Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
Author |
Topic |
Hillside
Starting Member
23 Posts |
Posted - 2007-08-10 : 09:35:39
|
Hi,How to find island and gaps in the below mentioned table.The row difference should exactly week difference with each row.DECLARE @Table1 TABLE (t06_t06_id INT, dte DATETIME, t06_value decimal(10,8))INSERT @Table1 SELECT 1, '2002-06-26 00:00:00', 5.1000 Union allSELECT 1, '2002-07-10 00:00:00', 5.1125 Union allSELECT 1, '2002-07-17 00:00:00', 5.0525 Union allSELECT 1, '2002-07-24 00:00:00', 4.8875 Union allSELECT 1, '2002-07-31 00:00:00', 4.8944 Union allSELECT 1, '2002-08-07 00:00:00', 4.7313 Union allSELECT 1, '2002-08-21 00:00:00', 4.8461 Union allSELECT 1, '2002-08-28 00:00:00', 4.7717 Union allSELECT 1, '2002-09-04 00:00:00', 4.5795 Union allSELECT 1, '2002-09-11 00:00:00', 4.6759 Union allSELECT 1, '2002-09-18 00:00:00', 4.4804 Union allSELECT 1, '2002-09-25 00:00:00', 4.4597 Union allSELECT 1, '2002-10-02 00:00:00', 4.5278 Union all SELECT 2, '2002-06-26 00:00:00', 5.1000 Union allSELECT 2, '2002-07-10 00:00:00', 5.1125 Union allSELECT 2, '2002-07-17 00:00:00', 5.0525 Union allSELECT 2, '2002-07-24 00:00:00', 4.8875 Union allSELECT 2, '2002-07-31 00:00:00', 4.8944 Union allSELECT 2, '2002-08-07 00:00:00', 4.7313 Union allSELECT 2, '2002-08-21 00:00:00', 4.8461 Union allSELECT 2, '2002-08-28 00:00:00', 4.7717 Union allSELECT 2, '2002-09-04 00:00:00', 4.5795 Union allSELECT 2, '2002-09-11 00:00:00', 4.6759 Union allSELECT 2, '2002-09-18 00:00:00', 4.4804 Union allSELECT 2, '2002-09-25 00:00:00', 4.4597 Union allSELECT 2, '2002-10-02 00:00:00', 4.5278Thanks |
|
khtan
In (Som, Ni, Yak)
17689 Posts |
|
Hillside
Starting Member
23 Posts |
Posted - 2007-08-14 : 06:03:35
|
hi,Thanks for your reply.I have checked your URL and I did not found soultion to my problem.As i said alreayd that I am looking date gaps in the above series.Thanks |
 |
|
blindman
Master Smack Fu Yak Hacker
2365 Posts |
Posted - 2007-08-14 : 10:46:23
|
Then you need to tell us exactly what output you would expect from your sample data.e4 d5 xd5 Nf6 |
 |
|
Hillside
Starting Member
23 Posts |
Posted - 2007-08-14 : 10:56:51
|
Thanks for your reply.These dates are has to be exactly week difference and We need to check wether these date series has got exact weeke difference.If the dates difference more than week I need find those date gaps with ID.For example:SELECT 1, '2002-06-26 00:00:00', 5.1000 Union allSELECT 1, '2002-07-03 00:00:00', 5.1125 Union all Need find this top and bottom date gapsSELECT 1, '2002-07-10 00:00:00', 5.1125 Union all |
 |
|
blindman
Master Smack Fu Yak Hacker
2365 Posts |
Posted - 2007-08-14 : 11:46:14
|
You still seem unclear on the meaning of the term "exactly".The link that was posted illustrated three methods for doing the sort of task you mention. If none of these were satisfactory, nobody else on the forum is going to waste there time trying to guess what you want.e4 d5 xd5 Nf6 |
 |
|
Hillside
Starting Member
23 Posts |
Posted - 2007-08-14 : 12:11:21
|
Hi,The date series which i have mentioned in my previous posting are weekly dates(7 days a week ) and I need find out if any one of the "week date" along with "values" is missing for particular IDS.Weekly date means:Every week we will receive some data and if data is missed on particular week than that will be missed in the above date series.I need find these date gaps by using ISLAND and GAPS methodHope this is very much clear to you....thanks |
 |
|
blindman
Master Smack Fu Yak Hacker
2365 Posts |
Posted - 2007-08-14 : 15:04:23
|
Nope. As far as I can see, the link provided by khtan satisfies this requirement.There is a general principle in sql coding: If you can't explain what you want, you can't code what you want.e4 d5 xd5 Nf6 |
 |
|
Hillside
Starting Member
23 Posts |
Posted - 2007-08-15 : 05:25:28
|
Hi,Please go through carefully in Khtan has provided solution to missing IDs(Like Identity column) and not the weekly date missing.In his second URL it will give the out of variety of dates.Still if you insist the solution is there then please let me know....how to approachThanks |
 |
|
khtan
In (Som, Ni, Yak)
17689 Posts |
Posted - 2007-08-15 : 08:09:56
|
the F_TABLE_DATE LEFT JOIN method
SELECT d.[DATE]FROM F_TABLE_DATE('2002-06-26', '2002-10-02') d left JOIN @Table1 t ON d.[DATE] = t.dteWHERE d.DAY_OF_WEEK = 4AND t.dte IS NULL KH[spoiler]Time is always against us[/spoiler] |
 |
|
blindman
Master Smack Fu Yak Hacker
2365 Posts |
Posted - 2007-08-15 : 11:30:08
|
quote: Originally posted by Hillside Please go through carefully in Khtan has provided solution to missing IDs(Like Identity column) and not the weekly date missing.
The algorithm is the same.e4 d5 xd5 Nf6 |
 |
|
|
|
|
|
|