Return 2 entries per group based on the two lowest values in group

Below is a query that I need to modify so that it returns just the 2
MastNUMs, Stores, and Distances where the Distances are the two lowest for a
mastnum. So the results would be more like this
MastNUM Store Distance
-- -- --
000000067 76 3.358330
000000067 70 7.082444
000000068 76 4.447685
000000068 70 5.516853
000000069 70 3.836682
000000069 76 6.331691
000000070 76 3.729323
SELECT TOP 15 MastNUM, Store, Distance
FROM DistTest
GROUP BY MastNUM, Store, Distance
ORDER BY MastNUM, Distance
MastNUM Store Distance
-- -- --
000000067 76 3.358330
000000067 70 7.082444
000000067 69 8.112116
000000067 112 19.924702
000000068 76 4.447685
000000068 70 5.516853
000000068 69 6.874022
000000068 112 18.622366
000000068 71 19.396528
000000069 70 3.836682
000000069 76 6.331691
000000069 69 8.924779
000000069 112 16.709897
000000069 71 17.462224
000000070 76 3.729323
--
KoryI can get the lowest for each with the following but how can I get the
lowest 2?
SELECT TOP 15 t1.MastNUM, t1.Store, t1.Distance
FROM DistTest t1,
( SELECT MastNUM, MIN(Distance) AS MinDist
FROM DistTest
GROUP BY MastNum ) as Dmin
WHERE t1.MastNum = Dmin.MastNUm AND t1.Distance = Dmin.MinDist
go
MastNUM Store Distance
-- -- --
000000067 76 3.358330
000000068 76 4.447685
000000069 70 3.836682
000000070 76 3.729323
000000071 76 4.046238
000000072 70 3.928709
000000073 76 4.663551
000000074 76 3.206388
000000076 76 4.745636
000000077 76 5.338428
000000078 70 5.121837
000000079 70 4.580213
000000080 70 4.338181
000000081 76 4.069455
000000082 76 4.465975
"Kory Yingling" <Mister2zx3@.yahoo.com> wrote in message
news:%23OJC2n8WDHA.536@.TK2MSFTNGP10.phx.gbl...
> Below is a query that I need to modify so that it returns just the 2
> MastNUMs, Stores, and Distances where the Distances are the two lowest for
a
> mastnum. So the results would be more like this
> MastNUM Store Distance
> -- -- --
> 000000067 76 3.358330
> 000000067 70 7.082444
> 000000068 76 4.447685
> 000000068 70 5.516853
> 000000069 70 3.836682
> 000000069 76 6.331691
> 000000070 76 3.729323
> SELECT TOP 15 MastNUM, Store, Distance
> FROM DistTest
> GROUP BY MastNUM, Store, Distance
> ORDER BY MastNUM, Distance
> MastNUM Store Distance
> -- -- --
> 000000067 76 3.358330
> 000000067 70 7.082444
> 000000067 69 8.112116
> 000000067 112 19.924702
> 000000068 76 4.447685
> 000000068 70 5.516853
> 000000068 69 6.874022
> 000000068 112 18.622366
> 000000068 71 19.396528
> 000000069 70 3.836682
> 000000069 76 6.331691
> 000000069 69 8.924779
> 000000069 112 16.709897
> 000000069 71 17.462224
> 000000070 76 3.729323
> --
> Kory
>|||I think I have found a way to get the lowest and 2nd lowest, but I hope
their is a better way than the following..
SELECT TOP 15 t1.MastNUM, t1.Store, t1.Distance
FROM DistTest t1,
( SELECT MastNUM, MIN(Distance) AS MinDist
FROM DistTest
GROUP BY MastNum ) as Dmin
WHERE t1.MastNum = Dmin.MastNUm AND t1.Distance = Dmin.MinDist
go
SELECT TOP 15 t1.MastNUM, t1.Store, t1.Distance
FROM DistTest t1,
(
SELECT MastNUM, MIN(Distance) AS MinDist
FROM DistTest t2
WHERE Distance > (
SELECT MIN(Distance)
FROM DistTest d2
WHERE t2.MastNum = d2.MastNum
GROUP BY MastNum
)
GROUP BY MastNum
) as Dmin
WHERE t1.MastNum = Dmin.MastNUm AND t1.Distance = Dmin.MinDist
order by T1.mASTnUM
go
MastNUM Store Distance
-- -- --
000000067 76 3.358330
000000068 76 4.447685
000000069 70 3.836682
000000070 76 3.729323
000000071 76 4.046238
000000072 70 3.928709
000000073 76 4.663551
000000074 76 3.206388
000000076 76 4.745636
000000077 76 5.338428
000000078 70 5.121837
000000079 70 4.580213
000000080 70 4.338181
000000081 76 4.069455
000000082 76 4.465975
15 record(s) selected [Fetch MetaData: 0/ms] [Fetch Data: 0/ms]
[Executed: 8/6/03 12:05:51 AM CDT ] [Execution: 16/ms]
MastNUM Store Distance
-- -- --
000000067 70 7.082444
000000068 70 5.516853
000000069 76 6.331691
000000070 70 6.957127
000000071 70 5.944537
000000072 76 6.277529
000000073 70 5.487792
000000074 70 7.193161
000000076 70 5.201532
000000077 70 5.412091
000000078 76 5.295454
000000079 76 5.395715
000000080 76 5.950446
000000081 70 5.920339
000000082 70 5.529576
15 record(s) selected [Fetch MetaData: 0/ms] [Fetch Data: 0/ms]
[Executed: 8/6/03 12:05:52 AM CDT ] [Execution: 93/ms]
"Kory Yingling" <Mister2zx3@.yahoo.com> wrote in message
news:uVgDQB9WDHA.2268@.TK2MSFTNGP11.phx.gbl...
> I can get the lowest for each with the following but how can I get the
> lowest 2?
> SELECT TOP 15 t1.MastNUM, t1.Store, t1.Distance
> FROM DistTest t1,
> ( SELECT MastNUM, MIN(Distance) AS MinDist
> FROM DistTest
> GROUP BY MastNum ) as Dmin
> WHERE t1.MastNum = Dmin.MastNUm AND t1.Distance = Dmin.MinDist
> go
> MastNUM Store Distance
> -- -- --
> 000000067 76 3.358330
> 000000068 76 4.447685
> 000000069 70 3.836682
> 000000070 76 3.729323
> 000000071 76 4.046238
> 000000072 70 3.928709
> 000000073 76 4.663551
> 000000074 76 3.206388
> 000000076 76 4.745636
> 000000077 76 5.338428
> 000000078 70 5.121837
> 000000079 70 4.580213
> 000000080 70 4.338181
> 000000081 76 4.069455
> 000000082 76 4.465975
>
> "Kory Yingling" <Mister2zx3@.yahoo.com> wrote in message
> news:%23OJC2n8WDHA.536@.TK2MSFTNGP10.phx.gbl...
> > Below is a query that I need to modify so that it returns just the 2
> > MastNUMs, Stores, and Distances where the Distances are the two lowest
for
> a
> > mastnum. So the results would be more like this
> > MastNUM Store Distance
> > -- -- --
> > 000000067 76 3.358330
> > 000000067 70 7.082444
> > 000000068 76 4.447685
> > 000000068 70 5.516853
> > 000000069 70 3.836682
> > 000000069 76 6.331691
> > 000000070 76 3.729323
> >
> > SELECT TOP 15 MastNUM, Store, Distance
> > FROM DistTest
> > GROUP BY MastNUM, Store, Distance
> > ORDER BY MastNUM, Distance
> >
> > MastNUM Store Distance
> > -- -- --
> > 000000067 76 3.358330
> > 000000067 70 7.082444
> > 000000067 69 8.112116
> > 000000067 112 19.924702
> > 000000068 76 4.447685
> > 000000068 70 5.516853
> > 000000068 69 6.874022
> > 000000068 112 18.622366
> > 000000068 71 19.396528
> > 000000069 70 3.836682
> > 000000069 76 6.331691
> > 000000069 69 8.924779
> > 000000069 112 16.709897
> > 000000069 71 17.462224
> > 000000070 76 3.729323
> >
> > --
> >
> > Kory
> >
> >
>|||The initial step creating this list is as follows:
SELECT g.MASTNUM, s.STORE, master.dbo.DistanceMiles( g.lat, g.long, s.LAT,
s.LONG ) AS Distance
INTO DistTest
FROM geomailing g, geoStore s
WHERE master.dbo.DistanceMiles( g.lat, g.long, s.LAT, s.LONG ) <= 20
GROUP BY g.Mastnum, s.Store, master.dbo.DistanceMiles( g.lat, g.long, s.LAT,
s.LONG )
ORDER BY g.Mastnum, master.dbo.DistanceMiles( g.lat, g.long, s.LAT, s.LONG )
Can anyone suggest a way that it only inserts into this table the 2 lowest
Distances ( master.dbo.DistanceMiles( g.lat, g.long, s.LAT, s.LONG ) ) ? So
as to avoid having to pull out the two lowest distances later?
Thanks.
Kory wrote in message news:%23OJC2n8WDHA.536@.TK2MSFTNGP10.phx.gbl...
> Below is a query that I need to modify so that it returns just the 2
> MastNUMs, Stores, and Distances where the Distances are the two lowest for
a
> mastnum. So the results would be more like this
> MastNUM Store Distance
> -- -- --
> 000000067 76 3.358330
> 000000067 70 7.082444
> 000000068 76 4.447685
> 000000068 70 5.516853
> 000000069 70 3.836682
> 000000069 76 6.331691
> 000000070 76 3.729323
> SELECT TOP 15 MastNUM, Store, Distance
> FROM DistTest
> GROUP BY MastNUM, Store, Distance
> ORDER BY MastNUM, Distance
> MastNUM Store Distance
> -- -- --
> 000000067 76 3.358330
> 000000067 70 7.082444
> 000000067 69 8.112116
> 000000067 112 19.924702
> 000000068 76 4.447685
> 000000068 70 5.516853
> 000000068 69 6.874022
> 000000068 112 18.622366
> 000000068 71 19.396528
> 000000069 70 3.836682
> 000000069 76 6.331691
> 000000069 69 8.924779
> 000000069 112 16.709897
> 000000069 71 17.462224
> 000000070 76 3.729323
> --
> Kory
>|||Does anyone have any suggestions on how to get just the 2 lowest values
grouped?|||I would imagine a lot of people have a lot of ideas but with what you have
given us it would be stabbing in the dark.
1. Can you post a simple table structure
2. Post sample data to enter in
3. Tell us what you expect to see as the end result.
--
Allan Mitchell (Microsoft SQL Server MVP)
MCSE,MCDBA
www.SQLDTS.com
I support PASS - the definitive, global community
for SQL Server professionals - http://www.sqlpass.org
"Kory Yingling" <kory@.removeme-mlsc.com> wrote in message
news:OPqoHUDXDHA.208@.tk2msftngp13.phx.gbl...
> Does anyone have any suggestions on how to get just the 2 lowest values
> grouped?
>

Return 0 if null

Hi all,
I have got a query that returns values based on a date range and grouped by
week. I need to return a value of 0 if there are no entries for that week. At
the moment it just skips that week all together.
Any ideas?
Thanks
Without more details, I'd suggest lookign into IsNull( variable, 0 ) or a
left outer join if you're using multiple tables and a lack of entries for
that week table is eliminating the week row. But to give a specific answer,
we'd need more details.
-Paul Nielsen, SQL Server MVP
SQL Server 2000 Bible, Wiley Press
"Andrew Jurgens" <AndrewJurgens@.discussions.microsoft.com> wrote in message
news:EBB765D2-2A3D-4151-A241-0B52A9B6E66C@.microsoft.com...
> Hi all,
> I have got a query that returns values based on a date range and grouped
> by
> week. I need to return a value of 0 if there are no entries for that week.
> At
> the moment it just skips that week all together.
> Any ideas?
> Thanks
|||Hi Paul,
My current query is as follows.
SELECT Format([Counting Type_QRY].Date,'ww') AS Expr1, [Counting
Type_QRY].Branch, [Counting Type_QRY].Type, Sum([Counting
Type_QRY].CountOfType) AS SumOfCountOfType
FROM (Branch INNER JOIN Type ON Branch.ID = Type.ID) INNER JOIN [Counting
Type_QRY] ON Branch.ID = [Counting Type_QRY].Branch
GROUP BY Format([Counting Type_QRY].Date,'ww'), [Counting Type_QRY].Branch,
[Counting Type_QRY].Type;
This works great counting my entries and putting then grouping by week.
Trouble is if there is no info for a week it skipps that week.
34 = 4
35 = 2
37 = 7
I need it to return
34 = 4
35 = 2
36 = 0
37 = 7
I am fairly new to this so please excuse the query if it is not great. Just
trying to grow my skills in the real world.
Thanks
"Paul Nielsen" wrote:

> Without more details, I'd suggest lookign into IsNull( variable, 0 ) or a
> left outer join if you're using multiple tables and a lack of entries for
> that week table is eliminating the week row. But to give a specific answer,
> we'd need more details.
> --
> -Paul Nielsen, SQL Server MVP
> SQL Server 2000 Bible, Wiley Press
>
> "Andrew Jurgens" <AndrewJurgens@.discussions.microsoft.com> wrote in message
> news:EBB765D2-2A3D-4151-A241-0B52A9B6E66C@.microsoft.com...
>
>
|||Andrew Jurgens wrote:
> Hi Paul,
> My current query is as follows.
> SELECT Format([Counting Type_QRY].Date,'ww') AS Expr1, [Counting
> Type_QRY].Branch, [Counting Type_QRY].Type, Sum([Counting
> Type_QRY].CountOfType) AS SumOfCountOfType
> FROM (Branch INNER JOIN Type ON Branch.ID = Type.ID) INNER JOIN
> [Counting Type_QRY] ON Branch.ID = [Counting Type_QRY].Branch
> GROUP BY Format([Counting Type_QRY].Date,'ww'), [Counting
> Type_QRY].Branch, [Counting Type_QRY].Type;
> This works great counting my entries and putting then grouping by
> week. Trouble is if there is no info for a week it skipps that week.
> 34 = 4
> 35 = 2
> 37 = 7
> I need it to return
> 34 = 4
> 35 = 2
> 36 = 0
> 37 = 7
> I am fairly new to this so please excuse the query if it is not
> great. Just trying to grow my skills in the real world.
> Thanks
Sounds like you need an OUTER JOIN. I don't know your data, so use an
outer join against the table that may not have a foreign key
relationship. If you wanted all Accounts from the accounts table even if
some of the accounts didn't have a comment in the comments table, you
would:
From Accounts Outer Join Comments on Accounts.id = Comments.id
David Gugick
Imceda Software
www.imceda.com
|||Thanks David,
My problem is that I am querying a date range but there may not be entries
for all dates within that range in the table. I still need to return all
dates even if there is no data. Hence my previous
This works great counting my entries and putting then grouping by[vbcol=seagreen]
I really appreciate your input. Have been stuck for a bit and need to get
out of this hole!
Thanks.
"David Gugick" wrote:

> Andrew Jurgens wrote:
>
> Sounds like you need an OUTER JOIN. I don't know your data, so use an
> outer join against the table that may not have a foreign key
> relationship. If you wanted all Accounts from the accounts table even if
> some of the accounts didn't have a comment in the comments table, you
> would:
> From Accounts Outer Join Comments on Accounts.id = Comments.id
>
> --
> David Gugick
> Imceda Software
> www.imceda.com
>
|||On Thu, 14 Oct 2004 01:59:20 -0700, Andrew Jurgens wrote:

>Thanks David,
>My problem is that I am querying a date range but there may not be entries
>for all dates within that range in the table. I still need to return all
>dates even if there is no data. Hence my previous
Hi Andrew,
Looks like you need a calendar table.
See http://www.aspfaq.com/show.asp?id=2516.
Best, Hugo
(Remove _NO_ and _SPAM_ to get my e-mail address)

Retrive old JobHistory

I have a job which only keeps 100 entries in the job history. When one new
entry comes, the oldest entry in the list is deleted.
I need retrieve the old job histories that have been removed from the
current job history list. How can I do it? Restore DB is not an option.
Thanks a lot,
LixinYou will need to restore msdb(as different db) and look at the table
sysjobhistory, I recommend you iether increase the number of history rows
for that job or write a job which copies the rows to an archive table.
Yovan
"Lixin Fan" <nospam@.hotmail.com> wrote in message
news:uYx1XbouDHA.2712@.tk2msftngp13.phx.gbl...
> I have a job which only keeps 100 entries in the job history. When one new
> entry comes, the oldest entry in the list is deleted.
> I need retrieve the old job histories that have been removed from the
> current job history list. How can I do it? Restore DB is not an option.
> Thanks a lot,
> Lixin
>|||Hi Lixin
Thank you for using MSDN Newsgroup! It's my pleasure to assist you with
your issue.
I think a trigger will be a workaround to you question.
Suppose you want to record the job history of Job_test, and the job_id is
={07C00C79-AAB0-49B0-BA0A-9E7C884440DB} ( you can get this information from
two tables in database 'msdb': 'sysjob' and 'sysjobhistory').
Suppose you have a table 'job_history' in database 'Database_test', and the
table 'job_history' has the same schema with the table 'sysjobhistory' in
msdb and you want to use this table to save the job history. You can add a
trigger to the sysjobhistory.
The thinking is: when the system add a new entry into the 'sysjobhistory',
check if this new entry is for job_test. If no, then take no action. If
yes, then the trigger will copy this entry in the job_history table in
database_test. By this way, you can record all the job history when the job
occurred.
The code for the trigger would be like this:
CREATE TRIGGER [jobtest] ON [dbo].[sysjobhistory]
FOR insert
AS
Declare @.job_date int
Declare @.job_time int
Declare @.job_idx varchar
Set @.job_idx=' ({07C00C79-AAB0-49B0-BA0A-9E7C884440DB}'
If @.job_idx<> (select top 1 job_id from [msdb].[dbo].[rundate] where
job_id=@.job_idx order by rundate,runtime asc)
--if the latest entry in the sysjobhistory is not caused by the job you
want to record
--no action
Begin
Return
End
--If the job you want to record add a new entry
Else
Begin
Insert [database_test].[dbo].[job_history].[rundate] value (select top 1 *
from [msdb].[dbo].[runtime] where job_id=@.job_idx order by rundate,runtime
asc
End
Note:
1) This code is just a thinking, not a tested one
2) This method will add the workload on your system, especially on a system
which has many concurrent jobs.
I hope this will help you to solve your problem. If you still have
questions, please feel free to post message here and I am ready to help!
Best regards
Baisong Wei
Microsoft Online Support
----
Get Secure! - www.microsoft.com/security
This posting is provided "as is" with no warranties and confers no rights.
Please reply to newsgroups only. Thanks.|||Great help. Thank you very much.
Lixin
"Baisong Wei[MSFT]" <v-baiwei@.online.microsoft.com> wrote in message
news:NMNq6HxuDHA.2900@.cpmsftngxa07.phx.gbl...
> Hi Lixin
> Thank you for using MSDN Newsgroup! It's my pleasure to assist you with
> your issue.
> I think a trigger will be a workaround to you question.
> Suppose you want to record the job history of Job_test, and the job_id is
> ={07C00C79-AAB0-49B0-BA0A-9E7C884440DB} ( you can get this information
from
> two tables in database 'msdb': 'sysjob' and 'sysjobhistory').
> Suppose you have a table 'job_history' in database 'Database_test', and
the
> table 'job_history' has the same schema with the table 'sysjobhistory' in
> msdb and you want to use this table to save the job history. You can add a
> trigger to the sysjobhistory.
> The thinking is: when the system add a new entry into the 'sysjobhistory',
> check if this new entry is for job_test. If no, then take no action. If
> yes, then the trigger will copy this entry in the job_history table in
> database_test. By this way, you can record all the job history when the
job
> occurred.
> The code for the trigger would be like this:
> CREATE TRIGGER [jobtest] ON [dbo].[sysjobhistory]
> FOR insert
> AS
> Declare @.job_date int
> Declare @.job_time int
> Declare @.job_idx varchar
> Set @.job_idx=' ({07C00C79-AAB0-49B0-BA0A-9E7C884440DB}'
> If @.job_idx<> (select top 1 job_id from [msdb].[dbo].[rundate] where
> job_id=@.job_idx order by rundate,runtime asc)
> --if the latest entry in the sysjobhistory is not caused by the job you
> want to record
> --no action
> Begin
> Return
> End
>
> --If the job you want to record add a new entry
> Else
> Begin
> Insert [database_test].[dbo].[job_history].[rundate] value (select top 1 *
> from [msdb].[dbo].[runtime] where job_id=@.job_idx order by rundate,runtime
> asc
> End
> Note:
> 1) This code is just a thinking, not a tested one
> 2) This method will add the workload on your system, especially on a
system
> which has many concurrent jobs.
> I hope this will help you to solve your problem. If you still have
> questions, please feel free to post message here and I am ready to help!
>
> Best regards
> Baisong Wei
> Microsoft Online Support
> ----
> Get Secure! - www.microsoft.com/security
> This posting is provided "as is" with no warranties and confers no rights.
> Please reply to newsgroups only. Thanks.
>

retrieving data from table with 7 million entries takes time

Can anyone help me on this...
when i select data from table using select statement it takes huge amount of time....The table contains 7 million entries and when i select by mentioning a criteria it takes around 45 secs..The system has 4GB RAM and Dual Processing CPU. The select statement does not contain any grouping and all..

Will it take this much time to retrieve data.?.
The table does include an indexed field,
So can anyone help me on the different things i can do to make the retrieval faster?

Andy

Hi Andy,

Could you post the DDL please, including the the definintion of the indexes and the query you're running. Having an index doesn't help you if you don't select on fields that are in the index. (And even then....)

Also, pulling all 7 million rows over the network can take some time, especially when the rows are 'wide' or the network is slow. Are you selecting all the rows or just a small subset?

What you can also use is Tools > Database Engine Tuning Advisor in Mamagement studio. I haven't used it, but for these kind of question it could very well help you out.

Regards,

GJ

|||When is used a criteria (WHERE clause) you force engine to use that index that mean speed. When is not, the system make full table scan and it have to go to 1,2,3 , to the 7 million records that need time. So, you have to build or invent an appropriate WHERE clause to speed and resolve all user request.|||

o The following query will get the 10 missing indexes would produce the highest anticipated cumulative improvement, in descending order, for user queries.

SELECT TOP 10 *

FROM sys.dm_db_missing_index_group_stats

ORDER BY avg_total_user_cost * avg_user_impact * (user_seeks + user_scans)DESC

You can get the missing index details in the following way:

The following query determines which missing indexes comprise a particular missing index group, and displays their column details.

For the sake of this example, the missing index group handle is 24.(You will need to change the handle value with handle values which comes up from the earlier query)

SELECT migs.group_handle, mid.*

FROM sys.dm_db_missing_index_group_stats migs

INNER JOIN sys.dm_db_missing_index_groups mig

ON (migs.group_handle = mig.index_group_handle)

INNER JOIN sys.dm_db_missing_index_details mid

ON (mig.index_handle = mid.index_handle)

WHERE migs.group_handle = 24 <<put your handle value here>>

For details on this refer to the following articles:

http://msdn2.microsoft.com/en-us/library/ms345421.aspx

Using Missing Index Information to Write CREATE INDEX Statements

http://msdn2.microsoft.com/en-us/library/ms345405.aspx