Libardo,
Your problem description is not so clear. I'm going to make some assumptions
which you can confirm or reject.
Your incoming file actually looks like
ID Name
1445566 AAA
1445556 AAA
2222222 CCC
22222222 CCC
123 Peter
123 Pefer
4546 Floyd Dockter
4546 Floyd Fockter
Most important, you always have exactly two records that need to be
compared. Each record has two variables to check. As you show, you want to
know whether the ids match and whether the names match.
The simplest way to do this may be to use the built in faciliy that spss
has. I don't know when it was first included but in 15 it is called Identify
duplicate cases and is in the Data dropdown menu.
If you don't have that facility, then here is how to do it. I also notice
that you want both records in a set to be marked. This is a bit tricky
because you don't have valid case id numbers. You'll need to create them
first. Note that I assume that an error exists and then change the marker
variable if it does not. That should catch missing values as well as
nonmatching ones.
compute case=$casenum.
if (mod(case,2) eq 0) case=lag(case)
Compute errorid=1.
Compute errorname=1.
If (id eq lag(id)) errorid=0.
If (name eq lag(name)) errorname=0.
Sort cases by case errorid.
If (case eq lag(case) and lag(errorid) eq 0) errorid=0.
Sort cases by case errorname.
If (case eq lag(case) and lag(errorname) eq 0) errorname=0.
Gene Maguin
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
