Need information on this Query

Hi guys,
can someone explain how this query works indetail.
SELECT DISTINCT (a.sal) FROM EMP A WHERE &N = (SELECT COUNT (DISTINCT (b.sal)) FROM EMP B WHERE a.sal<=b.sal);
Advance thanks for your help.
Kadiyala
Edited by: user10223931 on Mar 4, 2009 8:17 PM

Hi,
SELECT DISTINCT (a.sal),(SELECT COUNT (DISTINCT (b.sal)) FROM EMP B WHERE a.sal<=b.sal) FROM EMP A order by 2 descI hope this will help you.
The query will rank the records as per salary amount. and you will get the salary amount of the specified rank.

Similar Messages

Hello. I just upgraded to Lion OS X 10.7.3. I was previously running 10.5.7 which included the ENTOURAGE application. how do I retrieve needed information from this application that no longer runs on this new system?

Hello. I just upgraded to Lion OS X 10.7.3. I was previously running 10.5.7 which included the ENTOURAGE application. how do I retrieve needed information from this application that no longer runs on this new system?

The file you'll be looking for that has most of what you're looking for - messages, address book, etc. is the Entourage Database. The verison of Microsoft Office I run (on Lion) is Office 2008. By default the Entourage Database for this verison is at user/Documents/Microsoft User Data/Office 2008 Identities/Main Identity/Database. That folder (Main Identitty) also has your rules, signatures, and such. You'll have to recover this file from whatever backup you have. If you have a different version of Entourage, the file location will be a little diffferent.
Good luck
srb

Please need help with this query

Hi !
Please need help with this query:
Needs to show (in cases of more than 1 loan offer) the latest create_date one time.
Meaning, In cases the USER_ID, LOAN_ID, CREATE_DATE are the same need to show only the latest, Thanks!!!
select distinct a.id,
create_date,
a.loanid,
a.rate,
a.pays,
a.gracetime,
a.emailtosend,
d.first_name,
d.last_name,
a.user_id
from CLAL_LOANCALC_DET a,
loan_Calculator b,
bv_user_profile c,
bv_mr_user_profile d
where b.loanid = a.loanid
and c.NET_USER_NO = a.resp_id
and d.user_id = c.user_id
and a.is_partner is null
and a.create_date between
TO_DATE('6/3/2008 01:00:00', 'DD/MM/YY HH24:MI:SS') and
TO_DATE('27/3/2008 23:59:00', 'DD/MM/YY HH24:MI:SS')
order by a.create_date

Perhaps something like this...
select id, create_date, loanid, rate, pays, gracetime, emailtosend, first_name, last_name, user_id
from (
      select distinct a.id,
                      create_date,
                      a.loanid,
                      a.rate,
                      a.pays,
                      a.gracetime,
                      a.emailtosend,
                      d.first_name,
                      d.last_name,
                      a.user_id,
                      max(create_date) over (partition by a.user_id, a.loadid) as max_create_date
      from CLAL_LOANCALC_DET a,
           loan_Calculator b,
           bv_user_profile c,
           bv_mr_user_profile d
      where b.loanid = a.loanid
      and   c.NET_USER_NO = a.resp_id
      and   d.user_id = c.user_id
      and   a.is_partner is null
      and   a.create_date between
            TO_DATE('6/3/2008 01:00:00', 'DD/MM/YY HH24:MI:SS') and
            TO_DATE('27/3/2008 23:59:00', 'DD/MM/YY HH24:MI:SS')
where create_date = max_create_date
order by create_date

I need solution for this query

hi all,
could anyone please send me solution for this query these are the database tables am having
TABLE NAME :USERS
ATTRIBUTES
UNAME
PASSWORD
GROUPNAME
TABLE NAME:GROUPS
ATTRIBUTES
GROUPID
GROUPNAME
my requirement is that i need 2 acces the groupname of a particular user and the reamining groups to which i doesn't belong in A SINGLE QUERY
my result needs to be like this
Authorised group
consultant
UNAuthorised groups
sales
vender
recruiter
admin
if any body got the solution please send

hi All,
I have 3 tables 1)PREVILEGES (groupname, previleges(values y or n only))
2)GROUPS (groupid, groupname)
3) USERS (uname, groupname).
Here each user belongs to one group, each user hav a default previlege means example if user is consutant then he can access only consultant group. Means default previlege will be used. For default previleges, there is no record in Previleges table.
The Problem is that, I need to reterive the groupname from users which de doesn't belong as well as his previleges from previleges, If there is no values in previleges in the table it should return n. or value what is therey.
the different groups are
SALES,CONSULTANT,VENDER,RECRUTER,ADMIN

Need help in this query using Case Statement

I have the following query which is currently existing and I am adding few more conditions based on the new requirements which is based on a particular flag set as 1 or 0.
If it is set to 1 then I should use the old query as it is and if it is set to 0 then I should add the new conditions.
Basically when the flag is set to 0, I shouldnt be including some of the records that already exists and should include only new records. This is based on the plan_type_ids in (1,2,3,4).
Hence I am using the Case statement to check if the plan_type_id is in (1,2) then do a set of not exists and if the plan_type_id in (3,4) then do set of not exists.
But when I run this query it is giving me error. What am I doing wrong?
Is there any other simple way to combine all the not exists for all of those select statements as given after the line ------------------------- into a single one?
What am I doing wrong?
I tried putting the NOT EXists before the case too but that too didnt work.
Please help. Appreciate it.
Thank you in advance.
SELECT
ee.employee_id
,'WELCOMEMSG'
,DECODE( me.member_enrollment_id
,first_enr.enrollment_id
,20
,23
) status_id
,me.member_enrollment_id
,wk.welcome_msg_id
FROM wk
,employees ee
,MEMBER_ENROLLMENTS me
,plans pl
,( SELECT employee_id
,plan_type_id
,start_date
,plan_id
,MIN(MEMBER_ENROLLMENT_ID) member_enrollment_id
FROM ( SELECT me.employee_id
,DECODE(pl.plan_type_id,1,2,pl.plan_type_id) plan_type_id
,pl.start_date
,wk.plan_id
,me.member_enrollment_id
FROM wk
,PLANS pl
,MEMBER_ENROLLMENTS me
WHERE wk.done_by = nvl('TEST' ,wk.done_by)
AND wk.welcome_msg_name <> 'NONE'
AND pl.employer_id = wk.employer_id
AND wk.employer_id = 5
AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = wk.plan_type_id
AND pl.plan_id = NVL(wk.plan_id,pl.plan_id)
AND me.plan_id = pl.plan_id
AND me.coverage_effective_date <> NVL(me.coverage_end_Date, me.coverage_effective_date + 1)
AND me.coverage_effective_Date BETWEEN wk.start_date AND NVL(wk.end_date, me.coverage_effective_date + 1)
MINUS
SELECT me.employee_id
,DECODE(pl.plan_type_id,1,2,pl.plan_type_id) plan_type_id
,pl.start_date
,NULL plan_id
,me.member_enrollment_id
FROM wk
,PLANS pl
,MEMBER_ENROLLMENTS me
WHERE wk.done_by = nvl(NULL,wk.done_by)
AND wk.welcome_msg_name <> 'NONE'
AND pl.employer_id = wk.employer_id
AND wk.employer_id = 5
AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = wk.plan_type_id
AND pl.plan_id = wk.plan_id
AND me.plan_id = pl.plan_id
AND me.coverage_effective_date <> NVL(me.coverage_end_Date, me.coverage_effective_date + 1)
AND me.coverage_effective_Date BETWEEN wk.start_date AND NVL(wk.end_date, me.coverage_effective_date + 1)
WHERE employee_id = 100
GROUP BY employee_id ,plan_type_id,start_date ,plan_id
)first_enr
,MEMBER_EVENTS mv
WHERE wk.done_by = nvl(NULL,wk.done_by)
AND wk.employer_id = ee.employer_id
AND ee.employee_id = me.employee_id
AND ee.employee_id = 100
AND me.plan_id = pl.plan_id
AND me.coverage_effective_date <> NVL(me.coverage_end_Date, me.coverage_effective_date + 1)
AND me.coverage_effective_Date BETWEEN wk.start_date AND NVL(wk.end_date, me.coverage_effective_date + 1)
AND is_expired(me.employee_id,me.plan_id) = 'Y'
AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = wk.plan_type_id
AND pl.plan_id=nvl(wk.plan_id,pl.plan_id)
AND me.employee_id = first_enr.employee_id
AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = first_enr.plan_type_id
AND pl.start_date = first_enr.start_date
AND me.member_enrollment_id = mv.member_enrollment_id (+)
AND 'WELCOMEMSG' = mv.event_name(+)
AND mv.member_enrollment_id IS NULL
AND wk.welcome_msg_name <> 'NONE'
AND NVL(first_enr.plan_id,0) = NVL( wk.plan_id,0)
AND (FN_get_all_participant(wk.employer_id) = 1
OR
(FN_get_all_participant(wk.employer_id) = 0
AND (CASE WHEN pl.plan_type_id IN (1,2)
THEN NOT EXISTS (SELECT 'X'
FROM member_accounts ma
member_enrollments men3
plans pl3
plan_types pt3
WHERE ma.member_account_id = men3.member_account_id
AND ma.employee_id = me.employee_id
AND ma.plan_id = pl3.plan_id
AND pl3.start_date < pl.START_DATE
AND TRUNC(men3.coverage_effective_date) <> TRUNC(NVL(men3.coverage_end_date, men3.coverage_effective_date + 1 ))
AND pl3.plan_type_id = pt3.plan_type_id
AND pt3.plan_type_id in (1, 2)
UNION
(SELECT 'X'
FROM member_accounts ma
member_enrollments men3
plans pl3
plan_types pt3
WHERE ma.member_account_id = men3.member_account_id
AND ma.employee_id = me.employee_id
AND ma.plan_id = pl3.plan_id
AND pl3.start_date = (pl.start_date - 365)
\ AND TRUNC(men3.coverage_effective_date) <> TRUNC(NVL(men3.coverage_end_date, men3.coverage_effective_date + 1 ))
AND pl3.plan_type_id = pt3.plan_type_id
AND pt3.plan_type_id = wk.plan_type_id
UNION
(SELECT 'X'
FROM member_accounts ma
member_enrollments men3
plans pl3
plan_types pt3
WHERE ma.member_account_id = men3.member_account_id
AND ma.employee_id = men2.employee_id
AND ma.plan_id = pl3.plan_id
AND pl3.start_date < pl2.START_DATE -- '01-Jan-2011'
AND TRUNC(men3.coverage_effective_date) <> TRUNC(NVL(men3.coverage_end_date, men3.coverage_effective_date + 1 ))
AND pl3.plan_type_id = pt3.plan_type_id
AND pt3.plan_type_id = 2
UNION
(SELECT 'X'
FROM member_accounts ma
member_enrollments men3
plans pl3
plan_types pt3
WHERE ma.member_account_id = men3.member_account_id
AND ma.employee_id = men2.employee_id
AND ma.plan_id = pl3.plan_id
AND pl3.start_date < pl2.START_DATE -- '01-Jan-2011'
AND TRUNC(men3.coverage_effective_date) <> TRUNC(NVL(men3.coverage_end_date, men3.coverage_effective_date + 1 ))
AND pl3.plan_type_id = pt3.plan_type_id
AND pt3.plan_type_id = 1
WHEN pl.plan_type_id IN (3, 4)
THEN NOT EXISTS (SELECT 'X'
FROM member_accounts ma
member_enrollments men3
plans pl3
plan_types pt3
WHERE ma.member_account_id = men3.member_account_id
AND ma.employee_id = men2.employee_id
AND nvl(ma.account_end_date, sysdate) <= trunc(sysdate)
AND ma.plan_id = pl3.plan_id
AND pl3.start_date <= pl2.START_DATE
AND TRUNC(men3.coverage_effective_date) <> TRUNC(NVL(men3.coverage_end_date, men3.coverage_effective_date + 1 ))
AND pl3.plan_type_id = pt3.plan_type_id
AND pt3.plan_type_id in (3, 4)
END
AND (CASE WHEN pl.plan_type_id IN (1,2)
ERROR at line 89:
ORA-00936: missing expression

Maybe
SELECT ee.employee_id,
       'WELCOMEMSG',
       DECODE(me.member_enrollment_id,first_enr.enrollment_id,20,23) status_id,
       me.member_enrollment_id,
       wk.welcome_msg_id
FROM wk,
       employees ee,
       MEMBER_ENROLLMENTS me,
       plans pl,
       (SELECT employee_id,
               plan_type_id,
               start_date,
               plan_id,
               MIN(MEMBER_ENROLLMENT_ID) member_enrollment_id
          FROM (SELECT me.employee_id,
                       DECODE(pl.plan_type_id,1,2,pl.plan_type_id) plan_type_id,
                       pl.start_date,
                       wk.plan_id,
                       me.member_enrollment_id
                  FROM wk,
                       PLANS pl,
                       MEMBER_ENROLLMENTS me
                 WHERE wk.done_by = nvl('TEST',wk.done_by) /* same as wk.done_by = 'TEST' */
                   AND wk.welcome_msg_name 'NONE'
                   AND pl.employer_id = wk.employer_id
                   AND wk.employer_id = 5
                   AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = wk.plan_type_id
                   AND pl.plan_id = NVL(wk.plan_id,pl.plan_id)
                   AND me.plan_id = pl.plan_id
                   AND me.coverage_effective_date != NVL(me.coverage_end_Date,me.coverage_effective_date + 1)
                   AND me.coverage_effective_Date BETWEEN wk.start_date AND NVL(wk.end_date,me.coverage_effective_date + 1)
                MINUS
                SELECT me.employee_id,
                       DECODE(pl.plan_type_id,1,2,pl.plan_type_id) plan_type_id,
                       pl.start_date,
                       NULL plan_id,
                       me.member_enrollment_id
                  FROM wk,
                       PLANS pl,
                       MEMBER_ENROLLMENTS me
                 WHERE wk.done_by = nvl(NULL,wk.done_by) /* same as 1 = 1 */
                   AND wk.welcome_msg_name 'NONE'
                   AND pl.employer_id = wk.employer_id
                   AND wk.employer_id = 5
                   AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = wk.plan_type_id
                   AND pl.plan_id = wk.plan_id
                   AND me.plan_id = pl.plan_id
                   AND me.coverage_effective_date NVL(me.coverage_end_Date,me.coverage_effective_date + 1)
                   AND me.coverage_effective_Date BETWEEN wk.start_date AND NVL(wk.end_date, me.coverage_effective_date + 1)
         WHERE employee_id = 100
         GROUP BY employee_id,
                  plan_type_id,
                  start_date,
                  plan_id
       ) first_enr,
       MEMBER_EVENTS mv
WHERE wk.done_by = nvl(NULL,wk.done_by)
   AND wk.employer_id = ee.employer_id
   AND ee.employee_id = me.employee_id
   AND ee.employee_id = 100
   AND me.plan_id = pl.plan_id
   AND me.coverage_effective_date != NVL(me.coverage_end_Date,me.coverage_effective_date + 1)
   AND me.coverage_effective_Date BETWEEN wk.start_date AND NVL(wk.end_date, me.coverage_effective_date + 1)
   AND is_expired(me.employee_id,me.plan_id) = 'Y'
   AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = wk.plan_type_id
   AND pl.plan_id = nvl(wk.plan_id,pl.plan_id)
   AND me.employee_id = first_enr.employee_id
   AND DECODE(pl.plan_type_id,1,2,pl.plan_type_id) = first_enr.plan_type_id
   AND pl.start_date = first_enr.start_date
   AND me.member_enrollment_id = mv.member_enrollment_id(+)
   AND 'WELCOMEMSG' = mv.event_name(+)
   AND mv.member_enrollment_id IS NULL
   AND wk.welcome_msg_name != 'NONE'
   AND NVL(first_enr.plan_id,0) = NVL(wk.plan_id,0)
   AND (
        FN_get_all_participant(wk.employer_id) = 1
    OR
        (FN_get_all_participant(wk.employer_id) = 0
   AND   NOT EXISTS(SELECT 'X'
                      FROM member_accounts ma,
                           member_enrollments men3,
                           plans pl3,
                           plan_types pt3
                     WHERE ma.member_account_id = men3.member_account_id
                       AND ma.employee_id = me.employee_id
                       AND ma.plan_id = pl3.plan_id
                       AND pl3.start_date < pl.START_DATE
                       AND TRUNC(men3.coverage_effective_date) != TRUNC(NVL(men3.coverage_end_date,men3.coverage_effective_date + 1))
                       AND pl3.plan_type_id = pt3.plan_type_id
                       AND pt3.plan_type_id in (1,2)
                       and pl.plan_type_id IN (1,2)
                    UNION
                    SELECT 'X'
                      FROM member_accounts ma,
                           member_enrollments men3,
                           plans pl3,
                           plan_types pt3
                     WHERE ma.member_account_id = men3.member_account_id
                       AND ma.employee_id = me.employee_id
                       AND ma.plan_id = pl3.plan_id
                       AND pl3.start_date = (pl.start_date - 365)
                       AND TRUNC(men3.coverage_effective_date) != TRUNC(NVL(men3.coverage_end_date,men3.coverage_effective_date + 1))
                       AND pl3.plan_type_id = pt3.plan_type_id
                       AND pt3.plan_type_id = wk.plan_type_id
                       and pl.plan_type_id IN (1,2)
                    UNION
                    SELECT 'X'
                      FROM member_accounts ma,
                           member_enrollments men3,
                           plans pl3,
                           plan_types pt3
                     WHERE ma.member_account_id = men3.member_account_id
                       AND ma.employee_id = men2.employee_id
                       AND ma.plan_id = pl3.plan_id
                       AND pl3.start_date < pl2.START_DATE -- '01-Jan-2011'
                       AND TRUNC(men3.coverage_effective_date) != TRUNC(NVL(men3.coverage_end_date,men3.coverage_effective_date + 1))
                       AND pl3.plan_type_id = pt3.plan_type_id
                       AND pt3.plan_type_id = (1,2)
                       and pl.plan_type_id IN (1,2)
                    UNION
                    SELECT 'X'
                      FROM member_accounts ma,
                           member_enrollments men3,
                           plans pl3,
                           plan_types pt3
                     WHERE ma.member_account_id = men3.member_account_id
                       AND ma.employee_id = men2.employee_id
                       AND trunc(nvl(ma.account_end_date,sysdate)) <= trunc(sysdate)
                       AND ma.plan_id = pl3.plan_id
                       AND pl3.start_date <= pl2.START_DATE
                       AND TRUNC(men3.coverage_effective_date) != TRUNC(NVL(men3.coverage_end_date,men3.coverage_effective_date + 1))
                       AND pl3.plan_type_id = pt3.plan_type_id
                       AND pt3.plan_type_id in (3,4)
                       and pl.plan_type_id IN (3,4)
       )Regards
Etbin

Need to tune this query

Hi
I dont know anything about tuning
can anybody help to tune this query
SQL> set autotrace traceonly explain
SQL> select count(*) from ibs_x_t_receipts c
2 where exists
3 (Select 1 from ibs_s_i_doc_receipt a,ibs_x_t_bill b
4   where A.app_fr_doc_no = c.receipt_no
5     and a.app_to_doc_no = b.doc_no
6     and a.app_to_tr_type = 'TRN001'
7     and A.app_fr_tr_type = 'TRN002'
8     and b.bill_month <= '31-dec-2008'
9     and b.balance_ser_amt = 0
10     and b.balance_tax_amt = 0)
11 /
Execution Plan
   0      SELECT STATEMENT Optimizer=ALL_ROWS (Cost=212880 Card=1 Byte
          s=21)
   1    0   SORT (AGGREGATE)
   2    1     NESTED LOOPS (Cost=212880 Card=1 Bytes=21)
   3    2       VIEW OF 'VW_SQ_1' (VIEW) (Cost=212878 Card=1 Bytes=14)
   4    3         HASH (UNIQUE)
   5    4           NESTED LOOPS
   6    5             NESTED LOOPS (Cost=212878 Card=1 Bytes=66)
   7    6               VIEW OF 'index$_join$_002' (VIEW) (Cost=30021
          Card=91387 Bytes=3564093)
   8    7                 HASH JOIN
   9    8                   INDEX (RANGE SCAN) OF 'SIDOC_INDX1' (INDEX
          ) (Cost=171215 Card=91387 Bytes=3564093)
10    8                   INDEX (RANGE SCAN) OF 'SIDOC_INDX2' (INDEX
          ) (Cost=366130 Card=91387 Bytes=3564093)
11    6               INDEX (UNIQUE SCAN) OF 'ITL_PK2' (INDEX (UNIQU
          E)) (Cost=1 Card=1)
12    5             TABLE ACCESS (BY INDEX ROWID) OF 'IBS_X_T_BILL'
          (TABLE) (Cost=2 Card=1 Bytes=27)
13    2       INDEX (UNIQUE SCAN) OF 'RECEIPTS_PK2' (INDEX (UNIQUE))
           (Cost=1 Card=1 Bytes=7)it is running like anything, i didn't get the result.
Please help me
kanish

--untested
SELECT COUNT (*)
FROM ibs_x_t_receipts c
WHERE EXISTS (
          SELECT 1
            FROM ibs_s_i_doc_receipt a
           WHERE a.app_fr_doc_no = c.receipt_no
             AND a.app_to_tr_type = 'TRN001'
             AND a.app_fr_tr_type = 'TRN002'
             AND EXISTS (
                    SELECT 1
                      FROM ibs_x_t_bill b
                     WHERE a.app_to_doc_no = b.doc_no
                       AND b.bill_month <= '31-dec-2008'
                       AND b.balance_ser_amt = 0
                       AND b.balance_tax_amt = 0))regards,
friend
Edited by: most wanted!!!! on Apr 17, 2012 3:51 AM

Need Information on af:query component

Hi,
I need some information on the <af:query> component in JDev 11g.
When using ADF 11g's new component, ADF Query (af:query) whether the default "All Queryable Attributes" or on other newly defined View Criterias:
1     What are the disadvantages of this component, given that the result-set - without filtering - is large (between 1350 and 1700 rows, approximately).
2     Does it fetch all rows from the database, then performs filtering upon user's entered search criteria – i.e., hitting the "search" button?
3     Is there a way to add client-side validation on the search criteria attributes?
4     In general: What is the best practice to implement a Search Form in ADF 11g
Kindly provide your inputs for the same.
Thanks,
Vineet

user445872 wrote:
Hi,
I need some information on the <af:query> component in JDev 11g.
When using ADF 11g's new component, ADF Query (af:query) whether the default "All Queryable Attributes" or on other newly defined View Criterias:
1     What are the disadvantages of this component, given that the result-set - without filtering - is large (between 1350 and 1700 rows, approximately).
2     Does it fetch all rows from the database, then performs filtering upon user's entered search criteria – i.e., hitting the "search" button? You could tune your VO not to exeute the query by default (i.e., before applying a search criteria)
check this blog for sample: http://andrejusb-samples.blogspot.com/2009/08/jdevadf-sample-oracle-adf-tuning.html
As an other option, you could have a default view criteria applied and executed to filter to result set if it is applicable for your usecase.
3     Is there a way to add client-side validation on the search criteria attributes?I'm not sure about this though.
4     In general: What is the best practice to implement a Search Form in ADF 11g
In addition check this sample as well. It might be of help: http://andrejusb-samples.blogspot.com/2009/09/jdevadf-sample-programmatical-reset-for.html
regards,
~K

Help needed in tuning this query

This is the SQL i would like to tune for performace...
The table structure is given below.
The table has about 2 million rows.
From the second day onwards only around 10% of records will be between Load_Start_Time and Load_End_Time. Among these around 50% will have THAMES_URN_WITH_ROLE_SUFFIX as 'TP%'. At present there are unique and primary key indexes on THAMES_URN_WITH_ROLE_SUFFIX .
I'm sure that we need a b-tree index on STG_UPDATE_DATE_TIME as it is the column with good selectivity.
I'm primarily concerned about the substr, instr and reverese on THAMES_URN_WITH_ROLE_SUFFIX.
Also pls. let me know if the order of predicates is right.
Thanks in advance.
SELECT TPD_STG_TL_SF_LEGAL_OWNER.THAMES_URN, TPD_STG_TL_SF_LEGAL_OWNER.UPDATE_SOURCE, TPD_STG_TL_SF_LEGAL_OWNER.SOURCE_EXTRACT_DATE_TIME, TPD_STG_TL_SF_LEGAL_OWNER.LEGAL_OWNERSHIP_ISSUE_IND, TPD_STG_TL_SF_LEGAL_OWNER.CONTACTABLE_INDICATOR, TPD_STG_TL_SF_LEGAL_OWNER.ADMIN_CONTROL_INDICATOR, TPD_STG_TL_SF_LEGAL_OWNER.BANKRUPTCY_INDICATOR, TPD_STG_TL_SF_LEGAL_OWNER.ASSIGNED_INDICATOR, TPD_STG_TL_SF_LEGAL_OWNER.IN_TRUST_INDICATOR, TPD_STG_TL_SF_LEGAL_OWNER.DIVORCE_CASE_INDICATOR, TPD_STG_TL_SF_LEGAL_OWNER.POA_COP_INDICATOR, TPD_STG_TL_SF_LEGAL_OWNER.CLEANSING_STATUS, TPD_STG_TL_SF_LEGAL_OWNER.CATEGORY
FROM
TPD_STG_TL_SF_LEGAL_OWNER
WHERE
((TPD_STG_TL_SF_LEGAL_OWNER.STG_UPDATE_DATE_TIME > '$$Load_Start_Time'
AND
TPD_STG_TL_SF_LEGAL_OWNER.STG_UPDATE_DATE_TIME <= '$$Load_End_Time')
OR
(TPD_STG_TL_SF_LEGAL_OWNER.TPDB_PROCESSED_IDENTIFIER != 'Y') )
AND
SUBSTR(TPD_STG_TL_SF_LEGAL_OWNER.THAMES_URN_WITH_ROLE_SUFFIX,1,2)='TP'
AND
SUBSTR(REVERSE( TPD_STG_TL_SF_LEGAL_OWNER.THAMES_URN_WITH_ROLE_SUFFIX),INSTR(REVERSE( TPD_STG_TL_SF_LEGAL_OWNER.THAMES_URN_WITH_ROLE_SUFFIX),'-',1)-1,1)!='S'
========================================
CREATE TABLE TPD_STG_TL_SF_LEGAL_OWNER
THAMES_URN VARCHAR2(35 BYTE) NOT NULL,
UPDATE_SOURCE VARCHAR2(3 BYTE),
SOURCE_EXTRACT_DATE_TIME DATE,
LEGAL_OWNERSHIP_ISSUE_IND CHAR(1 BYTE),
CONTACTABLE_INDICATOR CHAR(1 BYTE),
ADMIN_CONTROL_INDICATOR CHAR(1 BYTE),
BANKRUPTCY_INDICATOR CHAR(1 BYTE),
ASSIGNED_INDICATOR CHAR(1 BYTE),
IN_TRUST_INDICATOR CHAR(1 BYTE),
DIVORCE_CASE_INDICATOR CHAR(1 BYTE),
POA_COP_INDICATOR CHAR(1 BYTE),
THAMES_URN_WITH_ROLE_SUFFIX VARCHAR2(39 BYTE),
THAMES_LEGAL_OWNERSHIP_IND CHAR(1 BYTE),
PRODUCT_HOLDING_ROLE_TYPE VARCHAR2(21 BYTE),
OWNERSHIP_SPLIT_PERCENTAGE NUMBER(3),
SOURCE_SYSTEM VARCHAR2(3 BYTE),
TITLE VARCHAR2(50 BYTE),
FIRST_NAME VARCHAR2(50 BYTE),
MIDDLE_INITIAL VARCHAR2(50 BYTE),
SURNAME VARCHAR2(50 BYTE),
GENDER CHAR(1 BYTE),
BIRTH_DATE DATE,
DEATH_INDICATOR CHAR(1 BYTE),
DEATH_RECORD_DATE DATE,
NATIONAL_INSURANCE_NUMBER VARCHAR2(9 BYTE),
GONE_AWAY_INDICATOR CHAR(1 BYTE),
ADDRESS_LINE_1 VARCHAR2(50 BYTE),
ADDRESS_LINE_2 VARCHAR2(50 BYTE),
ADDRESS_LINE_3 VARCHAR2(50 BYTE),
ADDRESS_LINE_4 VARCHAR2(50 BYTE),
ADDRESS_LINE_5 VARCHAR2(50 BYTE),
POST_CODE VARCHAR2(12 BYTE),
COUNTRY VARCHAR2(50 BYTE),
OVERSEAS_ADDRESS_INDICATOR CHAR(1 BYTE),
ORGANISATION_NAME VARCHAR2(50 BYTE),
TPDB_PROCESSED_IDENTIFIER CHAR(2 BYTE),
CLEANSING_STATUS VARCHAR2(30 BYTE),
CATEGORY VARCHAR2(50 BYTE),
SCHEME_NAME VARCHAR2(50 BYTE),
STG_CREATE_DATE_TIME DATE,
STG_UPDATE_DATE_TIME DATE,
LAST_UPDATED_DATE_TIME DATE
TABLESPACE TPDBS01A_DATA
PCTUSED 0
PCTFREE 10
INITRANS 1
MAXTRANS 255
STORAGE (
INITIAL 128K
NEXT 128K
MINEXTENTS 1
MAXEXTENTS 2147483645
PCTINCREASE 0
BUFFER_POOL DEFAULT
LOGGING
NOCOMPRESS
NOCACHE
NOPARALLEL
MONITORING;
CREATE UNIQUE INDEX TPD_STG_TL_SF_LEGAL_OWNER_PK ON TPD_STG_TL_SF_LEGAL_OWNER
(THAMES_URN_WITH_ROLE_SUFFIX)
LOGGING
TABLESPACE TPDBS01A_DATA
PCTFREE 10
INITRANS 2
MAXTRANS 255
STORAGE (
INITIAL 128K
NEXT 128K
MINEXTENTS 1
MAXEXTENTS 2147483645
PCTINCREASE 0
BUFFER_POOL DEFAULT
NOPARALLEL;
ALTER TABLE TPD_STG_TL_SF_LEGAL_OWNER ADD (
CONSTRAINT TPD_STG_TL_SF_LEGAL_OWNER_PK
PRIMARY KEY
(THAMES_URN_WITH_ROLE_SUFFIX)
USING INDEX
TABLESPACE TPDBS01A_DATA
PCTFREE 10
INITRANS 2
MAXTRANS 255
STORAGE (
INITIAL 128K
NEXT 128K
MINEXTENTS 1
MAXEXTENTS 2147483645
PCTINCREASE 0
));

> If indexes don't help, is the only way to gain REAL
performance gain on this type of query is to
partition the table
If a Full Table Scan on a table takes "too long" (and
this is purely subjective - "too long" is different
for every problem) and indexes can't help, and we
need <100% of the rows, and if partitioning will give
us the performance we desire, is it OK do it ? Also
would STG_UPDATE_DATE_TIME be the best column to do a
index partion ?
For this type of query, partitioning on STG_UPDATE_DATE_TIME would be good.
However, partitioning is not done just for one query. It has to fit in your total design. For example, if you are also doing queries with a predicate like <column with a local unique index but not the partition key> = <constant>, it would have to visit all partitions instead of just one, and those queries would take much longer to execute. You can make the index global, but there all kind of maintenance issues with that. In short: be careful about partitioning and think it through first.
> But is it a good option to partition tables with
total no. of rows between 2 million and 5 million ? I
thought partitioning is better done for tables > 20
million.
Please throw some light on that.
Generally, partitioning is done on bigger tables, yes. But it is not about some specific number of rows. You partition a table when all forecasted use of the table is known, so you can make a good judgement which usages will benefit, which ones won't be affected and which ones are likely to suffer. When the benefits outweigh the disadvantages, you partition...
Above I had an example about queries that are going to suffer. Partitioning is beneficial for example when once in some period you'd have to delete A LOT of old rows. Or when you have some data warehousing queries that need like 10% of only the most recent data in a huge table.
The decision is yours.
Regards,
Rob.

Need help on this query

hi ,
I have the following data and the expected results.
Have managed to return the results in a procedure but the performance is not good.
Could a analytic be used instead ? need advise
logic
1. if all type is of run then i'll simply get the id and its seq
2. however , if the type is a if ...else then i'll need to check against table B and whether passing the if else will determine the seq to be used and if it passes the if then it'll go to the next id, if not if will go the else until it reaches the end of that section of if .. else for the same id before going to the next id
table A
id     seq     type          checkname     checkvalue     check          fail          nextseq
id0     1     run
id0     2     run
id0     3     run
id1     1     if          ac          5          =          3
id1     2     run                                                  end
id1     3     elseif          ac          6          >=          4
id1     4     run
id1     5     else                                                  end
id1     6     run
id2     1     run
id2     2     run
id2     3     if          an          5          <=          5
id2     4     run
id2     5     else                                                  end
id2     6     run
scenario 1
table B
id     checkname     val
id0     aa          0
id1     ac          2
id1     ab          8
id2     ab          4
id     an          5
RESULTS :
id     seq
id0     1
id0     2
id0     3
id1     6
id2     1
id2     2
id2     4
scenario 2
table B
id     checkname     val
id0     aa          0
id1     ac          5
id1     ab          8
id2     ab          4
id     an          6
RESULTS :
id     seq
id0     1
id0     2
id0     3
id1     2
id2     1
id2     2
id2     6appreciate ur advise
tks & rgds

ok,
of course you can do it with model
but still it can be done with analytic functions.
one more question
what is the condition to be checked?
a.checkvalue || a.check || b.val
or
b.val || a.check || a.checkvalue
accroding to your OP it seems the second choise is correct, isn't it?
SQL> -- first scenario
SQL>
SQL> with tableA as (select 'id0' id, 1 seq, 'run' type, null checkname, null checkvalue, null check_, null fail, null nextseq from dual union all
2                   select 'id0',    2,   'run',       null,null,null,null,null from dual union all
3                   select 'id0',    3,   'run',       null,null,null,null,null from dual union all
4                   select 'id1',    1,   'if',        'ac',5,'=',3,null from dual union all
5                   select 'id1',    2,   'run',       null,null,null,null,'end' from dual union all
6                   select 'id1',    3,   'elseif',    'ac',6,'>=',4,null from dual union all
7                   select 'id1',    4,   'run',       null,null,null,null,null from dual union all
8                   select 'id1',    5,   'else',      null,null,null,null,'end' from dual union all
9                   select 'id1',    6,   'run',       null,null,null,null,null from dual union all
10                   select 'id2',    1,   'run',       null,null,null,null,null from dual union all
11                   select 'id2',    2,   'run',       null,null,null,null,null from dual union all
12                   select 'id2',    3,   'if',        'an',5,'<=',5,null from dual union all
13                   select 'id2',    4,   'run',       null,null,null,null,null from dual union all
14                   select 'id2',    5,   'else',      null,null,null,null,'end' from dual union all
15                   select 'id2',    6,   'run',       null,null,null,null,null from dual),
16 tableB as (select 'id0' id, 'aa' checkname, 0 val from dual union all
17             select 'id1', 'ac', 2 from dual union all
18             select 'id1', 'ab', 8 from dual union all
19             select 'id2', 'ab', 4 from dual union all
20             select 'id2', 'an', 5 from dual)
21             --
22             select id, seq from (select tt.*, first_value(decode(flag,1,seq) ignore nulls) over (partition by id order by seq) FV from
23             (select a.*, b.val, lag(nvl(nvl2(val,case check_ when '=' then decode(sign(val-checkvalue),0,1,0)
24                                          when '>=' then decode(sign(val-checkvalue),-1,0,1)
25                                          when '>' then decode(sign(val-checkvalue),1,1,0)
26                                          when '<=' then decode(sign(val-checkvalue),1,0,1)
27                                          when '<' then decode(sign(val-checkvalue),-1,1,0)
28                                     end, decode(type,'else',1)),val)) over (partition by a.id order by seq) flag, lag(type,1,type) over (partition by a.id order by seq) lag_type
29                              from tableA a, tableB b
30              where a.id=b.id(+)
31                and a.checkname=b.checkname(+)
32                order by 1,2) tt
33 ) where (type='run' and lag_type='run') or seq=FV
34 /
ID         SEQ
id0          1
id0          2
id0          3
id1          6
id2          1
id2          2
id2          4
7 rows selected
SQL> --second scenario
SQL>
SQL> with tableA as (select 'id0' id, 1 seq, 'run' type, null checkname, null checkvalue, null check_, null fail, null nextseq from dual union all
2                   select 'id0',    2,   'run',       null,null,null,null,null from dual union all
3                   select 'id0',    3,   'run',       null,null,null,null,null from dual union all
4                   select 'id1',    1,   'if',        'ac',5,'=',3,null from dual union all
5                   select 'id1',    2,   'run',       null,null,null,null,'end' from dual union all
6                   select 'id1',    3,   'elseif',    'ac',6,'>=',4,null from dual union all
7                   select 'id1',    4,   'run',       null,null,null,null,null from dual union all
8                   select 'id1',    5,   'else',      null,null,null,null,'end' from dual union all
9                   select 'id1',    6,   'run',       null,null,null,null,null from dual union all
10                   select 'id2',    1,   'run',       null,null,null,null,null from dual union all
11                   select 'id2',    2,   'run',       null,null,null,null,null from dual union all
12                   select 'id2',    3,   'if',        'an',5,'<=',5,null from dual union all
13                   select 'id2',    4,   'run',       null,null,null,null,null from dual union all
14                   select 'id2',    5,   'else',      null,null,null,null,'end' from dual union all
15                   select 'id2',    6,   'run',       null,null,null,null,null from dual),
16 tableB as (select 'id0' id, 'aa' checkname, 0 val from dual union all
17             select 'id1', 'ac', 5 from dual union all
18             select 'id1', 'ab', 8 from dual union all
19             select 'id2', 'ab', 4 from dual union all
20             select 'id2', 'an', 6 from dual)
21             --
22             select id, seq from (select tt.*, first_value(decode(flag,1,seq) ignore nulls) over (partition by id order by seq) FV from
23             (select a.*, b.val, lag(nvl(nvl2(val,case check_ when '=' then decode(sign(val-checkvalue),0,1,0)
24                                          when '>=' then decode(sign(val-checkvalue),-1,0,1)
25                                          when '>' then decode(sign(val-checkvalue),1,1,0)
26                                          when '<=' then decode(sign(val-checkvalue),1,0,1)
27                                          when '<' then decode(sign(val-checkvalue),-1,1,0)
28                                     end, decode(type,'else',1)),val)) over (partition by a.id order by seq) flag, lag(type,1,type) over (partition by a.id order by seq) lag_type
29                              from tableA a, tableB b
30              where a.id=b.id(+)
31                and a.checkname=b.checkname(+)
32                order by 1,2) tt
33 ) where (type='run' and lag_type='run') or seq=FV
34 /
ID         SEQ
id0          1
id0          2
id0          3
id1          2
id2          1
id2          2
id2          6
7 rows selected
SQL> Message was edited by:
Volder
1> In my query I don't use FAIL column
so I go consequently throug SEQ column order.
Using MODEL - you could use FAIL values as dimension values for getting the next row needed.
2> In my query If there're no needed values in TABLE B - the else part of your IF_CLAUSES will be triggered.
3> Instead of huge case you can just create function that will get inputted 3 params: 2 values and check operator - and will return 1 if true, 0 if it is false.
Just take these comments into account.

Need help over this Query

Hi, I need the Result set in the given below form.
Can any one please help me with this.
I am not able to get the 2nd most recent login date in the separate column.
Concerned Columns: Login date, login date2, login difference
Login date = Most recent login by the user
Login date 2= 2nd most recent login by the user
Login difference = difference between last 2 logins
Help!!
DECLARE @ownerid INT=304,@offset INT=330,@totalsec INT=86400
SELECT USERNAME, LOGIN_ID,EMAIL,DEPARTMENT,LOGIN_DATE, LOGOUT_DATE, IP_ADDRESS,MACHINE_NAME,TOTAL_LOGIN_TIME,LOGIN_NUMBER,
ROW_NUMBER() OVER (ORDER BY USERNAME, LOGIN_ID) AS ROWNUMBER
FROM
SELECT Usr.UserID [USERID], usr.UserName [USERNAME],usr.LoginID [LOGIN_ID],
usrcontact.Email [EMAIL],
UsrContact.Department [DEPARTMENT],
MAX(DateAdd(mi,@offset, UsrHis.LoginDate)) [LOGIN_DATE],
MAX(DateAdd(mi,@offset, UsrHis.LogoutDate)) [LOGOUT_DATE],
UsrHis.IPAddress [IP_ADDRESS],UsrHis.MachineName [MACHINE_NAME],
convert( NVARCHAR(50),(DATEDIFF(SECOND,UsrHis.LoginDate,UsrHis.LogoutDate))/@totalsec) + ' '+'Days'+ ' '+
convert( NVARCHAR(50),((DATEDIFF(SECOND,UsrHis.LoginDate,UsrHis.LogoutDate))%@totalsec)/3600) + ' '+'hours'+ ' '+
convert( NVARCHAR(50),(((DATEDIFF(SECOND,UsrHis.LoginDate,UsrHis.LogoutDate))%@totalsec)%3600/60)) + ' '+'Minutes'+ ' '+
convert( NVARCHAR(50),(((DATEDIFF(SECOND,UsrHis.LoginDate,UsrHis.LogoutDate))%@totalsec)%3600%60)) + ' '+'Seconds'
[TOTAL_LOGIN_TIME],
ROW_NUMBER() OVER (PARTITION BY USR.USERID ORDER BY USRHIS.LOGINDATE) LOGIN_NUMBER
FROM Az_User Usr Inner Join LoginHistory UsrHis
ON UsrHis.OwnerID=Usr.AppOwnerID
and
Usr.UserID = UsrHis.UserID
inner join UserContact UsrContact
ON UsrContact.CompanyID = Usr.AppOwnerID
and UsrContact.UserID = Usr.UserID
WHERE usr.AppOwnerID = @ownerid
GROUP BY Usr.UserID,Usr.UserName , usrcontact.Email , UsrHis.LoginDate,usr.LoginID,UsrContact.Department,
UsrHis.LogoutDate,UsrHis.IPAddress,UsrHis.MachineName,
Dateadd(D, Datediff(D, 0, Dateadd(mi, @offset, UsrHis.LoginDate)), 0)
) AS TEMP
WHERE TEMP.LOGIN_NUMBER <=2
Hi here is the Data type of the columns i am selecting.. and the result i am getting.
But here the total_login_time is wrong. .
I need difference between last 2 logins not length of the session.

It is unclear what you are trying to achieve...
DATEDIFF(d,Logindate,Logindate2)
AS [Login difference]
FROM.....
Best Regards,Uri Dimant SQL Server MVP,
http://sqlblog.com/blogs/uri_dimant/
MS SQL optimization: MS SQL Development and Optimization
MS SQL Consulting:
Large scale of database and data cleansing
Remote DBA Services:
Improves MS SQL Database Performance
SQL Server Integration Services:
Business Intelligence

Need help with this query

Hi,
I am using SQL Server 2008 Enterprise edition 64 bit on Windows serer 2008 enterprise edition 64 bit.
The below query works fine. I am trying to rewrite the code without using the sub query. In the end i should get the same results for both the query which we are going to rewrite and the below mentioned query.
Can any help me please.
SELECT
t1.DOCUMENT_NO,
RTRIM(CAST(t1."ENVIRONMENT_CD" AS VARCHAR(5))) + '*' + RTRIM(CAST(t1."ORDER_NO" AS VARCHAR(25)))
+
CASE WHEN (SELECT COUNT(DISTINCT S1.TEST_REPORTING_MATERIAL_SID)
FROM dbo.TBLTEST S1 ON S1.DOCUMENT_NO = T1.DOCUMENT_NO
WHERE S1.TEST_REPORTING_MATERIAL_SID > 0
AND S1.COMPANY_CD = T1.COMPANY_CD AND S1.INVOICE_SEQ_NO = T1.INVOICE_SEQ_NO
AND S1.DOCUMENT_ITEM_NO = T1.DOCUMENT_ITEM_NO
AND S1.ORDER_NO = T1.ORDER_NO) > 1 THEN
RTRIM(CAST(t1."TEST_REPORTING_MATERIAL_SID" AS VARCHAR(10)))
ELSE
END as sum_key,
t1.SALES_ANALYSIS_CD
FROM
dbo.TBLTEST_ALL T1
WHERE
t1.TIME_SID >= 20001001 ;
I tried to use a left outer Join , it did not work out. I tried to take the subquery and assign it to a variable and use in the above mentioned query. But i am not getting the same results.
I don't want to use the above query so i am planning to rewrite the code.
Thank You,

I doubt that the query you posted works fine, because:
FROM dbo.TBLTEST S1 ON S1.DOCUMENT_NO = T1.DOCUMENT_NO
This is a syntax error. I will have to assume that your inteded query reads:
SELECT t1.DOCUMENT_NO,
   rtrim(cast(t1."ENVIRONMENT_CD" AS varchar(5))) + '*' +
   rtrim(cast(t1."ORDER_NO" AS varchar(25))) +
CASE WHEN (SELECT COUNT(DISTINCT S1.TEST_REPORTING_MATERIAL_SID)
   FROM   dbo.TBLTEST S1
   WHERE S1.DOCUMENT_NO = T1.DOCUMENT_NO
   AND S1.TEST_REPORTING_MATERIAL_SID > 0
   AND S1.COMPANY_CD = T1.COMPANY_CD
   AND S1.INVOICE_SEQ_NO = T1.INVOICE_SEQ_NO
   AND S1.DOCUMENT_ITEM_NO = T1.DOCUMENT_ITEM_NO
   AND S1.ORDER_NO = T1.ORDER_NO) > 1 THEN
   RTRIM(CAST(t1."TEST_REPORTING_MATERIAL_SID" AS VARCHAR(10)))
   ELSE '*-'
END as sum_key, t1.SALES_ANALYSIS_CD
FROM   dbo.TBLTEST_ALL T1
WHERE t1.TIME_SID >= 20001001
Here is a query with a left join. Whether it is actually better than the one you have I am not sure.
SELECT t1.DOCUMENT_NO,
   rtrim(cast(t1."ENVIRONMENT_CD" AS varchar(5))) + '*' +
   rtrim(cast(t1."ORDER_NO" AS varchar(25))) +
CASE WHEN S1."EXISTS" IS NOT NULL
   THEN RTRIM(CAST(t1."TEST_REPORTING_MATERIAL_SID" AS varchar(10)))
   ELSE '*-'
END as sum_key, t1.SALES_ANALYSIS_CD
FROM   dbo.TBLTEST_ALL T1
LEFT   JOIN (SELECT 1 AS "EXISTS"
   FROM   dbo.TBLTEST
   WHERE TEST_REPORTING_MATERIAL_SID > 0
   GROUP BY DOCUMENT_NO, COMPANY_CD, INVOICE_SEQ_NO,
   DOCUMENT_ITEM_NO, ORDER_NO
   HAVING COUNT(DISTINCT TEST_REPORTING_MATERIAL_SID) > 1) AS S1
   ON S1.DOCUMENT_NO = T1.DOCUMENT_NO
   AND S1.COMPANY_CD = T1.COMPANY_CD
   AND S1.INVOICE_SEQ_NO = T1.INVOICE_SEQ_NO
   AND S1.DOCUMENT_ITEM_NO = T1.DOCUMENT_ITEM_NO
   AND S1.ORDER_NO = T1.ORDER_NO
WHERE t1.TIME_SID >= 20001001
Erland Sommarskog, SQL Server MVP, [email protected]

SQL I need help with this query Please help

List the names of all the products whose weight unit measure is “Gram”. Order the list by product name. Do not use JOINS, but use the IN clause with a sub-query.
select Name
from UnitMeasure
where Name= 'Gram'
order by Name
I did this, but it seem that the requirement is different.

As a guess:
Select Name from Product
where UnitMeasure in (Select Name from unitmeasure where name = 'Gram')
Andy Tauber
Data Architect
The Vancouver Clinic
Website | LinkedIn
This posting is provided "AS IS" with no warranties, and confers no rights. Please remember to click
"Mark as Answer" and "Vote as Helpful" on posts that help you. This can be beneficial to other community members reading the thread.

Physician/Researcher needs help with this query

Folks,
Thank you in advance for looking at my problem. You are helping to save lives and make this world a better place.
I am trying to find out the earliest date patients took steroid. I have:
select patientid, visitnumber, steroid_start_date from treatments where steroid_start_date is not null and trim(steroid_start_date) != '/'
order by patientid, substr(steroid_start_date, -4)
I then have:
patientid visitnumber steroid_start_date(this is a string column, not date or time, and it has so many junk in it, like /2003 , /)
4 3 03/2004
4 2 01/01/2005
10 2 08/01/2005
10 1 05/01/2002
What I need is:
4 3 03/2004
10 1 05/01/2002
I am not good with group by... having.... max.... How can I limit to one patient per row and this row is their earliest visit?
Thank you very much in advance. You are helping to save lives.

Here it is ->
satyaki>
satyaki>select * from v$version;
BANNER
Oracle Database 10g Enterprise Edition Release 10.2.0.3.0 - Prod
PL/SQL Release 10.2.0.3.0 - Production
CORE    10.2.0.3.0      Production
TNS for 32-bit Windows: Version 10.2.0.3.0 - Production
NLSRTL Version 10.2.0.3.0 - Production
Elapsed: 00:00:00.53
satyaki>
satyaki>
satyaki>with t
2 as
3    (
4       select 4 patientid, 3 visitnumber, '03/2004' steroid_start_date from dual
5       union all
6       select 4, 2, '01/01/2005' from dual
7       union all
8       select 10, 2, '08/01/2005' from dual
9       union all
10       select 10, 1, '05/01/2002' from dual
11    )
12 select patientid,
13         visitnumber,
14         steroid_start_date
15 from (
16         select k.*,
17                row_number() over(partition by patientid order by to_date(substr(steroid_start_date,-4),'YYYY')) rn
18         from t k
19         order by patientid
20       )
21 where rn=1;
PATIENTID VISITNUMBER STEROID_ST
         4           3 03/2004
        10           1 05/01/2002
Elapsed: 00:00:00.11
satyaki>Regards.
Satyaki De.

I need information on this Object

Hi, can one help me in this object.
On a monthly basis, after all data has been fed into One Source and the monthly One Source calculations have been performed, an extract will be run from One Source , converted via XI and loaded to BW. The current plants that will be extracted from One Source are: 6101, 6102, 6103, 6201, 6202, 6203, 6301, 6302, 6601, and 6602. As additional plants get converted to SAP; they will automatically start appearing on this extract. The One Source extract file will be pulled from a corporate server for processing. (One source is the Data Warehouse)

Hi Rahul,
what expecting you exctly, hot to aciheve your requiremnt in XI.
Suppose oneSource is File System.
you want to send data to BI. Use File to ABAP Proxy .
LET ME KNOW YOUR REQUIREMENT EXACTLY.
Regards,
Raj

I need Information on this intresting OBJECT

On a monthly basis, after all data has been fed into One Source and the monthly One Source calculations have been performed, an extract will be run from One Source , converted via XI and loaded to BW. The current plants that will be extracted from One Source are: 6101, 6102, 6103, 6201, 6202, 6203, 6301, 6302, 6601, and 6602. As additional plants get converted to SAP; they will automatically start appearing on this extract. The One Source extract file will be pulled from a corporate server for processing. (One source is the Data Warehouse)

Please post this in BI section

Need information on this Query

Similar Messages

Maybe you are looking for