How do I identify multiple records within a row

CathyVI — Wed, 27 Jan 2021 17:12:08 GMT

I have to following synthetic data similar to my data. I have millions of records so this is demo I created.

data test;

input Name $ sex $ weight height income;

cards;

Bob M 1 0 0

Mat M 0 1 0

Cat F 1 0 1

Dog F 1 0 0

Fred M 0 1 1

Sam M 0 0 1

Pat F 0 0 0

Jack M 1 1 1

;

run;

My goal is to identify when a row has more than 1 record within the variable weight, height and income.

For example the cat, fred and jack rows all have 1 record for each variable, how will I find these values. I want to identify row with more than 1 record separately from row with only 1 record.

So I want the count of cat, fred and jack separately and the count of bob, mat, dog, sam separately. I already identify the latter using this code;

data check;

set test;

if weight =1 or height =1 or income=1 then total_all=1;

else total_all=0;

run;

Re: How do I identify multiple records within a row

Reeza — Wed, 27 Jan 2021 17:19:29 GMT

First identify which group a record belongs to and then you can determine later how to summarize them separately. You don't indicate how that would work so I'm guessing. It would have also been really helpful to see what you expect as output from this input. If my solution doesn't work, please post what you expect as output.

data want;
set test;

if sum(weight, height, income)>1 then group = "2";
else group = "1";

run;

proc freq data=want;
table group;
table group*name / list;
run;

@CathyVI wrote:

I have to following synthetic data similar to my data. I have millions of records so this is demo I created.

data test;

input Name $ sex $ weight height income;

cards;

Bob M 1 0 0

Mat M 0 1 0

Cat F 1 0 1

Dog F 1 0 0

Fred M 0 1 1

Sam M 0 0 1

Pat F 0 0 0

Jack M 1 1 1

;

run;

My goal is to identify when a row has more than 1 record within the variable weight, height and income.

For example the cat, fred and jack rows all have 1 record for each variable, how will I find these values. I want to identify row with more than 1 record separately from row with only 1 record.

So I want the count of cat, fred and jack separately and the count of bob, mat, dog, sam separately. I already identify the latter using this code;

data check;

set test;

if weight =1 or height =1 or income=1 then total_all=1;

else total_all=0;

run;

topic Re: How do I identify multiple records within a row in SAS Programming

How do I identify multiple records within a row

Re: How do I identify multiple records within a row