distinct on selected variables and remaining as it is

[alexdsa310]

is there any way in proc sql if i can select distinct on few varibles and have remaining as it is.

as below i do not want distinct on b.term,b.acce and need for the rest

proc sql;
create table _values_ as
select (distinct b.table,b.column,b.origin,b.originwhere,b.algorithm), b.term,b.acce
from vcol a right join source_columns b
on a.origin=b.origin

quit;

any suggestion please

[Kurt_Bremser]

How about

proc sql;
create table _values_ as
select b.table,b.column,b.origin,b.originwhere,b.algorithm, b.term,b.acce
from vcol a right join source_columns b
on a.origin=b.origin
group by b.table,b.column,b.origin,b.originwhere,b.algorithm
;
quit;

?

[alexdsa310]

No need to mention distinct? sorry to ask

[Kurt_Bremser · Accepted solution]

As soon as the distinct does not extend over all columns, it becomes meaningless, as all rows from the dataset will be included anyway. Play around with the concept for a while, and you'll see what I mean.