T-SQL (SS2K8) :: Getting Minimum And Maximum Values In A Large Table
May 23, 2014
Table definition:
Create table code (
id identity(1,1)
code
parentcode
internalreference)
There are other columns but I have omitted them for clarity.
The clustered index is on the ID.
There are indexes on the code, parentcode and internalreference columns.
The problem is the table stores a parentcode with an internalreference and around 2000 codes which are children of the parentcode. I realise the table is very badly designed, but the company love orms!!
The table currently holds around 300 millions rows.
The application does the following two queries to find the first internalreference of a code and the last internal refernce of a code:
--Find first internalrefernce
SELECT TOP 1 ID, InternalReference
FROM code
WHERE ParentCode = 'M222'
Order By InternalReference
-- Find last ineternalreference
SELECT TOP 1 ID, InternalReference
FROM code
WHERE ParentCode = 'M222'
Order By InternalReference DESC
These queries are running for a very long time, only because of the sort. If I run the query without the sort, then they return the results instantly, but obviously this doesn't find the first and last internalreference for a parentCode.
I realize the best way to fix this is to redesign the table, but I cannot do that at this time.
Is there a better way to do this so that two queries which individually run very slowly, can be combined into one that is more efficient?
what I try is a little bit heavy, maybe, but I only miss the minimum/maximum fuction - or I didn´t found it; not here in the Forum and also not in the onlinehelp of the SQL Server.
What I try to do:
I have 2 columns in my table; a start- and an end-date. For this period of time between end and start i have to calculate the days for the years. Here my thoughts (for the current year):
Is the startdate <= 31.12.2004 and the enddate >= 1.1.2004 i have to calculate die datediff between max(1.1.2004/startdate) and min(31.12.2004/enddate)
like this sqlstatement:
SELECT CASE WHEN dbo.Phases.phasenstart <= CAST(CAST(YEAR(GETDATE()) AS varchar) + '-31-12' AS smalldatetime) AND dbo.Phases.phasenabschlussist >= CAST(CAST(YEAR(GETDATE()) AS varchar) + '-01-01' AS smalldatetime) THEN 365 ELSE 0 END AS Expr2, FROM dbo.Phases WHERE (phasenstart IS NOT NULL) AND (phasenabschlussist IS NOT NULL)
instead of 365 there must be the above calculation. Is start=3.1.2003 and end=30.1.2004 I expect as result only the 30 days in 2004.
I am trying to get the minimum and maximum values from a field in SQL Server 2008 Express, but I cannot even get started because I keep getting this error that I cannot figure out.
I have a problem (who not?) with a function which i'm using in a view.This function is a function which calculates a integer value of adate. For example: '12/31/2004 00:00:00" becomes 20041231. This isvery handy in a datawarehouse and performes superfast. But here is myproblem.My calendar table is limited by a couple of years. What happens isthat sometimes a value is loaded which is not in the range of theCalendardate. What we want to do is when a date is loaded is that thisfunction insert a minimum date when date < minimum date and a maximumdate when date > maximum date.Yes i know you're thinking : This is datamanipulation and yes this istrue. But now we loose information in our cubes and reports by innerjoining. So if we can use a minimum and a maximum than a user wouldsay: "This is strange, a lot of values on 1980/1/1!" instead of "Ithink that i have not all the data!"GreetzHennie
Because of an error in google or underlying site i can reply on my ownissue. Therefore i copied the former entered message in this message.-------------------------------------REPY----------------------------------Hi Maybe i wasn't clear. I want to dynamically check whether what thelowest date and the highest date is in the calendar table. Thepresented solutions has fixed dates and i don't want that.If i could store a global variable in SQL server (dynamic properties?)then it would be great. Fill this once and call it multiple times inmy intensively used function. Is this possible?GreetzHennie----------------------------Previously enteredissue-----------------------I have a problem (who not?) with a function which i'm using in a view.This function is a function which calculates a integer value of adate. For example: '12/31/2004 00:00:00" becomes 20041231. This isvery handy in a datawarehouse and performes superfast. But here is myproblem.My calendar table is limited by a couple of years. What happens isthat sometimes a value is loaded which is not in the range of theCalendardate. What we want to do is when a date is loaded is that thisfunction insert a minimum date when date < minimum date and a maximumdate when date > maximum date.Yes i know you're thinking : This is datamanipulation and yes this istrue. But now we loose information in our cubes and reports by innerjoining. So if we can use a minimum and a maximum than a user wouldsay: "This is strange, a lot of values on 1980/1/1!" instead of "Ithink that i have not all the data!"GreetzHenniePlaats een reactie op dit berichtBericht 2 van deze discussieVan:John Bell (jbellnewsposts@hotmail.com)Onderwerp:Re: Intensively used function in view needs a minimum andmaximum from a tableView this article onlyDiscussies:comp.databases.ms-sqlserverDatum:2004-12-30 03:56:25 PSTHiIf you LEFT or RIGHT JOIN to the calendar table you will get a NULLvaluefor the column, you can then is CASE to determine the valueCREATE FUNCTION ConvertDate (@datevalue datetime)RETURNS INTASBEGINDECLARE @dateint INTSELECT @dateint = CAST( CASE WHEN A.Date < '20030101' THEN '19800101'WHEN A.Date > '20051231' THEN '99991231'ELSE CONVERT(CHAR(4),C.[Year]) + RIGHT('0'+CONVERT(VARCHAR(2),C.[Month]),2) + RIGHT('0'+CONVERT(VARCHAR(2),C.[Day]),2)END AS INT )FROM ( SELECT @datevalue AS [Date] ) ALEFT JOIN CALENDAR C ON C.[Date] = A.[Date]RETURN @dateintENDGOJohn"Hennie de Nooijer" <hdenooijer@hotmail.com> wrote in messagenews:191115aa.0412300238.7dee0f85@posting.google.c om...[color=blue]>I have a problem (who not?) with a function which i'm using in a[/color]view.[color=blue]> This function is a function which calculates a integer value of a> date. For example: '12/31/2004 00:00:00" becomes 20041231. This is> very handy in a datawarehouse and performes superfast. But here is my> problem.>> My calendar table is limited by a couple of years. What happens is> that sometimes a value is loaded which is not in the range of the> Calendardate. What we want to do is when a date is loaded is that this> function insert a minimum date when date < minimum date and a maximum> date when date > maximum date.>> Yes i know you're thinking : This is datamanipulation and yes this is> true. But now we loose information in our cubes and reports by inner> joining. So if we can use a minimum and a maximum than a user would> say: "This is strange, a lot of values on 1980/1/1!" instead of "I> think that i have not all the data!">> Greetz>> Hennie[/color]Plaats een reactie op dit berichtBericht 3 van deze discussieVan:Hugo Kornelis (hugo@pe_NO_rFact.in_SPAM_fo)Onderwerp:Re: Intensively used function in view needs a minimum andmaximum from a tableView this article onlyDiscussies:comp.databases.ms-sqlserverDatum:2004-12-30 15:32:06 PSTOn 30 Dec 2004 02:38:51 -0800, Hennie de Nooijer wrote:[color=blue]>I have a problem (who not?) with a function which i'm using in a[/color]view.[color=blue]>This function is a function which calculates a integer value of a>date. For example: '12/31/2004 00:00:00" becomes 20041231. This is>very handy in a datawarehouse and performes superfast. But here is my>problem.[/color](snip)Hi Hennie,Is this conversion all that your function does? If so, you might wanttotry the following alternative (using CURRENT_TIMESTAMP as example;replaceit with your date column / parameter):SELECT CAST(CONVERT(varchar, CURRENT_TIMESTAMP, 112) AS int)You could put this in the UDF (probably at least as fast as yourcurrentCalenmdar-table based function), or use it inline as a replacement tothefunction call (probably even faster).It should work for all dates from Jan 1st 1753 through Dec 31st 9999.Best, Hugo--(Remove _NO_ and _SPAM_ to get my e-mail address)
I have a report that calculates mean, min, max, stddev of somer exercises for a class (pushups, situps, trunk lifts, etc). I also have to calculate those for the 1-mile run time (ex: data -- 7:56, 6:35, 9:45 ( in minuteseconds). Obviously the standard Avg(), Min(), Max(), StdDev() functions won't work for time. So I put in some custom code to convert the time to seconds so I can use the standard functions on the seconds and also code to convert that answer back to minuteseconds. Max() works, but for example when I try to calculate the min(), it includes nulls from the dataset as zeros (not every student has to do the mile run). So the min is always 0. How can I exclude nulls from being calculated. The min() function excludes nulls so it returns the correct min() on integer data. What else can I do?
--drop table #temp create table #temp (id int, idvalue int) insert into #temp(id,idvalue) select 1095,75
[code]...
I need to take the id value from maximum's id, and compare the rest id value from the table. i need to check the diffrence , if diffrence is more than 18, then i need to raise the flag as failure otherwise the whole test is success. i need to take 63 and compare rest 69,65,61,75.check the diffrence less than 18 or not.
I have 2 identical tables one contains current settings, the other contains all historical settings.I could create a union view to display the current values from table A and all historical values from table B, butthat would also require a Variable to hold the tblid for both select statements.
Q. Can this be done with one joined or conditional select statement?
DECLARE @tblid int = 501 SELECT 1,2,3,4,'CurrentSetting' FROM TableA ta WHERE tblid = @tblid UNION SELECT 1,2,3,4,'PreviosSetting' FROM Tableb tb WHERE tblid = @tblid
I am using this below query to sum and select maximum values from table. I have converted the cost column here and how can I possibly sum the cost column?
select ID, MAX(Dates) Dates,'$ ' + replace(convert(varchar(100), convert(money, Cost), 1), '.00', '') Cost, MAX(Funded) Funded from Application group by ID, Dates, Cost, Funded
I have a master table containing details of over 800000 surveys made up of approximately 400 distinct document names and versions. Each document can have as few as 10 questions but as many as 150. Each question represents one row.
My challenge is to create a separate spreadsheet for each of the 400 distinct document names and versions containing all the rows and columns present in the master table. The largest number of rows would be around 150 and therefore each spreadsheet will not be very big.
e.g. in my sample data below, i will need to create individual Excel files named as follows . . . "Document1Version1.xlsx" containing all the column names and 6 rows for the 6 questions relating to Document 1 version 1 "Document1Version2.xlsx" containing all the column names and 8 rows for the 8 questions relating to Document 1 version 2 "Document2Version1.xlsx" containing all the column names and 4 rows for the 4 questions relating to Document 2 version 1
I assume that one of the first things is to create a lookup of the distinct document names and versions assign some variables and then use this lookup to loop through and sequentially filter the master table data ready for creating the individual Excel files.
--CREATE TEMP TABLE FOR EXAMPLE
IF OBJECT_ID('tempdb..#excelTest') IS NOT NULL DROP TABLE #excelTest CREATE TABLE #excelTest ( [rowID] [nvarchar](10) NULL, [docName] [nvarchar](50) NULL,
table2 is intially populated (basically this will serve as historical table for view); temptable and table2 will are similar except that table2 has two extra columns which are insertdt and updatedt
process: 1. get data from an existing view and insert in temptable 2. truncate/delete contents of table1 3. insert data in table1 by comparing temptable vs table2 (values that exists in temptable but not in table2 will be inserted) 4. insert data in table2 which are not yet present (comparing ID in t2 and temptable) 5. UPDATE table2 whose field/column VALUE is not equal with temptable. (meaning UNMATCHED VALUE)
* for #5 if a value from table2 (historical table) has changed compared to temptable (new result of view) this must be updated as well as the updateddt field value.
Hello,I need to select the minimum between the result of a function and anumber, but i can't find a smart way. By now I'm doing like thefollowing, but I think is very expensive because it evaluates thefunction twice:select case when (myfunction())<100 then (myfunction()) else 100 endAny idea is appreciated.Thank youRegards--elimina carraro per rispondere via email
I have a small problem with a join clause, because i need to return all values from my table BL:
my code is:
SELECT cast(0 as bit) as 'Escolha',data, contado , ollocal ,origem, ousrdata,ousrhora FROM ( SELECT noconta,banco, u_area FROM BL
[code]....
In fact, i need to return 2 accounts (16,35) - x.NOCONTA IN (16,35), but I know that the problem is on the WHERE clause.How can do that, because i need all the condition on WHERE clause regarding my table OL, but also, i need to return my two accounts (16,35).
Like all location details stored from all months in these table
here Dr=debit,Cr=Credit Formula= 'Dr-Cr' to find the salary wavges of amount
so i made the query to find the amount for may
select fs_locn, fs_accno, amount=sum(case when fs_accno like 'E%' and fs_tran_type='Dr' then fs_amount when fs_accno like 'E%' and fs_tran_type='Cr' then fs_amount * -1 end ) from accutn_det where fs_trans_date between '01-may-2014' and '31-may-2014' groupby fs_locn,fs_accno
now i need the sum values of all costcenter for the particular account.how to do that?
I've already created a table and i wanna to insert values in that more than five hundred row ,that values are stored in Excel files, Here I've the doubt is it possible to insert values from excel sheet? I've current data base of ms sql 2000, if it is possible means, how to insert values using query?
I have a table with data in two columns, item and system. I am trying to accomplish is we want to compare if an item exists in 1, 2 or all systems. If it exists, show item under that system's column, and display NULL in the other columns.
I have aSQL Fiddle that will allow you to visualize the schema.
The closest I've come to solving this is a SQL pivot, however, this is dependent on me knowing the name of the items, which I won't in a real life scenario.
select [system 1], [system 2], [system 3] from ( SELECT distinct system, item FROM test where item = 'item 1' ) x pivot ( max(item)
SELECT CAST(DEL_INTERCOMPANYRETURNACTIONID AS NVARCHAR(4000)) COLLATE DATABASE_DEFAULT AS DEL_INTERCOMPANYRETURNACTIONID, 'SRC_AX.PURCHLINE.DEL_INTERCOMPANYRETURNACTIONID' FROM SRC_AX.PURCHLINE WHERE DEL_INTERCOMPANYRETURNACTIONID IS NULL UNION SELECT CAST(DEL_INTERCOMPANYRETURNACTIONID AS NVARCHAR(4000)) COLLATE DATABASE_DEFAULT AS DEL_INTERCOMPANYRETURNACTIONID, 'SRC_AX.SALESLINE.DEL_INTERCOMPANYRETURNACTIONID'
[Code] .....
My tabel is HST_MASTER.Control.
I want to have this query in a stored procedure. What syntax stored procedure i need to make to fill my table.
I have a table with a column AttributeNumber and a column AttributeValue. The data is like this:
OrderNo. AttributeNumber AttributeValue 1.-Order_1 2001 A 2.-Order_1 2002 B 3.-Order_1 2003 C 4.-Order_2 2001 A 5.-Order_2 2002 B 6.-Order_2 2003 C
So the logic is as follows:
I need to display in my query the values are coming from Order_1, means AttributreValues coming from AttibuteNumbers: 2001,2002,2003...and Order_2 the same thing.
Not sure how to create my Select here since the values are in the same table
I would like to know if the following sql can be used to obtain specific columns from calling a stored procedure with parameters:
/* Create TempTable */ CREATE TABLE #tempTable (MyDate SMALLDATETIME, IntValue INT) GO /* Run SP and Insert Value in TempTable */ INSERT INTO #tempTable (MyDate, IntValue) EXEC TestSP @parm1, @parm2
If the above does not work or there is a better way to accomplish this goal, how to change the sql?
For example we've got a row from [Formula_Calc] table 'F1+F3' as a string that needs to be transformed as 240+160=400
The below code works for the above example but if I pick 'F11+F3' instead , returns 2561 which comes from 2401+16. Probably replaces F1 value instead of F11 and adds 1st digit (1) if I got it right ...
DECLARE @formula NVARCHAR(100); DECLARE @Total NVARCHAR(100); SET @formula = 'F11+F3';
SELECT @formula = REPLACE(@formula,RowNo,Total) FROM [Totals]
How can we find maximum value on column level? Suppose we have table A with four columns col1,col2,col3,col4, Now my query look likes this:
Select col1, col2,col3,col4, (col1 + col2) as addcol1, (col2 + col3) as addcol2, (col3 + col4) as addcol3, Max(addcol1,addcol2,addcol3) as maxvalue from Table A
I am getting error as max accepts one argument, I cannot use case statement as table is already bulky and it will make my query more expensive.
1. PhoneID is nothing but the participant in the call. PhoneID = 1 is twice from above. Which means 2 particpants (Same call )with 2 numbers with their callstarttime and callendtime. Similarly for PhoneID =2, there are 4 participants. And the list goes on for a day and then for a month.
2. For example a phone call P1 with 2 participants is going on for a particular day. We should not consider the same phone call having 2 participants involved. So that the concurrency would be 2. We have to ignore that here.
3. Only to be considered is other Phone calls for that day. Lets say P1 having call with 2 participants, P2 having some 4 participants which fall in the time period of P1. Then we should consider P1 and P2 the common period
4. In order to find number of concurrent calls happened for a day basing on callstarttime and callendtime. What would be the query?
5. Should consider the Timeperiod or the bucket with 1 hour as the period.
6. A Phone Call P1, Phone Call P2, should have matching (common) time ( keeping all the scenarios) is required for this query.
Result for Concurrent calls for a day should be like below. Should get all the concurrent connections happened for a particular day.
Date|TimePeriod/Bucket(hr part)|Concurrentconnections| Jan-01-2015|01 to 02|3 Jan-01-2015|11 to 12|2 Jan-02-2015|04 to 05|5 Jan-02-2015|12 to 13|13 ........
ii) So once the above is achieved.
Have to find the Maximum concurrent connections for day from the above.
For below Maximum Concurrent connections are '3' Date|TimePeriod/Bucket(hr part)|Concurrentconnections| Jan-01-2015|01 to 02|3 Jan-01-2015|11 to 12|2
Hence the Result for Maximum Concurrent Connections would be
Date|TimePeriod/Bucket(hr part)|MaxConcurrentconnections| Jan-01-2015|01 to 02|3 Jan-02-2015|12 to 13|13 .............
I have an issue where I have multiple rows of data and I need to reduce a dollar amount by a fixed maximum. I am going to throw some code in here to give a rudimentary idea of the data and what the final result should be.
I need to run an update so that the result of the following query:
select LineNum, Code, Amt, MaxAmt from
@tblLooks like this:
LineNum Code Amt MaxAmt ----------- ---- --------------------- --------------------- 1 AA 10.00 50.00 2 AA 20.00 50.00 3 AA 20.00 50.00
(3 row(s) affected)
I have tried cursors but got unexpected results or the MaxAmt always defaulted to the original even if I updated it. This seems like a simple problem but I have been banging my head against the wall for 2 days now. I've written some pretty complicated updates with less effort than this and I must have some mental block that is keeping me from figuring this out.
How to count the number of values that exist in a row based on the values from an array of numbers. Basically the the array of numbers I want to look for are in row 1 of table [test 1] and I want to search for them and count the "out of" in table [test 2]. Excuse me for not using the easiest way to convey my question below. I guess in short I have 10 numbers and like to find how many of those numbers exist in each row. short example:
I have a script that I use after some amount of data massaging (not shown). I would like to be able to change the
1) denominator value (the value 8 in line 32 of my code) based on how many columns are selected by the where clause:
where left(CapNumber,charindex('_', CapNumber)-1) = 1where capNumber is a value like [1_1], [1_4], [1_6]...[1_9] capNumber can be any values from [1_1]...[14_10] depending upon the specialty value (example: Allergy) and the final number after the equal sign is a number from 1 to 14)
2) I'd like to dynamically determine the series depending upon which values correspond to the specialty and run for each where: left(CapNumber,charindex('_', CapNumber)-1) = n. n is a number between 1 and 14.
3) finally I'd like to dynamically determine the columns in line 31 (4th line from the bottom)
If I do it by hand it's 23 * 14 separate runs to get separate results for each CapNumber series within specialty. The capNumber series is like [1_1], [1_2], [1_3],[1_4], [1_5], [1_6], [1_7], [1_8],[1_9] ... [8_4],[8_7] ... [14_1], [14_2],...[14_10] etc.
Again, the series are usually discontinuous and specific to each specialty.
Here's the portion of the script (it's at the end) that I'm talking about:
--change values in square brackets below for each specialty as needed and change the denom number in the very last query.
if object_id('tempdb..#tempAllergy') is not null drop table #tempAllergy select * into #tempAllergy from dbo.#temp2 T
[Code] ....
If I were to do it manually I'd uncomment each series line in turn and comment the one I just ran.
