2005 Microsoft Corporation Developer Edition on Windows NT 5.1 (Build 2600: Service Pack 2)
Now I create two tables and try four select queries:
Code Snippet SET ANSI_NULLS ON SET ANSI_PADDING ON SET ANSI_WARNINGS ON SET ARITHABORT ON SET CONCAT_NULL_YIELDS_NULL ON SET NUMERIC_ROUNDABORT OFF SET QUOTED_IDENTIFIER ON GO create table t1 (id int primary key, id_bigint as cast(id as bigint)) GO create table t2 (id int primary key, id_bigint as cast(id as bigint) persisted) GO select * from t1 -- (1) -- Clustered index scan with two times Compute Scalar GO select * from t2 -- (2) -- Clustered index scan with one times Compute Scalar GO create index IX_t2 on t2 (id_bigint) GO select * from t2 -- (3) -- Index Scan with one times Compute Scalar GO select * from t2 where id_bigint = 0 -- (4) -- Index Seek with one times Compute Scalar GO drop table t1 GO drop table t2 GO SET ANSI_PADDING OFF
1. I don't understand why access to computed column raise scalar computation wto times? 2. I don't understand why access to persisted computed column raise any scalar computation? 3. I don't understand why access to persisted computed column over index required any scalar computations?
Can anyone from Microsoft SQL Server Team told me about this mistake? It's a BUG or I incorrect understand value of the "PERSISTED" word?
-- Thanks with avanced. WBR, Roman S. Golubin grominc[at]gmail.com
Is it possible to retrieve the formula associated with a computed column using t-SQL? I can use COLUMNPROPERTY( id, column, 'IsComputed') to find the computed columns, but how do I get the formula itself?
I've got two integer columns in the table, third one is computed by previous two ones division. That's fine, however sometimes can happen that divided by column is set to zero. How can i avoid division by zero exception, please? TIA
I want to use an if statement to compute the value of a column in SQL Database using other columns. I am supposed to check if a column is null or not and do the computations accordingly. Can anyone help?
Hi. I have this method in a class, it produces a string value based on the value of another property in the object (which represents a field in the database). I would like to turn this into a computed column in SQL server... but need help converting the formula if this is even possible. Thanks in advance.public string GetVendorEvalRating(int vendorevaltotal) { string vendorevalrating = "";if (vendorevaltotal >= 26) {vendorevalrating = "Critical"; }else if ((vendorevaltotal >= 10) && (vendorevaltotal <= 25)) {vendorevalrating = "Material"; }else if ((vendorevaltotal >= 0) && (vendorevaltotal <= 9)) {vendorevalrating = "Minor"; }return vendorevalrating; }
I'm new to computed columns so heres the issue.I have four fields in particlur which I am trying to use: LeaseStart - smalldatetimeLeaseEnd - smalldatetimeLeaseDuration - Calulation = (datediff(year,[LeaseStart],[LeaseEnd]))YearlyExpense -TotalExpense - numeric(18,0) Now I am trying to divide the field in LeaseDuration by a numeric column (totalExpense) in the column YearlyExpense. convert(numeric(LeaseDuration)) / totalExpense Any help would be great!
Select t1.*, (Select Max(Qty) From T2 where ID=t1.ID) as MaxQty FROM T1 t1 WHERE MaxQty>=10
But it fails as computed or inline query columns can not be added in where clause.
However following works: Select t1.*, (Select Max(Qty) From T2 where ID=t1.ID) as MaxQty FROM T1 t1 WHERE (Select Max(Qty) From T2 where ID=t1.ID) >=10
BUT IS IT OPTIMIZED? Please suggest an optimized way to handle such scenarios.