How to flag observations if all values of a variable within a by group are the same

[KPCklebspn]

Hi there,

I have a dataset that looks like this

ID    varA   

1       0

1       0

2       1

2       1

3       0

3       1

I want it to look like this (i.e. flag=1 if all values of varA are the same within ID by group):

ID    varA   flag

1       0        1

1       0        1

2       1        1

2       1        1

3       0        0

3       1        0

How would I do this? Thanks

[PeterClemmensen · Accepted solution]

Try this

data have;
input ID varA;
datalines;
1 0
1 0
2 1
2 1
3 0
3 1
;

proc sql;
   create table want as
   select *, count(distinct varA)=1 as flag 
   from have
   group by ID;
quit;

[KPCklebspn]

Thank you, worked! can you pls explain what this bit of syntax means:

count(distinct varA)=1 

[PaigeMiller]

proc summary data=have nway;
    class id;
    var vara;
    output out=stats std=std;
run;

data want;
    merge have stats;
    by id;
    if std=0 then flag=1;
    drop std;
run;

[novinosrin]

data have;
input ID varA;
datalines;
1 0
1 0
2 1
2 1
3 0
3 1
;
proc sql;
   create table want as
   select *, std(vara)=0 as flag 
   from have
   group by ID;
quit;