Delete duplicate entries with additional if conditions

[Tita]

I am trying to delete HHID duplicate code to leave only Households (not all members of the household). But i need to not just blinding delete the duplicate one, but i have to create a condition that if the person is the financial respondent of the family, then delete the other person, if not then delete person #1. How can i remove duplicate with an embedded if statement?

Thanks,

Tita

[Steelers_In_DC]

It's kind of hard to guess what your data looks like without an example.  You'd be better off adding a small sample of what you have and again what you want:

data have;

input hhid$ person$ head$;

cards;

01 lammar 0

01 sally 0

01 renoldo 1

02 Judy 1

02 Leroy 0

02 Jane 0

;

data responsible not;

set have;

if head = 1 then output responsible;

else output not;

run;

[Tita]

Hi Mark, Thanks for the response. I have the following dataset:

HHID RespID SpouseID  Couple  Respondent(Fin resp)  Spouse(Fin Resp)  And 600 other variables

20       2020      2010          1                 1                                 0                           ,,,,,,

20       2010      2020          1                  0                                1                           ........

10        1010      0               0                   1                               0                            ......

30          3010    3020          1                   1                               0                         ......

30          3020     3010         1                   0                                1                        ......

I want to have the following:

HHID RespID SpouseID  Couple  Respondent(Fin resp)  Spouse(Fin Resp)  And 600 other variables

20       2020      2010          1                 1                                 0                           ,,,,,,

10        1010      0               0                   1                               0                            ......

30          3010    3020          1                   1                               0                         ......

Leave only one household who will be the financial respondent if coupled. So i will have two statements :

if Couple=0, it is not duplicate, leave it like that;

else if couple=1 and Respond(fin)=1 then delete the spouse row (duplicate one);

run;

[Steelers_In_DC]

Here is a solution for you:

data have;

input HHID RespID SpouseID  Couple  'Respondent(Fin resp)'n  'Spouse(Fin Resp)'n;

cards;

20       2020      2010         1                 1                               0

20       2010      2020         1                 0                               1

10       1010      0            0                 1                               0

30       3010      3020         1                 1                               0

30       3020      3010         1                 0                               1

;

data want rest;

set have;

if couple = 0 then output want;

else if couple = 1 and 'Respondent(Fin resp)'n = 1 then output want;

else output rest;

run;

Go Pirates