cancel
Showing results for 
Search instead for 
Did you mean: 

select unique records in data services

Former Member
0 Kudos

Hi All!

Can I select unique records in data services ?

Example : 2 records in file

key1 100

key2 200

key -primary key in table.  I need only record2 in table.

When I use Query - and flag distinct - it doesn't work.

Accepted Solutions (1)

Accepted Solutions (1)

abhi2636
Participant
0 Kudos

Hi Zohin,

Can you please elaborate your requirement. Because 'distinct' is used to get a unique record from bunch of duplicate records. In your scenario we cannot use 'distinct' to get unique record. If you provide more information on your source data that might help to find a solution.

Regards,

Abhi

Former Member
0 Kudos

Hi Abhi!Thanks for your reply!

I want to do like InfoSourse in BW system! I have duplicate records in sourse (file or table) and in target (dataSourse in BW ) I want to see only 1 record.

I don't want to create Infosourse in BW

DS - > DataSourse BW - > Attribute  (NOT DS - > DataSourse BW - >IS ->  Attribute)

abhi2636
Participant
0 Kudos

Hi Zohin,

Try by enabling 'Auto Correct Load' option to 'Yes' in target editor properties under 'Options' tab as in figure.

Regards,

Abhi.

Former Member
0 Kudos

Tnank you!!!

Answers (3)

Answers (3)

Former Member
0 Kudos

Hi,Abhi!


When I try to use table generated from BW ( Data Source )  as target -  therу is NO  'Auto Correct Load'  option. 

And I have doubles records as result (


Query from file format


Query - BW target




Regards!

abhi2636
Participant
0 Kudos

Hi Zlobin,

There is no 'Auto Correct load' option for BW Datasource.

I think you can first load data into a table with 'Auto correct load' option enables in a dataflow and use this table as source in another dataflow and load into BW datasource.

Regards,

Abhi

ahalya_inturi
Active Participant
0 Kudos

Hi ,

If you want to pass only key 2 record to target

then apply condition in where clause in query transform of data services

where key_field = key2

Former Member
0 Kudos

Hi! No ) a lot of records and maybe duplicate records

former_member290153
Participant
0 Kudos

Hi Alexander,

Kindly go through the provided link ...


Regards,

Priyanka.