Pick Unique ID per Unique ID

[SteveNZ]

HI All,

I have a data set where I've matched on customer number (CRN), date (idate) and amount (fullpay) from two tables which has created a many to many data set where there are same payments on the same day.

For example if there were 3 payments in my full payment table and 3 payments in my deduction table I get nine lines (see below). Is there a way I can get a unique MATCH_ID for each unique PAID_ID?

This is what I have:

data have;
	length crn paid_id match_id fullpay deduct issued idate 8;
	informat idate anydtdte21.;
	format idate ddmmyy10.;
	infile datalines delimiter=',';
	input crn paid_id match_id fullpay deduct issued idate;
	datalines;
123456,4471,53360,66.03,66.03,0,5/09/2019
123456,4471,53361,66.03,41.07,24.96,5/09/2019
123456,4471,53362,66.03,0,66.03,5/09/2019
123456,4472,53360,66.03,66.03,0,5/09/2019
123456,4472,53361,66.03,41.07,24.96,5/09/2019
123456,4472,53362,66.03,0,66.03,5/09/2019
123456,4473,53360,66.03,66.03,0,5/09/2019
123456,4473,53361,66.03,41.07,24.96,5/09/2019
123456,4473,53362,66.03,0,66.03,5/09/2019
654321,11827,184385,57.4,16.8,40.6,15/03/2019
654321,11828,184385,57.4,16.8,40.6,15/03/2019
654321,11826,184385,57.4,16.8,40.6,15/03/2019
654321,11827,184384,57.4,57.4,0,15/03/2019
654321,11828,184384,57.4,57.4,0,15/03/2019
654321,11826,184384,57.4,57.4,0,15/03/2019
654321,11827,184383,57.4,57.4,0,15/03/2019
654321,11828,184383,57.4,57.4,0,15/03/2019
654321,11826,184383,57.4,57.4,0,15/03/2019
;

and this is what I need:

data want;
	length crn paid_id match_id fullpay deduct issued idate 8;
	informat idate anydtdte21.;
	format idate ddmmyy10.;
	infile datalines delimiter=',';
	input crn paid_id match_id fullpay deduct issued idate;
	datalines;
123456,4471,53360,66.03,66.03,0,5/09/2019
123456,4472,53361,66.03,41.07,24.96,5/09/2019
123456,4473,53362,66.03,0,66.03,5/09/2019
654321,11826,184383,57.4,57.4,0,15/03/2019
654321,11827,184384,57.4,57.4,0,15/03/2019
654321,11828,184385,57.4,16.8,40.6,15/03/2019
;

Any help gratefully received and many thanks in advance

cheers

Steve

[Kurt_Bremser]

I guess this should be fixed while you do the initial join. Could you show us the corresponding entries in the two tables from which you created "have"?

[PeterClemmensen]

So do you want to resolve this problem in your initial join or from have to want?

[Patrick]

@SteveNZ 

I really appreciate that you've posted the data as working SAS data steps.

You need to fix the join so that it's no more many:many

For your sample data show us what you've got in the two source tables.

[Jagadishkatam · Accepted solution]

please try below code

proc sort data=have;
by idate crn match_id  ;
run;

data want;
set have;
by idate crn match_id ;
if first.match_id;
run;

[SteveNZ]

Hi Jag,

This appears to have solved it at first glance (large amount of data). Thanks so much, I think I was over complicating things a bit my end but running your code seems to do the trick!

cheers

Steve