So if you have records which are something like this:
ID NAME IDS
--- ------- ------
1 Issam 649874 54898 3128A 848E
2 Issam 3128A
3 Issam 649874 54898 3128A 848E
4 Issam 251989898F 3128A 848E
5 Issam 987785
6 Issam null
7 NABIL 987785
8 NABIL 98778587
1, 2, 3, 4 ARE duplicates (since they are have a common ID, at least one).
7 and 8 ARE NOT
The rest are just other records.
Blocking column: NAME
Match-Commands column: IDS
What kind of comparison type would you choose? And what the param/other options you would specify for such a case.
Thanks.
Match Designer - Comparision Types
Match Designer - Comparision Types
Not only thoughts, but a little bit of experience.
-
- Participant
- Posts: 527
- Joined: Thu Apr 19, 2007 1:25 am
- Location: Melbourne
Based on the example you have there, I'd not use QS.
You want one of those IDs to match exactly.
I'd:
- do a pivot to get ID, NAME, ID2 (1 row for each of those IDs)
- sort on NAME,ID2 with a key change column
- add a SetID, incrementing when key_change is 1
- join distinct ID per SetID back to original data
You want one of those IDs to match exactly.
I'd:
- do a pivot to get ID, NAME, ID2 (1 row for each of those IDs)
- sort on NAME,ID2 with a key change column
- add a SetID, incrementing when key_change is 1
- join distinct ID per SetID back to original data